The Australian Monograph no. 22 Notional University Development Studies Centre and The University of the South Pacific Centre Food in Fiji: the for Applied Studies in Development produce and processed foods distribution systems

Michael W.P. Baxter Food in Fiji This monograph , plus two others , is a joint publication of The Un iversity of the South Pacific 's Cen tre for Applied Studies in Development and The Australian National Un iver­ sity ' s Development Studies Centre. The research on which this monograph is based was ftmded by the International Development Research Centre (IDRC) through a grant to The University of the South Pacific (USP) and the volume is published with the support of the IDRC .

ii Development StudiesCentre Monograph no. 22

Food in Fiji: The produce and processed foods distribution systems

Michael W.P. Baxter

This monograph is published in association with the Centre for Applied Studies in Development, Universityof the South Pacific, Suva, Fiji.

Series editor Gavin W. Jones

The Australian National University Canberra 1980 © Michael Baxter 1980

This work is copyright. Apart from any fair dealing for the purpose of private study, research, criticism or review, as permitted under the Copyright Act, no part may be -reproduced by any process without written permission. Inquiries may be made to the publisher.

National Library of Australia Cataloguing-in-publication entry

Baxter, Michael W .P. Food in Fiji.

{Development Studies Centre monograph; no 22 )· iSBN 0 909150 03 6

1. Food Industry and Trade - Fiji. I. Title. (Series: Australian National University, Canberra. Development Studies Centre. Monograph; no. 22 ISSN 0157-5767)

338.4'76413'0099611

Printed and manufactured in Australia by The Australian National University Sunnnary

Foodstuff imports, the involvement of racial groups - Fijians, Indians, Chinese and Europeans - in conunerce, and government policy affecting both these areas are basic concerns of this analysis of Fiji's food distribution system. Food consumption habits in Fiji are the product of historic trends and cultural bias as well as economic rationalization. Save by a ban on imports, food imports cannot be decreased tmless these factors are taken into accotmt. The food distribution system has two discrete parts - one for produce and the other for processed foods.

Produce enters connnercial trade through either public urban retail markets or the government wholesale/retail National Marketing Authority. There is little vertical trade development. Markets are patronized by connnercial and subsistence producer-vendors and also by non-producer vendors. The NMA was established in 1971, largely to increase efficiency of the internal produce market; in fact it has had limited success in breaking into the internal market and has concentrated on foodstuff exports. Fijians, Indians and Chinese participate in produce production and marketing and each has a distinctive role. Fijian participation in marketing is increasing, especially as non-producer vendors. Selling in markets is an effective connnercial training grotmd for Fijians as it overcomes key handicaps to their involvement in connnerce.

The processed food distribution system has few entry restrictions and there are many participants, especially at the retail level. It is generally under-capitalized and is a back-order system. There is little government regulation, except of prices, although these controls have the additional objective of influencing the structure of the trading system. Racial involvement is as varied as in the produce trade. Importing and wholesaling is handled largely by Europeans, Chinese and Indians. Fijians are deeply involved in retail­ ing: about 30 per cent of retail outlets are operated by Fijian individuals� (An additional 20 per cent are operated by Fijian co-operative societies.) The government has a

v vi

number of schemes to promote Fijian involvement in commerce in general and in the processed food trade in particular but these do not overcome the spatial and cultural disad­ vantages facing Fijians.

Ways are suggested to increase the efficiency of both the produce and processed food systems and to gain in them more equitable racial participation. Pref ace

The Pacific Towns and Rural-urban Distribution Systems Project is a two-year applied research and training program based at the University of the South Pacific, Suva, Fij i. With funds from the Internat ional Development Research Centre of Ottawa, Canada, the University of the South Pacific appointed two research fellows and three assistants to investigate four questions :

(a) What are the existing distribution systems operating within towns and linking towns to rural areas ? (b) What is the economic role of in digenous and non­ indigenous peoples in these systems ? (c) What are the current governmen t policies towards indigenous involvement in the distribution systems . and do government attitudes or regulations restrict such involvement? (d) What are the developmental implications of govern­ ment policies regarding the increased participation of indigenous groups in the distribut ion systems ?

Research was conducted in Fij i and in Tonga from August 1975: the Director of research in Fij i was Michael Baxter and the Director in Tonga was Epeli Hau'ofa. A similar but independent study in the New Hebrides was conducted by Pro­ fessor R. G. Ward and others of the Australian. Nat ional Un i­ versity. The findings of each country study will be released separately . Research in the three study countries was con­ ducted independently , although it focused on a similar 'basic problem' core. In all areas , attention was confined to the internal distribution systems of produce and processed foods . Animal products and seafoods largely have been untouched.

The approach adopted in Fij i was that research priority should be given to establishing a body of basic data on commercial food distribution that would serve as a source of informat ion for decision-makers. Some implications of the analysis of data are noted and recommendations are made that could solve fundamental problems in the system, but

vii viii

detailed 'solut ions ' have been left largely to decision­ makers .

Such an approach was adopted for three reasons. First, a 'data base' on food distribut ion is a key to economic planning. We have not attempted to provide exhaustive data on any aspect of the system, but have tried to suggest the depth of data that is readily available an d which is a pre­ requisite for sound planning. Second, attention was concen­ trated on the broad structure of the distribut ion system because efforts to 'imp rove ' marketing systems often falter when attention is focused on specific local problems without the benefit of a perspective on the total system. The successful solution of specific problems depends on an appreciation of the fundamental structure and processes of the system and the attitudes and motivation of participants.

Third , as well as on an appreciation of the system's structure , successful resolution of shortcomings in the distribution system is contingent upon decis ion-makers and researchers having common obj ectives . The conclusions of this study touch upon very basic issues such as racial participation in connnerce, ownership of and access to resources, the role of overseas capital and ownership in connnerce, and the degree of government involvement in the economy . The conclus ions highlight possible alternatives in the context of these basic issues . More detailed solutions can be developed once the nature and extent of desired change are known .

The consequences of this approach are apparent in the structure of this report. Following a first chapter that puts the present food supply situation in to perspective , attention focuses in turn on the produce and processed foods distribution systems , an easily-made distinction as there is practically no overlap in items handled by each system, although the business organizat ion of each, as well as racial involvement in them, have much in connnon .

Results of the study were originally published by the Un iversity of the South Pacific under the same title in September 1977. That report was quickly put together as a discussion paper for a post-study conference of researchers and decision-makers , which was held in Suva in Decemb er 1978. When it was apparent that the report would be republished, it was rewritten and polished to remove some of the roughness ix which had been included in order to provoke discussion at the conference . The arguments, conclus ions and the time context (i.e. the 'present' is 1976-77) of this version are essentially the same as the first draft : it is hoped, however, that the path to them is smoother.

Michael Baxter

Acknowledgments

The research for this report has depended on the assistance of a number of in stitutions and many people. The Internat ional Development Research Centre of Ottawa , Canada, made available generous ftmds for the project and its representative showed a keen and continuing in terest in our work. IDRC concern was particularly helpful in over­ coming some administrative problems an d in stressing the combined research and training obj ectives of the proj ect . Particular acknowledgement is made of assistance given by Dr Jacques Amyot, Regional Liaison Officer for the Social Sciences, IDRC , Singapore; Dr Cros Walsh , former Head of Geography at the University of the South Pacific; Mr Derek Medford, Director of the Centre for Applied Studies in Development, Un iversity of the South Pacific; and Professor R. G. Ward and Dr T.G. McGee, Department of Human Geography , Re search School of Pacific Studies, the Australian National University , who are working on a similar study in the New Hebrides and co-ordinated research in the three study cotmtries of the Distribution Systems Proj ect .

Grateful thanks are extended to the numerous government officers and private businessmen in Fiji, not to mention the several thousand patient market vendors and customers , shop owners and customers, farmers, produce transporters an d grocery importers and wh olesalers who were contacted, generally for formal interview. Almost without except ion , all responded patiently and showed considerable interest in the work.

Much of the field research on wh ich this report is based was conducted by Jenny Baines and, to a lesser extent, Raj esh Chandra (both were project research assistants) and by·many students of the University of the South Pacific. In the fourteen months from January 1976, sixty or so students worked a total of about 4000 hours on the proj ect . Those who contributed most include Abdul Shariff, Viliame Lomaloma, Vij ay Naidu, Vij endra Prasad , Holland Seeto, Tevita Ba, Meri Kaloumaira , Lily Bingwor, Ram Autar , Ranjit Singh , Jo Rokorasei, Isimeli Cokanasiga , Premlata Banfal , Eci Kikau,

xi xii

Rishi Ram, William Senegar, Inoke Raikoso, Mosese Uluiciciya and Joketani Delai. Without the help of my research assist­ ants and these students, and many other student helpers who are not named but who were no less vital in their particular roles, the project would have had to be conducted on a con­ siderably more limited scale : grateful thanks to all .

Invaluab le assistance in data analysis came from Janet Aisbett, formerly of the Australian National University, who handled all the computer pro cessing against tight dead­ lines. Jenny Baines learnt a new skill as Project Carto­ grapher. Premila Lakhan typed rough copy for fifteen months; the final version of the report was typed by Beverly Brandenburg in Melbourne. Again , my debt to this assistance is great . Bryan Begley of Honolulu was instrumental in con­ vincing me to spend time reworking the original version of the report and provided key, constructive criticism.

Final thanks are due to Epeli Hau'ofa for being an inspiring co-worker and to Barbara Hau'ofa for her dis­ passionate editing, and to my wife and parents for their involvement in different ways.

M.W.P.B. Contents

Pref ace vii

Acknowledgments xi

Abbreviations xix

Glossary xx

Chapter 1 Food production and consumption 1 in Fij i

Chapter 2 Produce marketing 43

Chapter 3 The processed foods distribution system 133

Ap pendix Interview with a Fij ian permanent vendor at Suva.market 246

Notes 250

References 275

Tables

1.1 Composition of Fiji's food imports , 19 74 8

1. 2 Fiji's food imports: 1874, 1921-76 9 1. 3 Proj ected surplus/deficit of root crops en tering connnercial exchange, 1973 11 1.4 Supply and demand of selected local vegetables 13

1. 5 Selected food imports, 1974 14

1. 6 Post-harvest life of connnon foods 19 1.7 Racial and distribut ion characteristics of urban income , 1973 28

1. 8 Nutritional cost of connnon foods 36

xiii xiv

Tables (continued)

1. 9 Characteristics of connnunit ies surveyed in Central Division food preferen ce study 37 1.10 Consumption of imported foods in selected communities of the Central Division 41

2.1 Revenue of selected markets, 1966- 75 50 2.2 Revenue and expenditure of selected markets , 1975 51

2.3 Vendor attendance at markets 58

2.4 Occupat ion of head of vendor 's household 69

2.5 Race of market vendors 70

2.6 Residence-to-market distance of market vendors 72

2. 7 Means of vendor transport to market 73

2. 8 Production systems of producer-vendors at Ba and Labasa markets 79

2.9 Production system of Fij ian producer-vendors at Ba and Labasa markets 80

2.10 Participation and income of Fij ian producer- vendors at Suva market 82

2.11 Markups from purchase to retail price by non- producer-vendors at Suva market 84

2.12 Trading activity of a Fij ian and an Indian non-producer-vendor at Suva market , 7-11 Septemb er 1976 86

2.13 Net profits of non-producer-vendors at Suva market 87

2.14 Composition of foodstuffs traded by market 89

2.15 Food trade variation by day in selected markets 91

2.16 Share of trade in main foodstuff categories by race and market 92

2 .17 Trade in main product lines by race 93

2.18 Trade composition by race of vendor 94

2.19 Distance of vendors ' residence from market for selected products sold by producer-vendors 98 xv

Tables (continued) 2.20 Correlation between quantity of produce and producers ' residential distance from market 99

2.21 Proportion of food sold by producer 102

2.22 Producer-vendor participation in sales of different foods 103 2.23 Composition of producer and non-producer trade 104

2.24 Vehicles transporting produce to market 107

2.25 Type of journey made by people bringing food to market in non-bus vehicular transport 108

2.26 Distance of journeys made by trucks and 'carriers ' transporting produce to selected markets 110 2.27 NMA trading, 19 71-75 116 2.28 Origin of NMA purchases by division and province 11 8

2.29 Product composit ion of NMA purchases, 19 71-76 119

2.30 NMA purchases by race of seller 123

2.31 NMA dalo buying prices , 19 76 125

3.1 Forms of commercial organization 137 3.2 Trade activity of regional co-operative wholesale associations 142 3.3 Import, wholesale and retail markups on price-controlled items 145 3.4 Location and ownership of retail and whole- sale-retail grocery outlets 161 3.5 Ownership of retail and wholesale-retail food stores, Central Division 162 3.6 Employment of assistance by retail stores 182

3.7 Nature of stock in retail grocery stores 185

3. 8 Number of different food types stocked by retail shops 186 3.9 Display and storage inventory of retail food stores 187 xvi

Tables (continued)

3.10 Mode of payment for supplies and food stock characteristics of retail stores in the Central Division. 189

3.11 Food stocks and store distance from Suva 191

3. 12 Store distance from Suva and food stock characteristics 192

3.13 Food stocks and retail store accessibility 193

3.14 Transport access and food stock characteristics 194

3.15 Storeowner's possession of business transport and food stock characteristics 195

3.16 Form of store ownership and food stock characteristics 197

3. 17 Distance of retail shops from Suva and years of oper at ion 199

3. 18 Number of items displayed and stored, by race of storeowner 200

3.19 Race of owner and foodstock characteristics of individually-owned stores 201

3.20 Race of store ownership and food stocks (all stores) 202

3.21 Food stocks of Indian-owned retail stores 203

3. 22 Occupation of storeowners 205

3. 23 Patronage at shopping centres and expatriate- owned supermarkets in Suva 211

3. 24 Trade characteristics of twelve grocery shops in or near Suva 215

3.25 Trade performance of consumer and consumer- marketing co-operative societies 232

3. 26 Amotmt and source of initial capital of Fijian and Indian retail shops 236

Figures

1 Dominion of Fiji xvi ii

2 Vegetable imports, 1955-74; value and quantity 10 xv ii

Figures ( continued) 3 Price trends of vegetables in Suva market, 1974-76 12

4 Agricultural production by province, 1968 16

5 Urban food consumption and composition of expenditure, by income quartiles 29

6 Urban food consumption by race 31

7 Food preferences of Central Division 34 communities

8 Rural-urban variation in food preferences of Central Division conmumities 38

9 Produce marketing channels 44

10 Marketing channels of dalo 45

11 Vendor numbers and racial composition at produce markets 63

12 Vendors at Suva market 64

13 Supply hinterlands of urban producer markets 66

14 Origin of food at Suva market 97

15 Grocery importing channels 149

16 Supply areas of retail grocery stores in the Central Division 165 ·w I KEY E18 I� :: '�01:- .. : Ot"'•\•Onol 8oundor1��

km 190 9 I I I I 5,0

WESTERN .) /. 1fl / tf' VASAWA r - -!'PGROUP . - - - - - q .. .··' � . '

QL1(1A

1s·s , 1e·s

LA U DIVISION QLAKEBA

\J' n. 0 D I V I S I 0 N

· .· INSET . rJ fvTOTOYA . · �u FULAGA� lJ INSET ROTUMA

E180"W

Figure 1 Dominion of Fiji Note on units of weight

Weights of commodities for sale are quoted in pounds , rather than kilograms , as imperial units are habitually used in the Fij i wholesale and retail areas .

Abbreviations

CASD Centre for Ap plied Studies in Development, University of the South Pacific, Suva.

FAO Food and Agricultural Organisation , UN.

FCA Fij i Co-operative Association Ltd.

FDB Fij i Development Bank.

FMF Flour Mills of Fij i Ltd.

MAFF Min istry of Agriculture, Forests and Fisheries.

NMA National Marketing Authority.

USP University of the South Pacific, Suva.

$ Fij i dollar , equivalent to approximately Australian $0.97 in 1978.

xix Glossary

be le Leaves of Hibiscus manihot

dalo taro (CoZocasia escuZenta)

dalo-ni-tana Xanthosoma taro

dhal lentils duruka Fiji asparagus (Saccharu.m eduZe Linn.)

Fijian ethnic Fijian gala la Fijian farmer living and working outside the village

Indian Indo-Fijian; a citizen of Fiji, but not a Fijian ivi Tahitian chestnut (Inocarpus fa giferu.s ) mataqali Fijian social unit ota fern (DipZazium proZiferum)

roti unleavened bread; staple in Indian diet

rourou dalo leaves

sharps semi-refined wheat flour (used for making roti)

taukei indigenous Fijian person

tikina Sub-provincial administration unit

yaqona plant (Piper methysticum) used to make the beverage also known as yaqona or (outside Fiji) kava

xx Chapter 1

FOOD PRODUCT ION AND CONSUMPTION IN FIJI

Food imports and food dependency

Pacific island countries depend on food imports: between 15 .7 per cent of imports to Guam and Norfolk Island and 38.5 per cent of imports to Niue, by value, are food­ stuffs.! This dependency on foreign sources for food is ironic since until contact with European traders and settlers during the first half of the nineteenth century, individual Pacific islands were largely self-sufficient in foodstuffs. Links with the extra-Pacific world have led to dependency on external food supplies as well as dependency in other fields.

Numerous factors have contributed to this situation in the Pacific. One is the creation of food preferences that cannot be satisfied by local production. Another is the tendency of colonial and later administrations to promote export production of indigenous or introduced agricultural products. Such products became the foundation of island monetary economies through either the indigenous production system being adapted to cash cropping or foreign owned and managed plantations using indigenous or imported labour. Links with external markets were given high priority to facilitate retur ns; internal links with the exception of those of the export sector, were largely ignored. The result was often a developed export sector contrasting with a poorly articulated internal trade system.

In such circumstances, whilst much of the indigenous population remained largely self-sufficient in food, the introduced population, especially the managerial and propri­ etary class, became as externally oriented in food supplies as in commercial interests. A major factor in the spurning of local foods by expatriate populations was cultural prefer­ ence. Also important, however, were difficulties of access to surplus local foods and of the establishment and mainten­ ance of effective links between producer and consumer, and

1 2

the perishable nature of local produce. Whilst the two pro­ duction systems - expatriate commercial plantation and indigenous, largely subsistent - remained isolated from one another, demand was limited for food that was not produced locally. With increased involvement in the monetary economy, however, the indigenous population adopted some imported foods as subsidiary foods and even as staples for similar reasons to those of the expatriate population: accessibility, taste and status.

There is concern in the Pacific and elsewhere about reliance on external food sources. In add ition to obvious financial cost, food dependency has a number of detrimental consequences. Apart from those linking commercial producers to export outlets, internal distribution systems are neglected. Agricultural activities decline, and with this there is a possibility that traditional technologies, including the means to exploit particular eco- and bio-types, will be lost. Additionally, substitution of locally-grown by imported foods may be nutritionally undesirable. Finally, dependence on external food sources can result in a wider loss of economic and political independence. While these reasons for limiting food imports are largely defensive, there are also positive reasons to promote domestic food production.

There is considerable scope for income generation and employment at all stages of the food system: producing, assembling, transporting, processing, wholesaling and retail­ ing. Local food production and concomitant support of an internal marketing system can be a significant stimulus of broader economic development. Development schemes based on agriculture or industry have frequently been ineffective because inadequate attention has been given to delivery channels linking producer and consumer. An efficient dis­ tribution system contributes to the internal integration of an economy, makes effective use of production and insti.ls a 2 degree of self-generation.

Because of the costs of food imports and the benefits deriving from the promotion of local production and trade, food imports are often the target of publicized initiatives. 'Grow more food' campaigns are politically irreproachable and appear to strike directly at the problem - limited local production. What they and other approaches disguise, however, is the fact that self-sufficiency in food supplies, or even substantial decreases in food imports, are not easily achieved. 3

This is not so mu ch because of the amount of foodstuffs actually imported but often because the cultural and histor­ ical context of food dependency is not fully appreciated. As mu ch as overcoming environmental or economic obstacles to local production , cultural preferences and consumption patterns must be able to be satisfied by local production . Fij i is a case in point .

Located in the South Pacific, equidistant between Aus tralia and Hawaii, the Dominion of Fij i is a group of 300 islands that have been independent from the United Kingdom since 1969. The population of almost 600, 000 is concentrated on the main islands of Viti Levu, where the capital, Suva, is located , and Vanua Levu, wh ich together account for 85 per cent of the nation's land surface . The proportion of the indigenous Fij ian population has declined since contact with Europeans , an d especially since the late nineteenth century when Indians were introduced . Today , 50 per cent of the population is Indian (Indo-Fij ian) , while Fij ians account for 44 per cent. The balance is comprised of other Pacific Islanders, part-Europeans/Fij ians, Europeans and Chinese. The problem facing Fij i is particularly interesting, then , and difficult because of the ethnic and cultural diversity of the population . This situation is unequalled in the Pacific.

Like other Pacific countries, and for the reasons noted above , the government of Fij i is concerned about the level of food imports; about one-fifth of imports by value are foodstuffs. Thus from time to time , as well as on a continuous basis , the need to decrease foodstuff imports is stressed. More than for most other nations , an appreci­ ation of these factors is central to designing a successful strategy to reduce this reliance on imported food. Conse­ quently, before moving to an analysis of Fij ian food distribution systems , it is useful to examine briefly the broader context of the system.

Trade systems and food habits

Although hampered by past neglect, there is much to build on in developing internal food marketing in Fij i: the long involvement in transfer and trade by Fij ians3 and the Indian 'cormnercial tradition ' are obvious advantages. Prior to contact with Europeans, extensive exchange systems operated 4

both within the Fij i islands and with Tonga , some 800 km to the east. Exchange in these systems was 'transfer' rather than 'trade ' in that transactions were interpersonal between donor and recipient ; trade involves a multiplicity of links between originator and recipient. Transactions were primarily reciprocal ('balanced movements between symmetric points in a system' ) and redistributive ('movements to a centre and then out again ') (Brookfield and Hart 1971 : 315-16) . Market­ type transactions took place but were less common . One market-type exchange in the 1850s was described as follows :

Heaps of curious looking articles were piled ab out on the beach - fantastic earthenware pots, long rolls of tappas , skins , carved calabashes , spears , oars , etc. ; and around th ese the crowd gathered, dancing and vociferating as if in the performance of some ceremony (Aylmer 1860 :235-6) .

External transfer was primarily in non-food items; food was more important in internal trade , especially in intra­ group exchanges . Because communities were largely self­ sufficient in food, and because of its perishability, food was not as significant in traditional exchange as other items . There was , however , a considerable demand for food by European vessels that visited Fij i to replenish supplies or to gather beche-de-mer an d sandalwood. The Europeans had lit tle difficulty in obtaining local foodstuffs because the items traded for food were highly prized and the mode of transac tion was familiar to Fij ians .

The establishment of plantations did en courage some internal trade of food. In the early years , plantation labour was supported largely with local foods . Some was grown on the plantation and some bought from Fij ians living nearby. To meet the demands of plantations and ships' stores , an extensive trade in less-perishable local foods soon developed . The trade was centred on European firms in Levuka . For example, in 1873 a Levuka firm advertised for sale 500 , 000 yams . Patt.of this supply had been ob tained a couple of months previously in the Waidina valley , where one trader had purchased 180, 000 yams in three days (The Fiji Times, 27 March 1917) .

Much of this early trade with European ships and plan­ tat ions was in kind, but Fij ians also quickly adapted to monetary exchange . When selling produce in Levuka , for instance , Fij ians would 5

..• always make a point of asking for a shilling at first, [but ] they will frequently take sixpence , or even a small piece of tobacco . They have got the names of our coins without having any very distinct idea of their value, especially on the outer islands, so that the new arrival may be asked "a pound" for half a dozen eggs (probably bad) . It must not be supposed that there are

shops and stalls, where a choice of fruits •.. is offered. All that is sold by natives in Levuka is hawked about , each hawker having only one article to dispose of, as a rule. His sto ck-in-trade , if capable of division , is suspended in two coarse coconut-leaf baskets from the ends of a stout

stick, which he carries across his shoulder ..• (Pechey 1870 : 15-16) .

Food production and consumption in pre- and early post­ contact Fij i was in some ways more intricate than today . There were fewer crop varieties but methods of product ion , processing and preparation were often more varied. A number of pro ducts, including dalo (taro) , breadfruit , coconut , ivi (Tahitian chestnut) and vudi (plantain) , were processed for preservat ion and storage. Culinary skills were also intricate, food being cooked by roasting, steaming in a buried oven , or boiling. One nineteenth century observer listed twelve breads , thirty puddings and twelve soups that were common fare (Calvert 1870:119) .

In contrast, the European settlers who became mo re numerous from the mid-nineteenth century had little variety in their food. At Levuka, the early centre of European settlement , and on the plantations, staples for Europeans were flour, biscuits, salt beef and canned fish. Supplies on plantat ions included yam and dalo , but it was felt that yam 'is a poor substitute for cabbage , and that mottled , soap-looking root , the taro , is ••. anything but appetis:ing ' (Cooper 1880, i:75) .

Even though imported salt beef and canned meat and fish were reported to be highly conducive to indigestion and dyspepsia (Brewster 1937: 134) and European fare comparable to ships' rat ions, lit tle attempt was made to integrate local foods into the diet. Produce purchased from Fij ians was mainly fowls, pigs and fruit (some of which - like watermelons, pineapples and oranges - had been in troduced by Europeans). 6

It was some relief to the local expatriate population when Chinese farmers became established around Levuka and produced for sale 'lima beans , cabbage , radish , watercress, spring onions and lettuce ' (Cooper 1880 , i:75) .

Consumption patterns of Fij ians were affected by the new items that entered their exchange transactions, and which included food as well as non-edible goods. One observer in the early 1880s averred that Fij ians

are taking kindly , very kindly to our imported food, tinned me ats, salmon and sardines , bread and biscuits. Tea with them is a favourite refreshment. They have overcome their early obj ection to the use of milk , will soon learn to appreciate eggs , and are even acquiring a decided penchant for Ross' ginger ale of the best Dublin brand (Webb 1884:250) .

Undoubtedly reports of such eating habits were romanti­ cized and even exaggerated. It is difficult to gauge the extent of imported food consumption , but it is likely to have been for many years confined to areas in trade contact with Europeans . Diet elsewhere was based on traditional and a few introduced crops until well into this century when foods like sugar , salt, flour and tea gained wider acceptance. Although some Fij ians might have developed a taste for imported processed foods early , for a long time consumption of these was negligible.

Compared to the Fij ian's minimal consumption of imported food, the establishment of the Indian populat ion from the late nineteenth century entailed considerable demand for non-indigenous staples. Indian foods have remained rice, wh eat flour and dhal (lentils), of which only rice has signi­ ficant local production.

Food imports and local product ion

The reliance on imported staples by Indians and the gradual development of demand for imported foods by Fij ians focus attention on the role of imports in food supply. Food imports in 1975 were valued at $46 million, or about 19 per cent of all imports (Table 1.1). While in both ab solute and relative terms Fiji's food imports are signi­ ficant , three points should be kept in mind . • 7

First, although the Indian populat ion has grown sub­ stantially - it now comprises one-half of the total population - and Fij ian consumption of imported foods has in creased, the contribution of foodstuffs to total imports has altered only slightly over the past century (Tab le 1.2). The demand for imported foods was established early in Fiji's post-contact history and while the composit ion of food imports has altered in response to changes in consumer populat ion and domestic production , the proportion of total imports: that is food. has remained relatively stable. The value of food imports has risen from an average of $10.5 million in 1960-69 to $31.7 million in 1970-76, but the in crease in value has been far in excess of increases in quantity. Data on vegetable imports (Fig. 2 ) . for example suggest that the growth in volume of import s is what would be expected from in creases that have occurred in population and urban settlement.

Second, while there are seasons of relative shortage , local food production is substantial an d the production of many crops is regularly characterized by over-supply . Local root vegetables are a case in point . During 1973, the market for root crops was under-supplied by an average of 300 tonnes a month from January to June , but in the second half of the year the average monthly surplus was 453 tonnes (Tab le 1. 3). Seasonal availab ility of produce is reflected in retail price fluctuations (Fig •. 3) •. Data on vegetable product ion and price trends are in complete, but what are available suggest that indigenous vegetables have more uniform production and price levels throughout the year than introduced 'temperate' vegetables (Tab le 1.4, Fig. 3) •.

With foods grown locally the supply problem is not pro­ duction volume as much as seasonality, and a need for consumers to accept seasonal alternat ives, particularly among non-root vegetables . Half the year is characterized by shortages of locally-grown root vegetables, a situation that can only be overcome by the encouragement of off-season crops, the wider cultivation of the storable yam, or the development of post­ harvest technologies. Consumers have considerable choice in the availability of other crops. The more perennial rourou, ota, bele, eggplant and okra are all good price substitutes for temperate vegetables in the off-season . 8

Table 1.1 a Composition of Fiji's food imports, 1974

% total Imported foods $'000 food imports

Meat and meat preparations 4,691 11.4 Dairy products, eggs 3,546 8. 6 Fish and fish preparations 7,761 18.8 Cereals and cereal products 15, 325 37.1 Fruit and vegetables 5,889 14.3 Sugar and sugar preparations 654 1. 6 Coffee, tea, spices 1,579 3.8 Miscellaneous foods 1,231 3.0

Animal feed 514 1.2 Live animals 113 0.3

Total 41, 302

a Total imports for 1974 were $219.3 million , 18.8 per cent of which was of food items .

Source : Fij i Trade Report , 1974. Table 1.2

Fiji Is food imEorts - 1874, 1921-76

Mean ann- Food as % b ual food Composition of food imports (%) . a tota1···imports Period imports stand . c Other Dairy Fruit/ $ mil. Mean Meat Fish Rice devi . cereals Erods veges 1970-76 31 . 7 18 .7 0.6 10 .2 22 .3 7.9 22.8 10.1 12 .4 1960-69 10 .5 20 .7 1.3 9.3 14 .1 10.1 27.5 9.4 13.9 d 1950-59 5.8 21 .5 2.1 11 .6 7.6 5.7 21 .1 9.7 15 .4 e 1940-49 1.8 24 .2 3.2 9.3 4.5 3.5 27 .8 6.5 10.4 1930-39 0.5 18 .8 2.5 9.0 7.4 7 .4 32.5 4.4 9.9 1921-29 0.6 21 .8 1.5 11 .6 6.4 5.1 37 .6 7.1 8.0 4th qtr 1874 15 .0 20.8 2.9 3.3 20 .4 6.7 2.9 a Excluding live animal s, alcohol , tobacco and animal feeding stuff . (Note that animal feed and live animals are included in data of Table 1.1.) b Figures in upper row of 'composition of food imports' are for 1970-73 (no t 1970-76) . c Figures for 1960-69 and 1970-73 include fish imported to a cannery in Levuka , almo st all of which is sub sequently exported . d 1952-54 figures no t available. e 1941 figures no t available . Source : Fiji, The Trade Report (annual) and Thurston n.d. :4. 10

Potato

Figure 2 Vegetable imports , value and quantity, 1955-74 (Source: Fiji Trade Reports) 11

Table 1. 3

Projected surElus/deficit of root era� entering commercial exchange� 1973 {tonnes Sweet Dalo Cassava Yam Total potato

January - 235 - 164 - 66 - 60 - 525 February - 173 - 164 - 66 - 60 - 463

March - 143 - 164 - 66 + 40 - 333

Ap ril - 224 - 164 + 89 + 40 - 259

May - 160 - 164 + 89 + 90 - 145

June - 180 - 164 + 89 + 90 - 165

July - 157 + 331 + 89 + 90 + 353

August - 107 + 481 + 89 - 60 + 403

September 87 + 481 + 4 - 60 + 338

October + 328 + 481 + 4 - 60 + 753

Novemb er + 108 + 481 + 4 - 60 + 533

December 24 + 479 + 1 - 60 + 396

Total - 1,054 + 1,750 + 260 - 70 + 866

Source : Fij i, Department of Agriculture .

The third point to be considered is that food imports consist largely of products that have never been produced · commercially in Fij i. Foodstuffs that are imports and which in terms of climatic requirements could be produced locally , include rice, pulses , spices , all meat except possibly mutton , and da iry products. These foods accounted for about one­ third of food imports in 1974 (Table 1.5) . However , some categories of imported food , such as 'Fruit and vegetables ', 'Fish and fish preparations' , that appear to be suited to import-substitution in fact include items such as potatoes , onions, garlic , off-season temperate vegetables and cann ed mackerel; these would be difficult to produce locally . Nevertheless, concern over the increased value of food imports should not ob s cure a basic fact : assuming that the total demand for products now imported but also produced 12

1975 1976 cents ...... 60 . \ . , . ....1" ...... ,' .. - ...... ����. . \ i \ · . . . \ . � � � \ · ���·b·��� . \: \ 40 I II 111 11 I I I I I I I I I I pt' ... .\ r-.. c�ese �abbage \} \ , /\. A ...·.. .. -... , ... v 20 ····· . Av �r�: · . 1, � · ····J vy ... v-v·····-\ N'\ f\l. '1 .. v -�/� ..... L t \. M � M M 1975 / 1976 cents '" 80 , /

60

40

20

M M M M

1975 1976

Figure 3 Price trends of vegetables in Suva market, 1974-76 (Source: Weekly price data in The Fiji Times) Table 1.4

Supply and demand of selected local vegetables (Estimates in tonnes for 1973)

January-April May-September October-December Total Surplus/ Production Demand Production Demand Production Demand Production deficit Let tuce 16 80 53 100 4 60 73 -167

Tomatoes 50 220 180 270 80 160 310 -340

Carrots 4 100 9 125 10 75 23 -277

Onions 0 1200 10 1500 35 1000 45 -3655

Source : Fij i, Department of Agriculture . 14

Table 1. 5 Selected food imports , 1974 Percent Percent change Percent Total Imports food in imports demand $' 000 tonnes imports a 1971-1974 imported by value value volume b Rice 5,525 20, 787 13 .4 4 7.9 71.2 Pulses 1,431 3,216 3.5 c.90.0 29.5 180.2 Spices 283 164 0. 7 l.1. a. 31.9 -61. 3 Potatoes 1,432 7,204 3.5 c.95.0 36.5 124.4 Onions 577 3,080 1.4 100.0 33.0 99.4 Garlic 230 539 0.6 100.0 77.9 64.3 c Vegetables 1,796 4.3 112 .5 tomatoes , fresh 45 51 0.1 -15.0 49.3 canned 142 262 0.6 100.0 100.0 199.6 beans 59 110 0.3 39 .2 132 .0 peas , fresh 52 86 0.2 100.0 30.3 194.9 peas, other 27 75 0.2 100.0 47 .l 71.1 other vegetables 205 380 0.9 100.0 27.8 102.6 other frozen vegs 56 86 0. 2 100.0 207.1 345.6 vegetables , other preparations ll9 126 0.3 100.0 6.8 61.1 Meat 4,691 11.4 141.7 beef , frozen , chilledd 477 377 1.2 267.7 319.4 17.7 beef , canned 422 296 1.0 61.3 29.9 mut ton , frozen, chilled 1, 716 1,816 4.2 100.0 173.4 24.5 mut ton, canned 356 188 0.9 100.0 109.2 4.6 pigmeate 302 135 0. 7 28.2 63.5 12.4 goatf ll7 ll6 0.3 24.1 189.l 51 .8 g poultry 823 899 2.0 c.50 .0 123. 9 43.1 Fish canned n.a . 7,091 n.a. 98.0 n.a. 88.1 fresh , frozen n.a. 387 n.a. 18.3 n.a. 239.6 . h Dairy pro d ucts 3,546 8.6 99.8 milk , all types 1,960 2,015 4.7 84.2 3.9 butter 69 7 662 1. 7 73.9 33.2 ghee 657 510 1.2 23.9 -25.2 a Because of home production and , in the case of some meat products, tmregistered slaughters, the proportion of commercial meat demand met through local product ion is difficult to estimate. Figures here are derived from import figures and local production estimates made by the Department of Agriculture in 1973. b Tonnes padi equivalent . c Includes dried and preserved vegetables as well as fresh ; excludes vegetables listed above. d Re gistered production of 3120 tonnes. e Registered production of 286 tonnes plus 20 per cent for informal killings. f Est imated product ion of 355 tonnes. g Estimated production of 1000 tonnes. h Includes eggs , honey and other products as well as milk , butter and ghee. Source: Fiji , Central Planning Office , 19 75, and Fiji Trade Report, 19 73, 19 74. 15 locally were met by local production , at least 12 per cent of all imports would continue to be foodstuffs un til con­ sumption habits chan ge .

In contrast to the foundation for internal food marketing provided by the involvement of Fij ians and Indians in trade , food consumption patterns are a conservative force . Before examining the consequences of this , another factor should be reviewed, that of Fiji's flexible and innovative agricultural system.

The agricultural system

�rops , product ion and change

About 11 per cent of Fiji's land area of 7095 square miles is suited to agriculture. Only 8 per cent of this potentially arable area was cultivated at the time of the 1968 Census of Agriculture (Casley 1969) , an area that had not changed greatly in the preceding twenty years (Blackie 1949) and which probably has not altered significantly since .

Agriculture in Fij i is characterized by considerable variations in the scale and nature of production systems . In the size of holdings , for instance, the 1968 agricultural census revealed the mean area of holdings to be 18. 1 acres, ranging between provinces from 40.3 acres (Cakaudrove) to 2.2 acres (Namosi) . The Central Division provinces of Rewa (5.2 acres), Tailevu (9.0) and Naitasiri (13.2) were among those with small average holdings . Similar variations were apparent in connnercial orientation of producers . Throughout Fij i, 18 per cent of landholders were primarily engaged in subsistence production . Provincial variation in the pro­ port ion of subsistence produce is considerable, rising to 71 per cent in Rewa. The proportion of holdings devoted to 'production mainly for sale ' ranged from 9 per cent (Rewa) to 94 per cent (Rotuma) : the Fij i average was 59 per cent .

There is considerable regional variat ion in crops grown and crop area (Fig. 4 ). Main crops grown for local consump tion are root crops and a variety of other 'native ' and 'introduced' vegetables , rice and tree crops such as banana, plantain , coconut and breadfruit . The chief root crops are dalo (Colocasia escuienta) , cassava (Manihot escuienta) and yaqona (Piper methysticum, the beverage known elsewhere as kava) . Cultivated and wild yams (Discorea spp.) , sweet potato , Xanthosoma taro ('dalo ni tana') and swamp E'22:J Dalo illIIIIIll Cassava Numbers indicate total crop area Other root crops � in hectares § Yaqona l:JJ Fruit (bananas. pineapple. watermelon) �� Rice t::::: :I Pulses. maize. groundnuts CJ Leaf vegetables

0

0 •0

LAU 0 1002

t:?

0 0

Q

0 10 50 km 100

Figure 4 Agricultural production by province, 1968 (Source : Casley 1969) 17 taro (Cyrtosperma charrrissonis , via) are other root crops. There are varied edaphic and climatic optima for these crops. Dalo, for instance, requires mo isture throughout the nine­ mon th growing period. Yams grow best wh ere precipit ation is markedly seasonal; swamp taro grows in brackish water. Yaqona grows well in condition s suited to dalo and the crops are frequently interplanted. Cassava and, to a lesser extent , xan thosoma taro grow satisfactorily on poor soils in dry areas and may be planted at any time of the year.

'Native ' ve getables are leafy vegetables established prior to intensive European contact , including taro leaves (rourou), fern (ota, Diplazium proliferum) and bele , the leaves of Hibiscus rranihot . In the wetter areas these vege­ tables are perennial . With the exception of a few items like eggplant, okra, some beans , kerela an d cucumber, in troduced vegetables are highly seasonal. They are abundan t in the cooler June-to-October period but are generally unavailable at other times. Given close attention , introduced temperate vegetables may be produced in mo st areas , particularly in the cooler regions that include the southern islands of Beqa and Kadavu and Viti Levu highlands. With the exception of bananas and plantain , mo st tree crops have a marked season­ ality but producers can take advantage of varietal differences for a longer product ion season .

The commercial agricultural sector is dominated by the production of sugar and coconut products that together account for 81.5 per cent of Fiji's exports ($95.4 million , 1974) . In comparison with these products , the value of other agri­ cultural exports, of which gin ger is the main one with exports of $0. 55 million in 1974, and of the internal commerc ial food trade , is slight. 5

While there are physical parameters to crop production , crop distribution is at odds with the broad uniformity of the physical environment. An explanat ion of the anomalies is that the culture of producers has a significant influen ce on crops grown . Root crops are mainly grown by Fij ians ; rice produc t ion is almost en tirely in the hands of Indians. Cultural biases in production have a strong spatial expression that reflects the settlement patterns of differen t races.

Rural populations are largely self-sufficient in food­ stuffs that are grown in Fij i. There is minimal intra-rural produce trade at either a local or regional level . Self- 18

sufficiency in basic local foods , however, is frequently disturbed by hurricanes and floods : at these times the main movements of locally-grown food between producing areas occur . Within the Eastern Divis ion , for instance , there are occasional large government-organized tran sferrals of yam and other produce from Lomaiviti to Lau.

The ease of production of indigenous crops and of in troductions like cassava and green vegetables, adequate land resources for home production , surplus labour supply and a shortage of cash to purchase food, all encourage rural self-sufficiency . Another significant factor is the post­ harvest characteristics of local produce (Table 1.6). Apart - from coconut and yam, most local crops have a post-harvest life of less than one week, many lasting only a few days . Some root crops such as cassava may be left in the ground for some time after the optimum harvest time and mo st tree crops have similarly flexible harvest schedules. But overall , flexible post-maturity periods do not compensate for short post-harvest lives . As ide from the drying and milling of rice , post-harvest technologies to preserve or store local food products are not connnon . Moreover, the tropical humid climate shortens the shelf-life of foods that in more mild climates have good storage qualit ies. Flour, sugar, dhal , potatoes, onions and garlic deteriorate quickly unless stored in particularly dry and ventilated conditions.

The high perishability of local produce and irregular transpo rt and conmnm.icat ion linkages contribute to a marked distance decay in market participation . Intermediate storage and handling of produce is discouraged for similar reasons. It is technically possible to increas e the post-harvest life of mo st products; when this does occur , there could be a con siderable ch ange in the source of produce and the structure of the produce trade .

The agricultural system is flexible, particularly where change does not involve modification of its basic structure . The chief crops of pre-contact Fij i were dalo, yam, banana, plantain and breadfruit , the distribut ion of which largely reflected variations in the production environment . Swamp and giant taro and 'arrowroot ', for instance , in some areas were emergency foods and in others were staples. Today , both are rarely consumed. Th e mo st significant post-contact root crop introductions are cassava , sweet potato and xanthosoma taro , although none has usurped the symbolic 19

Table 1.6

Post-harvest life of connnon foods

Post-harvest Food Commen ts life dalo 1 week Unwashed, uncut corm attached to stem

dalo-ni-tana 2 weeks Unwashed, uncut corm yams , all varieties 6-12 months Harvested mature . If sprinkled with ash, post�harvest life is increased

cassava 3 days Not cut or bruised. If buried and kept mo ist , will keep 7 days

sweet potato 1 week breadfruit 1-2 days Harvested mature plantain 3 weeks Picked green coconut 4-6 weeks

taro leaves 2-3 days fern (ota) 2-3 days be le 2-3 days

ginger ( green) 1 mon th

citrus fruit 1-2 weeks Depends on maturity when harvested chillies (large) 1-3 weeks

duruka 3-4 days

tomat oes 2 weeks Picked green Chinese cabbage 2-3 days

English cabbage 2 weeks

beans 4-5 days

eggplant 4-5 days

okra 4-5 days pumpkin 4 weeks

Irish potato 2-4 weeks

onion 2-4 weeks

dhal 4-6 weeks

rice 2-3 months Can be treated to last 6 months flour , sharps 8 weeks

Note : Foods stored in cool and dry place ; the cooler and less humid the weather , the longer post-harvest life. Source: Local observat ion . 20 function or prestige of traditional root crops. Cassava has become widely disseminated since its introduction in the mid­ nineteenth century, and is now probably the most widely planted root crop. It is the basis of home gardens in many urban and peri-urban areas. The widespread introduction of cassava - as of any other introduced crop - occurred largely because its cultivat ion did not require maj or changes in agricultural technology , labour organization or land tenure. Another significant in troduction is rice. In terms of the area llllder cultivation and place in local diet, rice is almost as important today as root crops.

Change has also occurred in the intensity of production . Pre-contact agricultural technology in cluded extensive irrigation schemes , for dalo , and large scale product ion , particularly for yams . In mo st areas both these methods of cultivation fell into disuse soon after initial contact and agricultural product ion became increasingly fragmented.

The trends toward crop divers ification and production disintensifi cation have been reinforced by involvement in pro duction for cash. This has contributed to a reduction in the area tmder subsistence crops and a change in the relat ive importance of crops in response to demands for labour and particular eco-types. The growth of urban populations has en couraged production for the in ternal market , but little of the production for this is by farmers who are 'commercial ' in the sense of producing solely, or even largely, for sale. More common is production by farmers who grow much of their own food and also rely on produce sales for a maj or part of their income .

Production systems and marketing

Basic forms of agricultural production are recognizable in Fij i. Production systems may be differentiated by tech­ nology , form and source of labour , disposal of production and, to a lesser extent , scale of operation and crops grown . At one end of a production system cont inuum are commercial farme rs who produce entirely for sale, employ wage-labour and frequently have large holdings. Connnercial farmers produce mainly for export , but a number grow food for the in ternal market. At the other extreme of the continuum are subsistence producers; they are most common in areas that are marginally integrated in the nat ional economy . Subsistence production is characterized by small , irregular holdings , 21 mixed-cropping, unpaid family labour , simple techn ology an d no production for connnercial exchange .

Between the extremes of connnercial and subsis tence pro­ duction lies a bro ad middle ground that en compasses various types of production involved in , but only partly committed to the cash economy . Two basic forms of subsistence-connnerc ial production are recognizable : the village and the small independent farm systems . Village farme rs are Fij ians using land under traditional (mataqali) tenure and producing primarily for domestic consumption . Land holdings are small and labour is from the family , although connnunal labour can be important for specific tasks . Production is 'tradit ional ' in that it is not me chanized, there is considerable in ter­ cropping an d sequential plantings and male and female work tasks are differentiated. Produce from the village production system that enters commercial exchange is mainly surplus from domestic requirements or products, such as fruits or animal and water products, that are semi-cultivated or collected . Towards the subsis tence end of connnercial-subsistence pro­ duction , crops are rarely planted specifically for sale .

The small independent farming system covers a numb er of production forms , all of which are distinguished from village production by the farmer having individual land tenure rights. Wage-labour is employed on a seasonal short-term basis, phases of production are mechanized and crops are cultivated specifi­ cally for sale. Little of the produce sold by small inde­ pendent farmers is collected from semi-cultivated or wild states . Among Fij ians, people farming individually leased mataqati land - one type of small independent farmer - may not appear significantly different from village farmers. However, not only do the farmers on leased land have individual tenure rights, but often they have a distinctive 'connnercial ' outlook: income from agriculture is required to maintain their style of living.

The three production modes - commercial, connnercial­ subsistence and subsistence - are as much a key to the under­ standing of produce marketing as food product ion . The pro­ duction mode has considerable bearing on the volume and price of produce which enters the market system. Farmers near the subsistence end of the production continuum may deliberately plant crops for sale or plan a surplus in a crop planted primarily for household consumption . However, whether the produce is actually put on the market depends on the producer 's need for cash and whether the effort and cost of harvest, 22

transport an d sale brings a return appropriate to his expect­ ations.

Village producers and many small independen t farmers bring irregularly to market limited amounts of produce which they invariably retail themselves. In contrast, mo re connner­ cial producers grow crops specifically to sell. Poor prices are as discouraging to connnercial producers as to others , but do not lead to violent fluctuations in their market involvemen t. The different production mo des and attitud�s to marketing are reflected in two distinct pricing systems . That of commercial producers functions primarily in response to supply and demand. The other system, the province of the less commercially involved producers, exists within the broad context of supply and demand factors but prices are strongly influenced by money requirements of producers and the returns they expect .

• The dual pricing system has two significant consequences for food marketing. First, if the amount of produce en tering the market system is in fact influenced as much by producer requirements as consumer demand , it will not be affected by guaranteed prices or markets, both of which are commonly advocated means of increasing the amount of produce entering the commercial system. Second, there is considerable corre­ lation between the ethnic origin of the producer , production mo de and crop grown . Fij ian producers are generally closer to the subsistence end of the production continuum and more irregular in market participation than the more 'commercial ' Indian or Chinese farmers. As production of crops is some­ what racially-specific, crops are associated with particular commercial systems and so with different pricing principles. The price of dalo , for instance, is affected by its production being largely in the hands of Fij ian village producers. Considerable research needs to be done into the pricing mechanisms associated with production-marketing modes.

The government and agriculture

A basic obj ective of government development policy is 'maximum possible self-sufficiency ' in agricultural production (Fij i, Central Planning Office 1975:65) . The parameters of the 'possible ' are not defined, but the Department of Agri­ culture directs its attention to the technical aspects of production of a wide range of food and other products. Extension staff of the Department act as market-intelligen ce 23

sources for farmers, and at times organize the sale of produce , but their chief role is to improve production techniques and raise productivity. For farmers, there are numerous formal sources of loans for production costs - the Department itself and the YMCA lend small amounts, mainly to Fij ians, and larger amounts are available from the Fij i Development Bank.

In contrast to this assistance for production , marketing is largely left to the producer. The Department of Agri­ culture sees marketing assistance primarily in terms of the National Marketing Authority, 'the key to stabilisation of prices and supply of agricultural produce for consumers and of remunerative prices to farmers ' (Fij i, Cen tral Planning Office , 1975:65) . Three per cent ($669, 000) of the Department 's 1976-80 capital expenditure budget has been allocated to 'development of marketing', that is the National Marketing Authority. Another $1 million under 'Rural Services ' will go to the establishment of markets in rural areas .

The consumers

In an examination of the problems of the food supply and dependence system of Fij i, attention invariably centres on production . Although they 'have a sign ificant in fluence on the level and form of food consumption , consumers receive little attention . Failure to take in to account the aspiration s and preferences of consumers reduces the chances of success of any program relating to food production or consumption. The con sumers of Fij i canbe seen from three perspectives : (a) the nature of the economy; (b) spatial aspects of the economy and settlement, particularly rural/urban locat ion and access to supply centres ; (c) the racial and cultural features of the population .

The nature . of the economy

Links between the indigenous subsistence economy and in troduced monetary activities were for many years tenuous and spatially circums cribed . Today , however, while features of the classic colonial subsistence-connnercial dual economy are still apparent, connection with the monetary system is increasingly pervas ive . After limited growth until the early 1950s, the economy has been relatively buoyant.6 Over the decade to 1975 real economic growth averaged 3 to 4 per cent 24 annually . The Gross Domestic Product was $476 million ($837 per capita) in 1975 and is predicted to in crease at an annual rate of 7 per cent between 1975 and 1980 . Primary industry contributed to 27 per cent of the GDP , secondary industry 19 .3 per cent , services 43.2 per cent , and in direct taxation the remainder.

The past three or four years , however , have been a period of combined stagnat ion and in flat ion that has reached even the mo st isolated areas . The annual rate of increase of the Con sumer Price Index rose from 4.1 per cent in 1970 to 14 .4 per cent in 1974. The rate of inflation in creased markedly in the early 1970s, contributing to the imposit ion of price control an d wage restraint me asures. However , apart from controls on prices of some twelve 'essential connnodities ', rents and in ter-island shipping rates, these restraints have now been lifted. In flat ion in 1977 is about 12 per cent per annum; real growth continues, although not at former levels. Th e state of the mon etary sector is suggested by employment data. During the 1960s and early 1970s , wage and salary employmen t increased 6 per cent annually. Since 1973 there has been a fall in the rate of employment expansion which is expected to average only 3 per cent from 1975 to 1980.

In spite of the recent economic downturn , Fij i is a relatively wealthy developing nation . In addition to in creases in real wage and salary growth, indications of general pros­ perity are numerous . One particularly relevant to food marketing and nsuco mption is the trend in vehicle ownership . From 1960 to 1974, the number of licensed vehicles increased threefold to 23,700. The number of private cars rose at an even faster rate to 12 , 700 in 1974 from 3200 in 1960. Weight­ ings of foodstuffs in consumer price index calculations derived from urban income and expenditure surveys are another indication of prosperity. From 547 (out of 1000) in 1959, the foodstuff weighting in the CP I decreased to 490 (1965) , 434 (1968) and most recently 400 (1973) . 7 Greatest increase in weightings between 1968 and 1973 occurred in 'Housing and household operat ion ' (from 332 to 360 in 1973) and 'Transport ' (66 to 84) . Surveys in 1968 and 1972 both con­ cluded that urban consumpt ion patterns have changed and become more varied with rising in come .

In creases in real wage and salary rates are enj oyed by less than one-half of the total labour force. Fiji's labour force in 1973 was estimated at 167,000 of whom 154 , 000 were 25 economically active , although only 61,000 of these were classified as wage or salary earners. The remainder was accollllted for by self-employment in both the agricul tural and non-agricultural sectors . Data on the self-employed are not available , but overall they are less well-off than wage or salary earners ; for most, annual income is probably under one-quarter the nat ional average. During the early 1970s, there was a decline in the real standard of living of farmers and the rural population in general (Fij i, Central Planning Office , 1975 :65) . Moreover , the proj ected one per cent annual growth (1976-80) in the 'sub sistence ' agriculture sector, that is self-employed commercial-subsistence farmers , will barely cover proj ected population increase.

Spatial aspects of the economy and settlement

Rural/urban dichotomies are strong in Fij i and have a significant expression in food consumption through their association with particular forms of livelihood an d degree of invo lvemen t in the monetary economy . Conunercial activity is concentrated in the Suva and utokLa a urban areas which accollllt for 25 per cent of Fiji's populat ion . Outside these areas and other towns and the sugar producing areas , there is limited involvemen t in the cash economy, particularly on the outer islands where 13 per cent of the populat ion lives .

The spatial isolation of many areas of Viti Levu an d Vanua Levu is being overcome by road development , but distance decay of economic activity from towns and roads is marked. Efficient linkage of the islands with the conunercial system has not been achieved. Marked variation in levels of livelihood is equally apparent at very local levels . Even areas that are 'urban' by definition have a sharp juxta­ position of rural (including agriculture for home consumption) and urban (monetary employment) activities . 'Rural ' and 'urban ' populations are in no way mutually exclusive ; the labels mask a broad range of economic activities and standards of living within each .

Population size has changed greatly over the past century. From 127,000 at the first census in 1881 , the popu­ lation has increased to 588,000 in 1976, 36 per cent of whom live in urban areas . 8 Population is increasing at ab out 2 per cent annually ; 39 per cent is less than fifteen years of age and 50 per cent less than twenty. Population composition has also altered drastically . In 1881, Fij ians represented 90 per cent of the populat ion; they now represent 44 .2 per 26 cent. Other racial components are Indians (49.8), part­ Europeans (1. 8) , Ro tumans !311 d Pacific Islanders· other than Fij ians (2. 4) , Europeans (0.8) and Chinese (0.8) . 9

Racial and cultural food consumption habits

The racial composition of the populat ion of Fij i affects food consumption in two ways . First , variations in the amount and type of food consumed that could be expected as a reflection of differen ces in income are obs cured by the fact that there are marked differences in income by race. Second , racial , and cultural , groups have traditional food consumption preferences and habits. Each of these factors is examined below.

Useful information is available on income , race and food consumption , at least for the urban populat ion . The 1973 household income and expenditure survey, for instance , reveals a mean average Fij ian fortnightly household income of $100 .5 compared to $115.3 for Indians and $186.3 for the mainly Ch inese and European 'others ' (Tab le 1.7 ). Fij ians comprise 54.5 per cent of the first income quartile , compared to 35 .2 per cent of the fourth . The respective rates for Indians are 44 .1 and 52 .7 per cent and for 'others ' 1.3 and 12 .1 per cent . From another perspective, 29.9 per cent of Fij ian households are in the first quartile and 20.6 per cent of the fourth ; for Indians the rates are 21.2 and 27.1 per cent; and for 'others ', 5.1 and 51.3 per cent. These rates are for urban populat ions where 97 per cent of Fij ian respondents received salary or wages compared to 70 per cent of Indians (Fij i, Bureau of Statistics, 1974:21) : it is likely that in rural areas income differentials , by race , are significantly greater .

Chan ges in income are reflected in both the proportion of income devoted to food and the type of food consumed. The 1972 urban household income and expenditure survey revealed that the proportion of income spent on food was 50.7 per cent in the first quartile , decreasing to 36.7 per cent in the fourth (Fig. 5) . Food groups in which there are the greatest decreases are 'baking products, pulses and cereals ' (12.8 to 5.1 per cent) and 'fruits and vegetables ' (8. 3 to 3.8 per cent). Expenditure increases in 'beverages ' (2.8 to 5.1 per cent) and remains relatively even in other categories. 27

Little in formation is available on food demand elas­ ticit ies although some inference is possible from the house­ hold expenditure and income surveys. An index of elasticity calculated from the range of demand and the average demandlO reveals greatest variation in demand occurs with meals (90.7 per cent) , beverages (84 per cent) and tobacco (71.1 per cent). Amon g 'essential ' foods , bakery products, pulses an d cereals (62 per cent) and fruit and vegetables (58.3 per cent) have greatest elasticity. Least elasticity occurs with milk , cheese and but ter (24.1 per cent) , oils and fats (27 .3 per cent ) and eggs (27.8 per cent) . Fresh fish (36.3 per cent), confectionery (37.1 per cent), sugar (38. 5 per cent ) and meat and canned fish (41. 5 per cent) comprise a middle level of elasticity. Data on in dividual foods are not available. It is likely, however, that at least amon g the wage-earning population there is a higher demand elasticity for traditional root crops than for introduced dry goods like wheat flour products, rice an d sugar .

The second way in which the racial composition of the population has a significant effect on food consumption is that racial - an d ethnic and cultural - groups have trad­ itional food consumption habits and preferences. Each race has a basic diet , the core of wh ich changes but slowly. The traditional Fij ian etdi , for instance , comprises starchy root vegetable staples , such as taro , yam, cassava and sweet potato, a small variety of indigenous vegetables like rourou, bele and ota, an d seafood an d other animal protein . Connnonly Fij ian cuisine has a coconut cream (lolo) base and little additional seasoning. Diet of other Pacific Islanders in Fij i is essentially that of the Fij ians.

Makin g allowance for vegetable and me at-eating diets, the traditional Indian diet comprises large amounts of cereal in the form of rice an d wheat flour products, pulses, a range of green vegetables, of which eggplant , okra, tomato and beans are the mo st common , and small amotmts of animal protein. Indian cooking is characterized by masala-based curries and con siderable us e of on ion , garlic , vegetable oil and ghee .

The other main minority groups , the Chinese and the Europeans , also have distinct diets. Chinese diet is based on rice, leafy vegetables and me at . European consumption of local produce , aside from introduced temperate vegetables N 00 Table 1.7

Racial and distribution characteristics of urban income , 1973

Q Q Q Q Number l 2 3 4 h'holds

Fij ian (% households 54 .5 46 .3 40.1 35.2 281 in quartile) Indian (% households 44 .1 51.3 51. 9 52. 7 321 in quartile) Other (% households 1. 3 2.5 8.0 12 .1 39 in quartile) Mean fortnightly household in come ($) 45.6 70.9 108 .1 213 .0

Range of fortnightly household income ($) 0-5 7 58-88 89-137 >138

Numb er households 154 160 162 165

Note: Informat ion derived from a survey of 641 randomly sampled households in six urban areas over six weeks in October-Novemb er 1973. Source : Fiji, Bureau of Statistics , .1974 :10 , 18. 2 9

(/) c (lJ Q CJ) ::J(/) C'CJ ro 0 Qi Qi (lJ > Cl. c (lJ CJ) 0 .D cii o C'CJ o.=Q) ..c£ £ -o (/) -00 3o C'CJc .� () 0 0 0 ,c u.. u I ..c � �

1st quartile S:::::::::::::totalJ ••expenditurellll $ 7 136

2nd quartile

ELHYY\\TYTUtotal expenditure11111 $10 674

3rd quartile El\jjjjjj:\H\\jj\:H\\jjjjjjjjj/jjjjj]:total ·••: expenditure11111 $12 766

4th quartile

total111111111 expenditure11 $ 20 337

1000 5000 dollars

(/) cli Q) u (/) -0 2iC'CJ ::J -mQ) Q) -0 c ..c Q) 0 Qi c TOTAL EXPENDITURE ON FOOD (/) CJ) c. C'CJ () () .;:: Q) > � > Q)cli ..c � Qi (/J CJ) (/) � :; -"'" U..UJ�g; (/)::J C'.l C'CJ :;Cl. Q) U: � I I I I 11 I JI lst quartile. $ 3 611 \\I I I

2nd quartile. $ 5 031 I I I '1 1

3rd quartile. $ 5 224 II :I I I 1.1 11 11 I .. : /! f I II I 1· I If 11 1 I 4th quartile $ 7 464 10 per cent 100

Figure 5 Urban food consumption and composition of expenditure , by income quartiles (Source : Fij i Household Income and Expenditure Survey, 1973, Bureau of Statistics, Suva , 1974 :22) 30

and some meats, is limited. Of all groups, the European is mo st dependent on non-local foods .

Tradit ional diets are also tempered by numerous influ­ en ces . Religious beliefs , for instance , have significant dietary ramifications . Amon g the Indians , Hindus will not eat cattle products, Muslims avo id pork and vegetarians shllll all meat . Although there are few traditional food taboos among Fij ians, Christian sects have introduced a number to their Fij ian adherents.

Traditional diets are no longer as exclusive as before . A decline in racial exclusiveness and religious zeal and an increase in urbanization have contributed to more shared diet patterns. Bread and other bakery products, rice , canned fish and all seafoods , beef, cassava , dale and rourou are consumed by urban Indian and Fij ian alike . In spite of these changes, foods of the tradit ional diets remain bas ic even in urban areas .

Racial differences in diet are readily recognisable in findings of the 1968 urban household and expenditure survey, one of the few sources of in format ion on food consumption by race (Fig. 6) . In comparison to other races, Fij ians made a great er proportion of food purchases in fish and root crops and a smaller share in bakery products, cereals and pulses , fruit , oils and fat . In dians spent more on bakery products, cereals , pulses , oils and fat and fruit than Fij ians , Chinese or Europeans. European (and part-European) households spent least on bakery products, cereals and pulses , oil and fats, but considerably more than others on meat and beverages . Chinese households had greater expenditure on meat , sugar and confectione.ry than other races.

In all urban households , breakfast generally includes some cereal products and tea. Fij ians commonly eat bread or biscuits, or sometimes starchy vegetables from the previous day , with tea. Indians are likely to have a vegetable curry and roti with their tea. In rural areas Indians eat a similar breakfast, but Fij ians are likely to have a later , heavy breakfast, eating rice or left-over starchy vegetables as well as tea. For both Indians and Fij ians the evening meal is the chief meal of the day, differing from lllllch mainly by the greater variety of foods consumed.

The role of purchased food in diet has definite racial and spatial characteristics. No connnunity in Fij i is 31

EUROPEAN CHINESE INDIAN FIJIAN 0

Bakery products. 10 pulses. cereals ..... ·1 . Meat

Fish

Milk. cheese etc. Oils. fats

Fruit vegetables

Sugar. confectionery Tea. coffee. cocoa

Other food

Beverages. alcohol

Root crops

Tobacco

100 per cent 36.8 44.4 38.6 52.4 Per cent household expenditure on food

$ 99 $ 61 $ 74 $ 44 Average fortnightly household expenditure (all items)

Figure 6 Urban food consumption by race (Source : A preliminary report on the 1968 urban household in come and expenditure survey in Fiji, Bureau of Statistics , Suva , 1968 ) 32 completely self-sufficien t in food: and few are far from a store where some foods may be purchased. Every Fij ian village household has access to land for food gardening from which basic vegetable requiremen ts are met : there is very little purchase or cash exchange of these foods in rural villages.

Rural Fij ians purchase essential foods such as flour, sugar , canned mackerel and corned beef, salt, tea, rice and canned evaporated milk at privately-owned or co-operat ive society stores in their villages . The stores do not have refrigerat ion . Perishable foods in cluding fish and other seafoods and meat may be purchased occasionally at a shop or market in a town or perhaps at a nearby Chinese or In dian store. In mo st rural areas , produce is rarely purchased and expenditure on processed food is limited. In villages, each adult consumer probably spends $4-$5 on food each mon th , although the actual amotm.t is largely dependent on the avail­ ab ility of cash .

Seasonality of food production contributes to consider­ ab le variat ion in consumption patterns in those rural Fij ian connntm.ities largely self-suffi cient in food. Rural house­ holds do not purchase vegetables to compensate for seasonal shortages of starchy vegetable staples; rather, rice and dry biscuits are consumed more frequently. In towns, the type of food consumed varies more than in rural areas , but is also , to a large extent, seasonal . Particularly in urban areas , but also in some rural zones, the dissemination of cassava has done much to stabilize the seasonal availability of starchy vegetables.

Indians are more dependent on purchased food than Fij ians because local production of their staples, rice and pulses, is limited and in terms of energy expended more difficult than the production of root vegetables. Rural households, however, frequently have garden plots wh ich supply mo st of their green vegetables. Rural Indians produce a greater share of their animal protein requirements than Fij ians. In urban an d peri-urban areas , Fij ians and Indians sometimes have considerable home gardens but their actual contribution to food con sumption is unknown . Europeans and Chinese in urban areas purchase practically all their food.

The survival in Fij i of traditional consumption habits in the face of social and economic change suggests the conserv­ atism of diet. Non-economic criteria like taste and ascribed 33

status of foods contribute to dietary conservatism, but also con sumpt ion habits are often the product of rat ional evalu­ ation and selection . When traditional foods are omitted from that diet , it may be because they do not fit the pro­ gressive self-image of consumers. Equally important, how­ ever, are economic conside rations: out side producing areas traditional foods are often in short supply and expensive . Informat ion on the nut ritional cost of food (Tab le 1.8) suggests that consumers in Fij i purchase food wisely. Rice , dhal , flour, sharps, bread, cassava and canned mackerel form an economical basis of diet. Nutrition from local alternat ive starches and protein is markedly more expensive .

In addition to cost con siderations is the fact that a food's utility in a cash economy differs greatly from that in subsistence societies . Many indigenous foods of the Pacific have short storage lives and are highly perishable. Seasonality and the available size units cause sharp fluctu­ ations in availability and consumpt ion . Moreover, preparation of indigenous foods by traditional means can be a lengthy process. Problems of availability, storage an d preparation may be acceptable where time is not at a premium and food supplies are largely consumer-grown . But if foods must be purchased and prepared in the time schedule peculiar to wage labour , different considerat ions are important . In urban areas, for instance, as well as being cheap , readily avail­ able, easy to prepare an d involvin g minimal cooking time , food needs to be available in units large enough to supply a family, but small enough to limit the number of outsiders whose meal-time visits are more free-loading than sociable .

Food consumption - a case study

Food consumption patterns are the product of a variety of factors. To put the main influences in Fij i in to pers­ pective , that is, the consumer 's race , economic status and residen tial location , it is useful to consider a case study of food consumption conducted in the Central Division in six rural connnunit ies; two Fij ian peri-urban villages, another peri-urban settlemen t an d three urban connnunities in the greater Suva-Nausori area (Table 1.9, Fig.7). 11

The diet of rural Fij ians in the Central Division is based on home-grown starchy vegetables. In five of the six en tirely Fij ian communities surveyed, root crops , plantains , bananas and breadfruit are the most connnonly eaten foods ; 34

CJcereals [IlililIJ]Root vegetables h/J Meat. fish � Vegetables. fruit CJother -- Main road

Food preferences of Central Division comnum ities ( Source : Fieldwork) 35 in one , almost one-half of the recorded 'person-servings ' - one person eating a particular food, irrespective of amoun t, at one meal - are of starchy vegetables. For all Fij ian households , 29.4 per cent of person-servings are of starchy vegetables ; cereals and cereal products are almost as commonly consumed, meat only slightly less so. Vegetables other than the staple starches account for much of the balance . Within these broad categories particular foods are favoured. Bakery products and rice are frequently consumed. Cassava is a maj or food, being consumed on over 60 per cent of possible occasions in two communit ies an d over 50 per cen t in two others . Canned fish, almost solely mackerel, accounts for mo st imported 'meat '; local meat is primarily fish and sea­ food although beef is also important. Taro leaf , bele , ota an d varieties of 'cabbage ' account for almost all other vege­ table consumption. Coconut cream is frequen tly used , being served on 40 per cent or more of possible occasions in four communities .

Food consumption frequencies, that is 'food preferences', of Indian households differ greatly from those of Fij ians (Fig. 8) . Data were collected for 31 Indian households, although on ly one samp le community, Raralevu , was composed en tirely of Indians. In the Indian households , cereals and cereal products account for 50.6 per cent of total person­ servings ; non-starchy vegetables follow with 17 .4 per cent , meat accounts for 15 per cent and starchy vegetables for slightly less. Eggs and sweets are more frequently consumed by Indians than by Fij ians but remain negligible.

The frequency with which cereals and cereal products are consumed by Indians and root crops by Fij ians are about equal . Each group also con sumes vegetables and fruit with about the same frequency. Starchy vegetables are less fre­ quently eaten in Fij ian urban and peri-urban households than in Fij ian rural households , but urban dwellers apparently compensate by eating cereals and cereal products twice as often as rural Fij ians . Fij ians consume meat an d fish more frequently than Indians (24.3 to 15 .0 per cent of person­ servings), and the rate of meat and fish con sumption does not differ greatly among Fij ian households. In contrast, meat and fish consumption is markedly more common in urban and peri-urban than in rural Indian households.

Although Indians and Fij ians may appear to have similar consumption patterns at the level of broad food categories , consumption is in fact markedly different. Among 'cereals ', 36

Tab le 1. 8

Nutritional cost of common foods (cents)

Cost (cents) per : Food 100 lOg lg lOOg kg calories protein fat CHO dalo 40 3.5 20.0 15.4 dalo-ni-tana 28 2.4 13.8 10.6 cassava 17 1.1 24.4 8.6 4.6 yam 52 5.0 26.0 20.0 Irish potato 38 5.1 19 .0 22 .4 plantain 26 2.4 31 .0 15.5 10.0 breadfruit 24 2.1 16.1 6.1 9.3 jackfruit 38 5.3 25.6 12 .8 24.0 carrot 92 27.9 92 .1 131.6 pumpkin 40 11.1 4.0 50. 0 fern 38 11 .5 10.5 2.9 100.0 taro leaf 49 17.5 24. 5 16.3 12 2.5 lettuce 267 140.5 190. 7 890.0 eggplant 26 11.8 26.0 65.0 Engl ish cabbage 125 44.6 62 .5 41. 7 312 .5 Ch inese cabbage 33 18 .6 26.0 17.3 130.0 amaranth 80 28.6 40.0 26.7 200.0 okra 67 20.2 33.4 111.2 long bean 109 32 .1 54.5 181. 7 tomato 88 44 .0 88.0 22 .0 pawpaw 33 8.5 54.2 3.6 banana 31 2.7 31.0 10.3 11.5

rice 40 1. 1 5.7 7.9 5.0 dhal 55 1. 7 2.8 2.8 9.5 flour 35 1.0 3.2 2.2 4.9 sharps 35 1.0 3.5 2.5 4.7 dry biscuits 97 2.7 9.7 4.0 13.4 bread 48 1.9 6.5 3. 7 9.1

chicken 255 18.3 13.4 3.6 beef (mod. fat) 152 5.8 9.5 0.7 mut ton (mod. fat) 139 5.6 9.3 0. 7 fish (lean) 144 15 .6 6.7 22.7 canned mackerel 99 5.0 4.8 1. 3 canned corned beef 249 11.0 10 .8 1. 7

Note : Calcul at ed for net costs of food (i.e. pre-cooked edible port ion) . Fifteen per cen t allowed for waste with chicken , beef and mutton, 10 per cent with fi sh ; net costs of othe r foods from experiments.

Source : Costs from observat ion in Suva, February 1977; nut ritional data from South Pacific Health Service , Food Composition Tab les for Use in the South Pacific , n.d. Table 1. 9

Characteristics of corrnnun ities surv�ed in Cen tral _l)_i_vision �o-� J� renc�s study

Km Sample Mean weekly Sources of H 'holds Connnun ity to Access Settlement type ethn ic h 'hold net Income sources purchased with Suv com nco e o gardens ______� ______E_._ i _ _� __J_$_L______l_o_ _g__ �aralevu 24 Frequent buses to Dispersed rural 1 10 66 Wage employment in Suva and Local stores Nausori and Suva settlemea t N&unori; rice farming , mainly (3) , Nausori for home consumption

Wainihuku 15 Frequent buses to Re cent middle-income 7 53 Semi-pro fessional and semi- Local super­ Nausori and Suva ; sub-divis ion F 3 3killed employment in Suva market and shop oriented to Suva and Nausori

Cal i a 43 Half-mile from Navua; Rural-urban fringe 6 22 Un skilled labour ; two semi­ Lo cal stores regular buses to Suva of Navua F skilled workers , two farmers (2 ), Navua

Davuilevu 16 On King's Road one Urban settlement 1 6 48 Semi-pro fessional and semi- Local stores mile from Nausori ; within Nausori town F 4 skilled; in three households, (2 ), Nausori frequen t buses boundaries husband and wife employed

Vuni vi vi Hill , 19 Walking distance to Upper-class resi­ F 7 118 Higher level professional; Nausori, Suva 8 Nausori Nausori; mo st den tial area to 1 2 husband and wife employed in families have car Nausori Mix 1 four households

Suvavou Frequen t buse s to Peri- urban village F 8 so Semi-professional and semi­ Village store , Suva and Lami on Queen 's Road Mix 1 skilled ; three households , local stores (2) , ( 3 km) husband and wife emp loyed Suva

Wailoku 13 Terminal of bus Peri-urban village ; F 10 41 Un skilled and semi-skilled; Village (3) , and All route; frequent population descended in two households , husband local stores (2) buses to Suva from Sol omon Islands and wife employed

Nadoi 25 15 min utes boat to Village in the Rewa F 10 21 Two daily Suva wo rkers ; one Vi11 age stores All Lakia landing , thence De lta boat operat or; irregular (2) bus to Suva income from copra and market

Nailega 46 Buses to Suva or Village on King 's F 10 14 Two weekly Suva workers ; Village store All Korovou (3 km) Road, Ve rata road-side and market Korovou produce sales

Nasautoka 82 Buses to Suva or Village on King's F 10 20 Sale of produce at urban Vil lage store All Korovou (32 km) Road , Wainibuka markets; one local teacher

Tuvatuvavat u 93 5 km walk to road , Village off King 's F 10 20 Sale of produce at urban Village stores All thence bus to Suva Ro ad , Wainibuka markets (3) or Korovou (42 km)

Nuku 6 7 1 hour by boat to Village on midreaches F 8 23 Two boat operators, one Village stores All Navua , thence bus to of Navua River teacher; market sales at (4), Navua Suva Navua ------38

Cereals and nI cereal products

Starchy vegetables

Meat (including poultry and fish)

Fruit and vegetables

Other

PERI-URBAN & URBAN RURAL ALL AREAS

Figure 8 Rural-urban variation in food preferences of Central Division commun ities (Source : Fieldwork) 39 for in stance , both consume rice , but ro ti (unleavened bread) and dhal are exclusively consumed by Indians. In Raralevu roti is eaten on almost 60 per cent of possible occasion s and dhal on 24 per cent . In dians as well as Fij ians consume taro and cassava , but jackfruit an d Irish potatoes are con­ sumed almo st exclusive ly by In dians . Mut ton , ch icken , duck an d goat are preferred meats for In dians, wh ile beef and pork are popular amon g Fij ians . Taro leaf and cabbage , the two chief leafy vegetables of Fij ians , are also eaten by Indians, but eggplant, okra , tomato and a number of varieties of bean are their basic vegetables. In addit ion , Indians make chutneys and pickles from a wide range of ve getables and fruit . Frequen cy of consumption of fruit , eggs and sweets by both Indians and Fij ians varies greatly between connmm ities .

Th e rural Fij ian communit ies have relatively un iform patterns of food preferences but there are considerable differences between rural and urban jFi ian households. In rural households, 39.8 per cent of all person-servings are of starchy vegetables compared to 27.2 per cent in urban and peri-urban Fij ian households. Vegetables and fruit are con­ sumed mo re frequently (22.4 compared to 16.5 per cent of person-servings ) and cereals and cereal products less commonly (14.5 to 28. 9 per cent in rural than in urban or peri-urban households. Consumption frequencies of foods of Fij ian households in two peri-urban villages, Suvavou and Wailoku, are practically identical with consumption in other Fij ian urban and peri-urban hous eholds.

Differen ces between rural and urban diet are much less marked amon g In dians than Fij ians . For rural , peri-urban and urban In dian households, one-half of person-servings are of cereals an d cereal products - in cluding lentils. Starchy vegetables (13.2 against 9.7 per cent of person­ servings) and meat (15.8 versus 10 .7 per cent), but fewer vegetables and fruit (17.7 to 24. 7 per cent) are more frequently consumed in urban an d peri-urban than rural In dian households.

Data from the community studies suggest the extent of penetrat ion by imported foodstuffs . 'Imported foods ' we re taken as all cereals and cereal products excluding rice , Irish potatoes , canned meat and fish, frozen and canned vegetables and fruit . Consumpt ion of imported foods ranges from 10.4 to 48.6 per cent of total person-servings (Tab le 1.10) . If imported rice was in cluded, the share of imported 40 food, undoubtedly, would be higher . Im ported food consumption is highest in the Indian and largely Indian connnunities. Consumption of imported food in the Fij ian pe ri-urban villages is higher than in rural villages , but is still not equal to the level in the Indian connnunit ies . The effect of income and source of income on the rate of local food consumption is suggested by the fact that the three Fij ian villages with greatest involvement in wage labour (Nadoi, Suvavou and Wailoku) have the highest frequencies of imported food con­ sumption among the Fij ian co nnnunit ies .

The type of imported food consumed varies between commu- nities according to racial composition and rural-urban status (Tab le 1.10) . In the five rural Fij ian villages , imported foods are almost equally divided between canned mackerel and bakery product s - largely dry biscuits. In the peri-urban villages bakery products are more frequently consumed and cann ed fish considerably less frequently than in rural villages. Th ere is also greater diversity in imported foods in peri-urban than in rural connnunit ies. In Raralevu, rot i, dh al and potato account for 76 per cent of imported fo od, by person-servings. These foods are also important in the peri-urban connnunit i es, but bakery products, particularly bread, are more connnonly consumed there.

There was no evidence from the food preferences surveys of a causal relat ionship between in come level and consumption of imported foods , even within the same racial gro up . For in stance , the largely Fij ian upper-class community of Vunivivi Hill, Nausori , has the lowest frequency of consumption of imported foods of all urban and peri-urban communities .

Two conclusions may be drawn from the connnunity surveys of food preferences. First, traditional racial dietary patterns remain strong, particularly in rural areas but also in towns. Traditional diets have been maintained in the face of pervasive forces towards uniformity and suggest the conservative force of 'tradition ' in diet . At tempts to change eating habits without making allowance for the conser­ vatism of traditional diet have little chance of success , particularly as some reasons for tradit ional diet that at first sight seem based on tradition rather than logic actually have sound economic justification . A second conclusion is that people in many conmnm ities purchase min imal amounts of food and that there are significant differences in rural­ urban consumption habits, part icularly in terms of consumption of purchased foods . Table 1.10 a Consumption of imported foods in selected communities of the Central Division

Per cent share of imported food

rl 0 i::: C1l .w ..c .w Q) Q) C1l UJ C1l N CJ) % person- .w ri l-1 • Q) 0 Q) .w CJ) servings CJ) Q) 0 4-l l-1 rl ·ri .w 0.. 13 b .w (.) 4-l .0 ;:l (.) "Cl "Cl - Ctl l-1 imported >. (.) � l-1 ;:l l-1 ;:l ..c Q) Q) ..c 0 .w 4-l Q) "Cl •ri rl Q) "Cl CJ) � � u:i .w Q) ri � 0 .w Ctl ..c 0 • Cl) .w 00 "Cl Ctl l-1 0 ..c .w l-1 l-1 .0 0.. l-1 "Cl 0 0.. � (.)� H (.) Jj � � � Raralevu 48 .6 7.7 43. 7 18.3 0.5 14 .0 7.4 0.6 5.9 1.9

Wainibuku 43.5 15 .8 39.6 12 .9 3.3 14 .2 11 . 3 0.3 2.6

Cali a 40.7 13.4 37.6 16.0 16 .4 9.9 2.9 3.8

Davuilevu 46 .2 19 .0 35 .3 9.5 5.5 9.1 14 .1 1. 7 4.8 1.0

Vunivivi Hill 29.3 49 .5 11. 3 2.3 10 .8 7.6 9.0 1. 3 5.2 3.0 Suvavou 36 .6 43. 9 6.4 1. 7 10 .9 7.0 21. 5 4. 7 3.9

Wailoku 30 .3 48.7 3.5 8. 8 3.7 31 . 8 0.7 2.8

Nadoi 23. 8 39 .6 4.2 5.9 7. 3 37 .6 5.4

Nailega 16.9 38. 6 3.0 5.1 49 .3 4.0

Nasautoka 11. 3 41.4 6.6 9. 3 39 .7 3.0 Tuvatuvavatu 10 .4 45.9 54 .1

Nuku 16.9 18. 7 17. 4 1. 7 61 . 5 0. 7 ------a Beverages and mis cellaneous foods excluded ; rice treated as 'local '. b A 'person-serving ' is the unit of one person eating a particular food , irrespective of amount , at one meal.

Source: Food preferences study , Central Division , July 1976. .i::-- � 42

Summary

The argumen t of this chapter can be summarized in three maj or points. First , the marketing/distribut ion system of any one food or group of foods must be seen in th e context of both the broader food sy stem and the socio-economic environment : change in one part has ramificat ions for the broader system.

Second, the in ternal food distribution system in Fij i is poorly developed. This is due to past emphasis on external rather than internal trade links and government failure to support internal market ing , rather than to inherent weaknesses in the system. There is a substantial foundation on to wh ich improvements may be built : a long involvement in trade by Fij ians an d Indians, innovative agricultural systems , increas­ in g internal demand an d improving tran sport-conununications facilities. A realistic marketing policy would take advantage of these existing features and be developed from the pers­ pective of the total food system.

Third, food consumption habits do not change rapidly, particularly as many have sound economic justificat ion . The mo st certain way of decreasing food imports is to estab­ lish food preferences that can be met from local sources. Without drastic changes in local food preferences, food imports in Fij i will continue to be around 12 per cent of total imports by value . Food consumption habits cannot be changed by fiat . Awareness of the basis an d significance of food preferences is a prerequisite for successful inter­ vention in food consumption habits, including the consumption of non-local foods.

It is against the background of these general points that the following analyses of the produce and processed foods distribution systems should be read . Chapter 2

PRODUCE MARKETING

This chapter analyses Fiji's produce marketing system. Attention focuses on the chief parts of the contemporary system - the urban produce ma rkets and the National Marketing Authority - although evolution of the system is also des­ cribed. Key participants an d their roles are identified , as are some structural impediments to efficiency in the marketing system. The final section of the chapter summarizes problems present in the marketing system and suggests possible remedies.

Produce marketing in Fij i reflects features of the agr icultural system. Commercialism of production varies by product, region and race of producer. Probably no more than 15 per cent of total vegetable and fruit product ion en ters commercial trade. l Mo st produce en tering the market system comes from commercial-subsistence farmers who are committed to marketing to a greater or lesser degree and who retail small amo l.lllts of their own produce .

There are no guaranteed markets or guaran teed prices . When a crop is ready for harvest, a farmer must decide whether to wholesale or retail it at a produce market or to sell to a large-scale in termediary buyer such as an assemb ler­ wh olesaler of the government National Marketing Authority (NMA) . Except for Chinese farmers arol.llld Suva who grow on in formal in structions from Chinese ma rket vendors or assemb ler­ wholesalers , produce rarely has an assured commercial outlet .

The movemen t of local produce in Fij i is largely uni­ directional f rom rural producer to urban consumer (Figs 9, 10) . Much of the trade is in the hands of the producers themselves . Produce most commonly en ters the commercial system by being taken to an urban public market by the pro­ ducer where he retails it or, less frequently, sells it to a non-farming permanen t vendor. Some sales are also made to consumers in rural areas at roadside stalls or at the few recently-established rural markets.2 43 44

Rural consumer Producer

Roadside sales RURAL ZONE Wholesale buyer URBAN ZONE Hawker

Nonproducer 1 Producer Other retai I outlet vendor : vendor ------_L ------MARKET

URBAN CONSUMERS

Figure 9 Produce marketing channels

Vertical trade links are not highly developed although there is considerable variation by product . Few items pass through more than one intermediary and through more than two is un common ; much is sold without such involvement. Independent assemb ler-wholesalers and the NMA also purchase in rural areas , but generally for resale to in stitutions and hotels or for export , rather than for sale to individual consumers. Ap art from coconuts at 'gen eral stores ' and a range of produce at a few supermarkets and greengrocers in the larger towns , very lit tle fruit or vegetables is retailed through shops. 45

Hired I Buyer I Producer Hired .1 Public Transport Transport

National Marketing Authority

MARKET Export Wholesale Roadside ------, r------buyer - sales ·assembler Nonproducer : : Nonproducer ------, r------Producer 1 1 Producer

Government Hotels, 1 Individuals institutions restaurants 1 : ------�------�------CONSUMERS

Figure 10 Marketing channels of dalo

The governmen t has shown considerable interest in in ternal produce marketing over the past ten years . However , apart from in direct involvement through financing transport , connnunications and agricultural extension , emphasis has been placed mo re on establishing addit ional and different market outlets than on building onto the existing structures , of which the urban produce markets are the mo st important. Support is given to the creation of rural markets and to co-operative marketing societies , but more characterist ic of government involvement is the priority given to the NMA. Before examining the contemporary system, however, it is useful to review its antecedents as it is very much a product 46 of past experience . A review of prior developmen ts illumi­ nates two basic issues affecting the present system: the nature of government involvement and differential partici­ pat ion by different races .

Government concern with marketiE_g_: an historical perspective

Early markets

Food producers in Fij i have always taken considerable in itiative in the establishment of the commercial food trade. At earliest contact with Europeans, Fij ians bartered food for supplies and soon entered into cash exchange . Indian settlers quickly established weekly foodstuff and general merchandise bazaars within their settlemen ts and arotmd the cane mills. Apart from the bazaars that catered almost exclusively to Indian vendors and customers, the first formal market was built in Suva in 1892 from a subscription of 6d. per head from Fij ians . Five years later it was reported to be

... rapidly gaining in the favour of the natives, numbers of whom come from neighboring provinces to earn an honest shilling by the sale of fruits and vegetab les ... Th e canoes that brought these wares down the rivers may be seen drawn up along the beach opposite the market-house (Thomson 1896:9) .

Produce markets in towns and at sugar mills were a boon to town dwellers and rural producers alike , but little was done to en courage the development of internal marketing. 3 No other markets were established - the Suva market existed as the 'Cummin g Street Bazaar' tmtil moved to its present site in 1950 - and Indian bazaars were left alone by both government and sugar company.

Interest in internal produce marketing revived during the late 1930s , the impetus probably being the 1937 visit to Fij i of Sir Frank Stockdale, Agricultural Adviser to the Secretary of State for the Co lonies. In the report of his visit , Stockdale noted that Fij i was forttmate in being largely self-sufficient in food supplies and that an increasing diversity of foods was being produced and consumed by both Indians and Fij ians . Nonetheless, he was crit ical of the tmder-development of in ternal market in g systems : 47

Fiji, for the in ternal market ing of agricultural produce - part icularly foodstuffs - is very back­ ward. Central markets have not so far been created or assisted by the Administrat ion an d little thought seems to have been given to the developmen t of rural market centres. Some markets do exist , it is true , at the factories in the sugar-growing district s, but much improvement in the organisation can be made . The position in Suva it self seemed to me to be capable of considerable improvement. The contrast with the position in other dependencies was marked an d it did not seem to be reco gnised by the District Administration that agricultural pro­ duction cannot be expected to deve lop along sound lines unless facilities for the dispo sal of produce

are afforded ••.

The organisation of facil ities for the internal market is of even greater importance to the small producer than the developmen t of facil it ies for eX,ort (Stockdale 1938:38) .

At the time of Stockdale 's visit , produce marketing was largely semi-regular trading by pro ducers , either from a fixed site, often on or near a town main street an d at a fixed time - Saturday morning being the most popular - or by mob ile hawking. The market sites , like the In dian bazaars and even the Suva 'market ', were unimproved and trading un regulated. Immediately following Stockdale 's visit , concern for internal mairketing was apparent. Interest came from two sources : first , local authorities anxious for 'more orderly ' trading and for in come from vendors to supply improvements for this purpose; second, from producers seeking more effective outlets. The establishment of markets in Sigatoka and Nausori illustrates these forces.

The Sigatoka market , opened in 1939 , replaced the Saturday mo rning trading which had been conducted for many years from store verandahs and the banks of the Sigatoka River. The 'market ' was a demarcated area with no improve­ ments : the only shelters in th e area were those constructed by vendors . Vendors were not charged for use of the market , which functioned only on Saturday, but produce sales else­ wh ere in the town were prohibited. Within two mon ths the new market was being used by forty Fij ian and twenty Indian producers (Surridge 1940:32) . 48

Whereas the impetus for the creation of a market at Sigatoka came from local administrative officers dissatisfied with marketing arrangements - chiefly the overflow of pro­ ducers on to 'pub lic ' thoroughfares - the initiative at Nausori was from exempted Fijian farmers (galala) seeking an outlet for their produce . With money advanced from the Depart­ men t of Agriculture , a market building was erected and opened in January 1940, timing that proved fortuitous as the market became a maj or source of produce for the war-time military forces . 4 In 1942 , 1500 tonnes of produce valued at £20,000 was sold. The market was paid a connnission on sales, produce

being accepted from 'any Fij ian grower ..• and occasionally even Indian and Chinese produce is handled' (Harvey 1943 : 36) . As a stimulus to production and to improvements in ·production methods , input distribut ion and marketing, the market was a success, but this was largely due to the extra­ ordinary circumstances of war-time demand and intensive government involvemen t.

Investigations

Largely in response to the implications of the Stockdale report and war-time dislocation , two other investigations in to agriculture were conducted in the mid-1940s. The resultant reports (Shephard 1944 , Patterson and Dodds 1945) emphasized the impo rtance of in ternal marketing and revealed a fear that the improvements achieved during the war period would be short-lived without positive government act ion . The Patterson and Dodds report included reconnnendations that a marketing officer be appointed to assist small-holders dispose of pro duce , that rural ma rket areas be established and that co-operative marketing ventures be stimulated. All this would be in addition to continuing developmen t of external markets.

Coincidentally with the investigat ions of Patterson and Dodds and Shephard, there was considerable act ivity in internal marketing. Old markets were reconstructed and others established - with loans from the government - under local government regulat ion s and the country's first Markets Ordinance (1945) . Markets were re-established or foWlded at Sigatoka, Labasa and Nadi in 1948, at Nausori in 1949, at Suva in 1950 and at Tavua , Ba and Lautoka in 1952 . Markets we re established at Vaileka an d Levuka during the 1950s. Competit ion with the new outlets was curbed either by local By-laws or the Markets Ordinance. The Markets Ordinance , for in stance , made it possible to prohibit sales 49

of produce within three miles of an offic ial market , except at a licen sed wholesale or retail outlet . Health regulations and other legislation were also used to concentrate market ing in the new cen tres . For fifty years prior to the opening of the ma rket at Lautoka in January 1952 , wh ich was built with an £11,000 government loan , a Saturday mo rning bazaar had been held on the ro adside opposite the Colonial Sugar Re finery 'Toplines '. In June 1952 the bazaar was closed by local Health Authorities and the Township Board; vendors were forced to sell in the new market.5 Unregulated produce trading continued in other centres on ly until an official ma rket was established an d local market regulations were en acted; this included the prohib ition of tmlicensed produce sales outside markets. Formal markets were established at Korovou and Navua during the 1960s and later at Savusavu (1973) and Waiyevo (1974) . Today all 'urban ' centres have an official public produce market ; Suva has two , a second having been opened at Raiwaqa in 1971. There are also three small official 'rural markets ' at rural service centres.

The formal markets quickly showed signs of success. Nausori market, for example , was described by the District Connniss ioner as 'a peculiarly successful innovation '. In July 1949 it was patronized by over 1300 Fij ians, 536 Indians and 82 Chinese vendors, each of whom had a daily gross income of one to two pounds. 6 Presumab ly most vendors at the markets were producers, although at Nadi and Sigatoka some sales were handled by Fij ian co-operative societies.

Local government bodies realized that , in addition to being of benefit to producers and consumers, markets had other advantages . It was apparent that market revenue could easily outstrip expenditure, even allowing for capital investment. In 1951, for in stance, the market was the source of 42 per cent of the locally-raised revenue of Nausori township . Many markets continue to be important sources of cash revenue . In maj or towns they raise more than $20,000 per annum (up to $173,000) with as little as one-quarter of this being accotmted for by expenditure (Tables 2.1 and 2.2) .

Internal food marketing re tained a low priority in governmen t plans, despite Stockdale 's report , the broad concurrence of Shephard 's and Patterson and Dodds' findings, and the success of the new urban produce markets. Close involvement was difficult in that markets we re the respons­ ib ility of local government bodies and thus were effectively V1 Table 2.1 0

Revenue ($ '000) of selected markets:2 1966-75

1966 1967 1968 1969 1970 1971 1972 1973 1974 1975

Suva 80 .2 85 .8 88 .5 98.8 104.7 146 .5 148 .9 158.l 172. 5 Nausori 2.5 2.7 5.5 5.6 6.6 8.7 15 .3 15 .8 20 .2 Nadi 6.4 6.6 8.8 10.1 15 .2 15 .7 17.5 25 .8 Lautoka 34 .9 Ba 9.1 8.0 12 .4 11 .2 11 .5 22 .3 Sigatoka 1.5 1.5 1.9 3.8 9.3 10 .0 13. 3 18 .6 Labasa 6.4 7.0 7.2 9.8 12 .0 16 .3 18 .8 18 .5 22 .1 Savusavu 1.6 2.6 3.2 Levuka 0.2 0.2 0.2 0.5 0.5 0.4 0.4 0.5 0.7

Note : .., - ' indicates no information available .

Source : Local government records , Ministry of Urban Development, Housing and Social Welfare, Suva . Table 2.2

Revenue and expenditure ($ ' 000) of selected markets, 1975

Marke t revenue Expenditure Total city/ Revenue Expenditure as % of total as % of revenue town revenue revenue

Suva 172.5 80.2 46 .5 ) ) 1,824 .9 9.8 Raiwaqa 5.9 27.5 466 .1 ) ) Nausori 20.2 6.3 31 .2 77. 7 26 .0 Lautoka 34 .9 21. 7 62 .2 452.5 7.7 Sigatoka 18 .6 4.6 24 .7 74.6 24 .9 Nadi 25 .8 13 .5 51 .2 n.a. n.a. Ba 22 .3 12 .0 53.8 114.9 19.4 Labasa 22.1 5.6 25 .3 n.a. n.a. Savusavu 3.2 n.a. n.a. 22 .5 14.2 Levuka 0.8 0.6 75 .0 18 .4 4.3

No tes: 'General Revenue ' is comprised primarily of rates , licence fees and service charges, an d exclud es income sources such as electricity and sewerage accounts. Accounts of markets administered by Rural Local Authorities (Korovou, Navua , Tavua , Waiyevo and Vaileka) are not available . Source : Local government records, 1975 , Ministry of Urban Development , Housing and Social Welfare , Suva . 52 beyond the ambit of direct central governmen t influence. Several attempts were made to develop in ternal marketing. In 1948, for example , a marketing officer was appointed. Within a year , however, he was tran sferred to the Departmen t of Co-operatives, wh ich was then involved in produce market in g. 7 Three Economic Development Officers were appointed to develop commercial participation by Fij ian producers in the mid-1950s. Th ey , too , were generally tmsuccessful in a difficult task , notwithstanding their concentration on export crops rather than produce for the local market . 8

The Spate and Burns reports

The Spate and Burns reports (Spate 1959; Burns, Watson and Peacock 1960) are import ant for their being watersheds in attitudes towards Fij ian economic development . Both reports emphasized the central role of marketing in agri­ cultural development. Spate claimed, 'marketing is perhaps the mo st intractable of the purely economic factors hampering Fij ian development ', and stressed the importance of the in ternal market (Spat e 1959:48) . Problems associated with marketing were isolated by Spate as the growers ' failure to fulfil contract commitment s and their insistence on advance cash payment, the government 's acceptance of ad hoe marketing organizat ion , a lack of understanding of the role of middlemen , poor communicat ions and a lack of storage facilities. Spate believed the marketing situat ion could be improved with schemes for particular (export ) commodities and involvement of economic developmen t officers and co-operative societies, but the report did not include many specific recommendations. Th e Burns report was even less specific in its recommen d­ ations: improved communications and roads 'would help lessen the degree of market imperfection ' and changes in land tenure in the long run will, it stated, 'make the Fij ian more familiar with the elements of economic organizat ion in a monetized economy , en couraging him , particularly , to familiarize himself with market conditions '.

Th e findings of the Spate and Burns report s reflected government attitudes to in ternal marketing. It was accepted that it was necessary to stimulate internal trade but means to this end were not clear. Whether through this uncertainty or because of a low priority to marketing in reality, the problem was approached piecemeal and received few resources. The national developmen t plans of 1961-65 and 1964-68, for in stance, failed to mention internal marketing at all . Even the 1966-70 plan had on ly one reference to internal marketing. 53

Reasons for the lack of connnitment to marketing were prob ably shortage of funds , ab sence of leadership an d hostility towards middlemen .

Development during the 1950s and 1960s

In spite of the weak lead from government , trade between rural producer and urban consumer continued and even expanded during the 1950s and 1960s. Some trade movements were remarkably extensive given the state of conununicat ions and roads . In the early 1950s , for example , dalo an d other vegetables from the Suva and Nausori areas were occasionally sold by producers and assembler-wholesalers in the Western Division . 9 In the 1950s, such movements were confined largely to Indian pro ducers an d assemblers , but during the following decade Fij ian producers became increas ingly involved in supplying urban markets. One source of produce was small in dependent farmers on settlement schemes organized, from 1961 , by the Land Development Authority. The schemes were established to grow export crops, but many farmers soon turned to produce such as root crops, which could be sold on the local market. For example , the Lomaivuna scheme in Naitasiri was established to produce bananas for export . By the late 1960s farmers on the scheme had converted to mixed cropping for the Suva market.10

Another factor behind in creased Fij ian participat ion in marketing - and a strong influence on governmen t at titudes toward marketing - was the insistence of producers in Lomaiviti and even Lau that they receive assistance in putting produce, particularly yaqona, on the Suva market . They wanted a market in g organizat ion that would give higher returns to growers than were obtained through the established mi ddlemen channels . Government response marked another phase of government involvement in marketing. The Department of Co­ operatives - through marketing unions - and the Fij ian Af fairs Board with a produce-purchasing agency at Suva wharf operating in conj unction with the Department of Agriculture , became involved in these marketin g efforts. The Koro and Gau Marketing Un ions we re the mo st efficiently organized, with agents in Suva to handle shipmen ts and retail market sales . Both , however, we re short-lived an d financially unsuccessful . The Fij ian Affairs Marketing Scheme was phased out on ce the National Marketing Authority was established.

In addition to the government-supported market ing organizat ions, considerable entrepreneurial links existed 54 between islands and urban markets. Between Beqa and Suva , and the Yasawas and Lautoka, where distances were short and there were few non-Fij ians on the islands , trade was largely in the hands of Fij ian producers. 11 Lomaiviti, Kadavu , Lau and Cakaudrove were served either by itinerant or resident Indian buyers (in the case of yaqona) or by producers bringing their own goods to market. As kin settled in Suva , growers informally 'consigned' an increasing amo tmt of produce to resident relat ives for sale in the market. This form of 'con signment ' transfer remains a chief method of market participat ion for island producers .

Notwithstanding the ad hoe government involvemen t in marketing, reports of at least two FAO experts and the short­ term appointment of a local market investigation officer in 1961 , 12 an event that contributed greatly to the awareness of the significance of in ternal marketin g was the visit to Fij i in 1965 of Maj or H. C. Biggs , a marketing expert , sponsored by the British Ministry of Overseas Development. Biggs was to investigate the organizat ion of marketing of local produce and to recommend improvements that should and could be made to the efficiency of the system. His Report on the Marketing of Agricultural 'Produce in Fiji (Biggs n.d. ) was important not only for its specific recommend­ ations but also for focusing attention on the distribut ive and generative features of the internal food market in g system.

Awareness creat ed by the Biggs investigation was possibly a maj or factor behind the 1968 decision of the Developmen t Committee that the government should be an active agent in the marketing of produce , both locally and overseas. A marketing division was established in the Ministry of Natural Resources to promote produce sales , provide market intelligence and to act as a general focus of public and private sector activity to improve produce distribut ion . In 1969 the mar­ keting division was transferred to the Ministry of Commerce, In dustry and Co-operatives , whence came the suggestion that a connnercial marketing organization should be created. This in itiative led to the format ion of the National Marketing Authority in 1971. It should be noted , however, that wh ereas Biggs did encourage the development of a nat ional marketing body , he saw the parallel improvemen t of existing marketing facilit ies as a basic means of general market improvement. 55

Legislat i�n a.lid change

Governmen t involvement in internal produce market ing is reflected through legislat ion relating to the produce markets and the NMA. Legislat ion relating to the NMA will be covered below . Market legislat ion is reviewed here to give an idea of changes in both the form of marketing and government at t itudes toward marketing.

Prior to the mid-1930s, no legislation dealt specifically with markets, although they fell under public health and other legislation . The first market legislation was enacted by the Suva Town Board under the Towns Ordinance, 1935. The Suva (Markets) By-laws , 1937, established the role of the market as the chief form of urban produc e outlet : the sale of produce , or any other goods, elsewhere in town without a 1 licen ce was prohibited . 3 The By-laws established general market organizat ion and administrat ive procedures , in cluding the powers of the Market Master, fee schedules and the general conduct of the markets, wh ich included a ban on shouting or advertising by the vendors.

The first legislation dealing specifically with markets outside gazetted towns was the Markets Ordinance , 1945 . Al though the Ordinance does not apply to markets within towns , it has had considerable influence on their legislation through being taken as a guide to areas of authority. The influence of the Ordinance was strongest in the late 1940s when many local government bodies were enacting or revising market regulations . The current un iformity in form an d function of markets in Fij i is in part due to the similarity of market regulat ion s.

The market regulations developed under the Markets Ordinance prohibited the sale of produce within two miles of a market save through a licensed retail or wholesale outlet .14 Re gulat ions under the Ordinance were more detailed than any previous market regulat ions , despite the fact that trading in the markets to which they applied did not approach the sophisticat ion of town markets . Produce was de fined as 'agricultural and forest produce , fish, fresh meat , fresh fruit an d vegetables , poultry and eggs' . Advertising, shouting and auct ions within markets were illegal . At other markets 'produce ' was broadened to include local tobacco and handi­ crafts and at Nausori, Levuka and Suva included rice, split peas , pulses and 'goods of that nature '. It was only at Ba 56 and Labasa that this round of 1947-48 market regulat ions did not define goods that could be sold.

An impression of the changing nature of market trade is suggested by amendments to market regulations. Men tion of producer and non-producer trade was first made in 1953 when the Suva and Levuka regulat ions were amended to give preference in the allocation of selling space to 'producers and their agents selling direct to the public'. Another indication of the increasing sophistication of both vendors and consume rs was that in 1954 the Suva market regulations were amended to make man datory the display of selling prices.

A new phase in legislative development occurred in the 1960s. At Suva an d Lautoka, regulations made in 1962-66 outlawed wholesaling, prohibited the exchange , subletting and multiple occupancy of stalls, required the prices of fish to be marked with figures three in ches high and allowed handicraft dealers to sell 'provided that stall space is available ', the preference in selling space going to producers . This latter provis ion has become a bone of contention. In June 1972 the Suva City Council proposed to end the prefer­ ential treatment given producers , but the Central Farmers Association won an inj un ction in the Supreme Court that pro­ hibited the move . The inj unction still stands and handicraft dealers are prohibited from using the market after 2pm on Fridays and all day Saturday. 15

Market regulat ions are intricate but all are not enforced , which is probably just as well as some are inappropriate to the present situation . The regulations most common ly not enforced deal with the display of prices on products, the pre ference given to producers or their agents , the prohibit ion of wholesaling and the type of produce offered for sale. Most vendors dealing with potatoes, onions an d rice display prices; few others do . Conflict is common between producer and non-producer vendors , also between foodstuff and handi­ craft vendors at the markets where there are handicraft dealers , over the shortage of selling space and the occu­ pation of prime selling sites . Where regulations favouring pro ducer vendors are en fo rced at all , it is usually by limiting the space available to handicraft sellers an d the times they may operate , rather than by restricting stall space for non­ producer foodstuff vendors . Al though wholesale transactions are illegal in mo st markets, they are the only means by wh ich non-producer vendors are ab le to ob tain stock so they continue both within and around market buildings. If wholesaling 57

farme rs or non-producers are encountered in side a market (while making a delivery) , they may be charged the standard retail vendor's fee but this regulation is not common ly en forced .

Market regulat ions were drawn up with local producers in mind at a time when both perman en t vendo rs and imported food sales in markets were insignificant . Today , the place of both is often questioned by producer vendors who are unable to find selling space , and also by grocery storekeepers who fear competit ion . The issues are clouded because imported items are mainly basic staples (rice and pulses) an d luxury products (apples , etc. ), of wh ich there is limited or no local production . Imported items that are unprocessed foods have generally been permitted in markets, but the sale of imported food remains a bone of contention , especially when related to confl ict between producer and non-producer vendors .

Both documented and popular opinion suggest that the internal marketing of local products has changed greatly since official produce markets became firmly established in the early post-war period. The vo lume of trade , for instance, has increased in response to substantial growth in the urban and consumer populations and the numb er of producers : an indication of trade volume is the in crease in vendors as revealed by market income in the ab sence of better data (Tab le 2.3). Another change is greater market in tegrat ion with the expansion of producing areas that are linked to purchas ing cen tres. Th is has occurred mo st markedly with the involvement of the NMA in the produce trade . The extens ion and improvement of roads through Viti Levu an d Vanua Levu has also en larged cons iderably the area accessible to market. Wh ile hinterlands have expanded, there has been also some fractioning of supply areas with the establishment of local markets as at Korovou and Navua and areas without formal markets are now being served by either official town (Waiyavo an d Savusavu) or rural markets. Un til the early 1960s, the nature of market trade can only be in ferred,16 but it is likely that the number of vendors at all markets was small ; weekend as opposed to weekday markets were mo re significant than today , permanent vendors were few and mo stly Indians, and Fij ian part icipation was limited to end-of-week producer vendors, not discotmting more common hawking and roadside selling. 16

Un til the involvemen t of the NMA , marketing charges have been more in scale than in form. Main findings of Kamekamica' s \JI 00 Table 2.3 ---- a Vendor attendance at ma rkets

b Different vendors Number of vendors Fri .-Sat . vendors Urban population over week Market as % ('OOO, September 1976) Mon .-Thurs. Fri. Sat . Total % of whom are week ' s total daily perm. vendors

Suva 117 .8 250 450 650 58 680 27 Raiwaqa (n .a. )C 14 14 14 33 20 70 Nausori 12 .8 50 70 300 65 330 16 Korovou 0. 3 10 20 30 56 45 0 Vaileka 3.8 20 40 40 50 65 25 Tavua 2.1 45 70 100 49 130 22 Ba 9.2 50 90 300 61 270 25 Lautoka 28. 9 110 200 400 58 450 23 Nadi 13.0 70 100 150 47 250 36 Sigatoka 3. 6 40 70 100 52 130 26 Navua 2.6 25 80 100 64 12 0 21 Levuka 2.8 1 20 20 90 50 0 Waiyevo 0.1 0 10 35 100 35 0 Savusavu 2.3 10 10 30 50 45 7 Lab asa 13. 0 70 90 180 49 210 22 a Vendor figures are approximate . b ' Numb er of vendors ' is derived from daily estimates of total vendors . c Population of Raiwaqa is included in that of Suva .

Source : Bureau of Stat istics (urban populat ions) and market surveys , January 1976. 59

1962-63 market surveys (Kamekamica 1966 ) were that markets were supplied primarily from nearby, adj acent areas ; middlemen were more important in some markets and for some products than others; and that the share of trade handled by racial groups did not correspond with their representat ion among all vendors.17 These characteristics are still identifiab le but are more marked. Apart from the greater market in tegration of the economy , the main change since the early 1960s has been in specializat ion by vendors, which in cludes Fij ian vendors , producers and non-growing suppliers . Specialization in these fields was noted by Kamekamica, but apparently it was on a much smaller scale than today.

The situation described by Biggs (n .d.) also differs from the present more in scale and sophisticat ion than in basic marketing form. Biggs saw Suva and Lautoka as in cipient regional markets: they are unden iably so today . Producer vendors remain important in market trade , but they accoun t for considerably less than the 80 per cent of trade estimat ed by Biggs . Greatest change has occurred among non-producer vendors. As ide from a considerable in crease in number, 18 their trade has expanded into root crops and mo st other vegetables, rather than being confined almost exclusively to imported goods as Biggs reported. One of Biggs ' central recommendations to improve efficiency of produce marketing was the creation of a professional marketing class. Ind ications from the number and role of middlemen suggest this is being achieved , though not with the degree of concentration Biggs advocated. It appears also that the numb er of rural assemb ler­ wholesalers - or at least their importance in market supply - has in creased , but informat ion is not available to permit precise evaluation of the extent of change.

Al though production remains the main area of government concern in limiting food imports , the significan ce of an effective distribution system is also acknowledged . The concern is , however, both recent and selective . Individual marketing entrepreneurs who are either vendors or assembler­ wholesaiers and even the established produce markets have not received strong tangib le support. New rural and urban markets have been built , but otherwise the government has devoted its resources to establishing the NMA and orienting rural production to the Authority. Despite these limitat ions, this involvement of the government is a maj or change in food ma rketing in Fij i. 60

Urban nroduce markets

Market s in Fij i

Since earliest Western contact , local produce has been sold in Fij i by hawkers in areas where a non-farming popu­ lation has become established and later at Indian 'bazaars '; the creation of formal market places is relatively recent . As ide from one in Suva , formally administered markets were first established in the late 1930s; none of the present market structures is much older than twenty years and a number are considerably more recent . Today there is a public market in each of Fiji's two cit ies (Suva and Lautoka) , six towns (Nausori, Labasa, Nadi , Ba, Sigatoka and Levuka) and six townships admin istered by township boards or local rural authorities (Vaileka , Korovou , Navua , Savusavu , Tavua and Waiyevo). There is also a second market in Suva at Raiwaqa. All markets are administered by local government bodies.

At one level , markets in Fij i appear surprisingly similar. The building is likely to be open-sided an d iron-roofed, situated in a fenced-off compound that allows easy enforcemen t of ma rket regulations. Goods are mainly sold from wooden or concrete benches within the building, although the ground out side (but still within the compound) may also be marked off as selling areas . Apart from an office for the market master, a lockable room for storing of produce and perhaps a sect ion exclusively for selling fresh fish , there is little internal construction . Most vendors sell locally-grown , unprocessed food pro ducts that they set out on a selling area for which they pay a daily fee ranging from 2 5 or 30 cents to about a dollar, depending on type of goods sold, location in market and the actual market. 19 Vendors who do not grow food buy their stocks at wholesale prices from farmers , or less connnonly from non-producers, found on the edges of the market ; practically all transactions within the market are retail to urban-dwelling consumers .

Weekday market hours are generally from 7am to 5pm. Markets close at lpm on Saturday and do not open again until Monday morning , although vendors may try to sell surplus produce outs ide the market compound until dark. Food unsold one day may be left on the selling table until the next or , for a charge, locked in the storeroom. At mo st markets, stalls mu st be completely cleared for fumigation at least on ce a mon th , on a Saturday after trading, until the following 61

Monday . Otherwise , produce can be left in stalls at the own ers ' risk from day to day . In a few market s some lock-up shops are available for hire by vendors . As permanent vendors continuously store produce within their stalls, many markets take on a more permanen t appearance than perhaps would be in ferred from market regulat ions.

Three basic types of goods are offered for sale; prepared food, tmprepared food and non-food. About 90 per cent of vendors sell only un prepared food; unprepared foods combined with non-foods , which comprise mainly plastic shopping bags and sasa ( coconut rib brooms) , are sold by most other vendors . Apart from handicraft ven dors in markets visited by tourists, all vendors selling non-food items also sell foodstuffs. Non-food sales are not important in the total context of market sales. Around the market en trances and even at a few places inside , some people sell ready-to-eat foods prepared by the vendor and his family. Common snack foods are Indian sweets, peanuts, sliced pineapple and water­ melon , cooked corn and cordial drinks . Complete meals are not sold. All sales in markets are for cash . Prices of goods are rarely marked and are generally non-negotiable. Sales techn iques are passive : vendors do not actively pursue sales.

A numb er of other products are sold in addition to vegetables and fruit . Commonly presented products in clude fresh and sea water foods like turtle, crab , prawn , shell fish, beche-de-me r and seaweed; eggs and live poultry, but very few other animal products; yaqona and locally-grown rolled tobacco; rice and spices . Some imported fruit such as apples� grapes , plums and pears and vegetables such as lettuce , tomatoes , potatoes , onions and garlic are sold; many spices and some rice and yaqona are also imported. Vendors specialize in the type of goods sold: some sell only green vegetables and small amounts of fruit wh ile others deal in local root vegetables. As fish an d sea foods must be sold in a separat e section of the market, their vendors generally do not sell other produce . Sellers dealing exclus­ ively in potat oes , onions, spices , tobacco and yaqona, like those who sell only eggs , stand out as being trading specialists.

Markets are the chief source of fresh fruit , vegetables and sea food for the maj ority of urban dwellers; home gardens or the consumer' s own fishing activities are the only other 62

important source of supply . The consumer 's race has consid­ erab le in fluen ce on the type of food purchased; indeed , markets are characterized as much by the race of the consumer as that of the vendor. 20

Much market attendance by consumers is fairly casual in intent . Many people wander into the market to look around, and perhaps have a snack, especially at lunch time � in the afternoon after school or work and on Saturdays . Markets, even when very crowded as is often the case with the larger ones, are interesting and stimulating social centres.

Vendors vary considerab ly between markets in number, racial composit ion and permanence, that is whether they are producers or non-producers; the relative significance of week-day and end-of-week sales also varies (Tab le 2.3, Figs 11-12). Suva market is by far the largest with about 250 vendors present each day from Monday to Thursday and over 600 on Friday and Saturday . At the eleven markets where permanent vendors are most important, they comprise around one-quarter of all vendors. There is a correlation between market size , measured by the number of consumers or vendors , and the populat ion of the centre in which the market is situated. The five smallest markets, apart from Raiwaqa because it is a suburb of Suva, are in the country's smallest towns and have the lowest rates of non-producer vendor parti­ cipat ion . In spite of the differences in size and vendor composition , however, markets in Fij i have much in common and in many ways it is possible to treat markets and market vendors as one.

Market types

The fifteen public produce markets can be divided in to several catagories according to various characteris tics . The vendor's race and type of produce are two that are examined below. Others are the numb er of vendors and import­ ance of the end-of-week market. 21

By the number of vendors, the markets fall into five size categories. Generally the ranking of markets in terms of vendor numbers over Monday to Thursday is maintained on Friday and Saturday (Table 2.3). The market with the greatest number of vendors throughout the week is Suva market and as this market has twice as many vendors present on any day as has the one at Lautoka , the next largest market, it is best Race of vendor

Fi1ian CJ Indian � VLevuka0 Chinese r � ���J

700

350 200 100 50 Number of vendors

10 km 50

Figure 11 Vendor numbers an d racial composition at produce markets 64

f ' \ \ .... \ \ ,,I I I \ .... _1 ' I '' .- - ', .._ I -/ l l ,,""\ I / I I ,,. J I / 1_/ /

/ / / / ( J I N�!.'fASIRI I l, \ \ ... "'-,,,,.. --...... ;:··· I

Number of vendors ISO Fipan � Permanent 100

dian 60 In _ Non .:.:: permanent 20 Chinese ;·::. a 0 ---- Province boundary km 30

· ...... ····· Tikina boundary

Figure 12 Vendors at Suva market : attendance by vendors ' race , permanency, day of attendance and resi­ dential location (Source : Fieldwork) 65

regarded as forming a single category .

Lautoka market , with 110 week-day vendors, is in a second category . A third size category comprises Labasa and Nadi markets, each with ab out seventy vendors during the week , rising to 150-180 on Saturday . End-of-week markets ar e mo re important in a fourth group comprising Nausori, Tavua � Ba, Sigatoka and Navua . Ba and Nausori, with Saturday markets about six time s greater than on Monday to Thursday , have larger end-of-week markets than Labasa and Nadi, but have a lower trading level during the week. At all markets, apart from those in a final, fifth group , about one-third of vendors sell on Saturday and one-half on Friday and Saturday combined. The fifth group comprises Raiwaqa, Korovou, Vaileka, Levuka , Savusavu and Waiyevo markets. Here , there are no , or very few , sales on week days. At all these markets save Raiwaqa where activity is negligible on any day , end-of-week sales are the main activity.

Two main types of market , the Local Produce Outlet (LPO) and the Trading Centre Market (TCM) , are identifiable by the extent of their hinterlands. LPOs have restricted hinterlands and are located in small centres servicing the limited demands of the surrounding rural populat ions. Trade is handled largely by producer vendors for whom the market is often a prime sour ce of in come . Waiyevo , Levuka and Korovou, and to a lesser extent , Navua an d Savusavu , markets are LPOs . In contrast , the hinterlands of TCMs are extensive , even con­ taining other market s (Fig. 13) . TCMs are no less important to producer vendors than other markets but non-producer permanent vendors are the trading core , accounting for one­ fifth to one-third of vendors . Suva , Lautoka, Ba, Nadi, Nausori, Labasa and Sigatoka markets are TCMs . Suva an d Lautoka markets have extensive hinterlands, but apart from this and their larger trade volumes, these markets are not significantly different from other TCMs .

Tavua , Vaileka and pos-sibly Navua market s comprise a middle group between the TCMs and LPOs . These market s are essentially outlets for local producer vendors , but they also have a smallnumb er of permanent daily vendors .

It may be possible to chart an evolut ionary progression from Local Produce Out let to Trading Centre Market, but in doing so on e must not lose sight of vital factors in market format ion and growth. The socio-economic character of the 66

· � ..·· · ·· · ··· · · Supply area. named market

Supply area. Suva market

Supply area. La utoka market

0 10 km 50

Figure 13 Supply hinterlands of urban produce markets (Source : Fieldwork) 67 populat ion of the hinterland and centre where the market is located has a fundamental influence on market form. Similarly, market size reflects both demand of consumers living in the centre wh ere it is located and , to some extent, the degree of the hinterland's involvement in cash agriculture . Non­ producer vendors and the population of the larger urban centres are largely Indian so it is likely that markets in the larger towns will have a core of perman en t Indian vendors . On the other hand, all markets started as LPOs . It remains to be seen whether the present LPOs , which are the smallest markets and those mo st dominated by Fij ian vendors, will become TCMs as demand increases and whether the non-producer trade en tailed in this extens ion will be taken on by Fij ian or, as in the past, Indian vendors .

The market vendors

General characteristics. Market vendors in Fij i are not conspicuously differen t from the population from which they are drawn . Th ere is no distinct occupational class of 'market vendors ' although vendors do share a number of connnon characteristics. There is considerable range in the scale of vendor operation . Permanent and non-permanen t vendors represent two distinct forms of selling activity. Although a significant numb er of vendors are urban dwelling and permanent, mo st vendors come from rural areas , and live near the ma rket at wh ich they sell . There are no 'mobile ' traders in the sense of people who systematically sell at different markets. Vendors rarely sell at other than their 'regular ' market : on ly 10 per cent of vendors covered in the January 1976 surveys claimed that during the previous year they had sold at any market othe r than wh ere they were being interviewed.

Almost all vendors are self-employed and work alone. Only one-third of the 5 per cent or so who are not self­ employed are paid employees. Occas ionally stalls are operated by village or women 's clubs, the earnings of wh ich are pooled to a connnon cause. Ab out one-fifth of vendors have help in selling. Assistance is generally limited to one additional person , often a relative of the vendor. Only permanent vendors have paid assistance in selling, but many permanent vendors have no help at all , while non-permanent producer vendors are likely to be helped by unpaid relatives.

Male vendors are slightly more common than female (43.1 68 per cent are female) . Three-quarters of vendors are married; one-half are household heads and another one-third are wives of household heads. Almost all vendors have had some formal education but few have pro ceeded beyond primary school. Only 5 per cent of vendors have not had any formal educat ion , although 19 per cent have had less than one year . About one-third have completed primary school and 5 per cent have had more than ten years of formal education .

Market vendors come from rural backgrounds (Table 2.4). Sixty per cent of the heads of household, wh ere the head is not a market vendor, have agricultural occupations. Un skilled labour and semi-skilled trades account for a further 14 per cent . The fathers of vendors were more likely to have had rural occupat ions (77 per cent) than their vendor children . As only 5 per cent of vendors ' fathers were permanent market vendors there is no evidence that market selling is a here­ ditary occupation , although this could be attributed to the fact that daily markets are a relatively recent phenomena.

The races of Fij i are represented among vendors in roughly the same proportion as in the total population : Indians and Fij ians acconnt for 54.7 and 43.1 per cent of vendors respect ively; 1. 1 per cent are Ch inese and 0.6 per cent are 'Other Pacific Islanders ' (Table 2.5). Indian vendors are predominantly Hindu� Only nine of 1107 Indian vendors enumerated were Guj eratis, traditionally the commer­ cially�oriented Indian group . There is no apparent regional bias in place of origin of Fij ian vendors , who come from practically all the tikina of Fij i, although vendors from Lau and Lomaiviti are more numerous at Suva than other markets. While race and place of origin of vendors may show no signs of bias , each race has distinctive forms of involvemen t in marketing.

Both vendors and their families are highly dependent on the market for income . For almost three-quarters (72.6 per cent) of vendo rs , market sales are the only income source. Of vendors who have other sources of income , farming is the mo st important for 71.7 per cent , including cane farming for 136 vendors (29 per cent) in the Western Division . The next mo st common source of income is nnskilled labour (8.2 per cent). About half the vendors ' families are totally dependent on market sales for income : market sales are of minimal importance for only on e-tenth of families . Among vendor families with other in come sources , agriculture scores low (39.l per cent of 381 responses) . Agriculture is in fact 69

Table 2.4

Occupation of head of vendor's household (if not vendor)

Occupation Frequency Per cent

Farmer or villager 491 58.0 Cane farmer 18 2.1 Agricultural sub-total 509 60.1

Fisherman 8 0. 9 Skilled trade 21 2.5 Semi-skilled trade 64 7.6 Unskilled labour 53 6. 3 Professional 16 1. 9 Semi-professional 19 2.2 Shopkeeper, businessman 11 1. 3 Market vendor 58 6.8 Non-specific employment 25 3.0 Un employed - any reason 63 7.4 Not available 199 missing Total 1,046 100. 0

Source : Market surveys, January 1976. 70

Table 2 .5

Race of market vendors

Race Frequency Percentage

Fij ian 866 43.l Rotuman 8 0.4 Other Pacific Islander 5 0.2 Sub-total 879 43. 7 Mo slem 162 8.1 North Indian Hindu 726 36 .2 South Indian Hindu 159 7.9 Guj erati 9 0.4 Punj ab i 23 1.1 Un specified Indian 22 1.1

- Sub-total 1,101 54.8 Chinese 23 1.1 Part-European 4 0.2 Other 1 o.o Not available 2 missing Total 2,010 100.0

Note: Figures refer to different vendors, irrespective of how many days they are at market. If race was assigned on the basis of vendor-days, the Indian and Chinese shares would be larger . For example , at Labasa 79 .7 per cent of different vendors are Indian, while 86 per cent of vendor-days are Indian. At Ba, the only other market where a count of vendor-days was made in November 1976, Indians accounted for 70.9 per cent of vendors and 81.0 per cent of vendor-days.

Source: Market surveys, January 1976. 71

probably mo re important than this suggests as where a vendor's family has a farm, input from memb ers other than the vendor, especially wh ere he is household head, was probably not listed as an addit ional source of income . Unskilled labour and semi-skilled trades are the chief other in come sources for vendor families.

Most vendors live near the market at wh ich they sell . The average distance between home an d market for all vendors is 16km (Table 2.6). On e-half of the vendors live less than llkm from the market at which they sell an d three-quarters live within 16km. Only 10 per cent live mo re than 32km away . There is considerable variat ion in the distance trav­ elled according to the day of the week, reflecting the greater end-of-week involvement of producer-vendors an d con comitant hinterland expansion and contraction . For instance, the average distance travelled on Monday and Tuesday is lOkm and in creases throughout the week to become almost double on Friday and Saturday .

The distance travelled by permanent non-producer vendors remains relat ively constant during the week but is less than one-half the market-to-residence distance covered by non­ permanent producer vendors. Travel distance of non-permanent vendors increases sharply on Wednesday and remains high for the remainder of the week . Difference in travel distance between producer and non-producer is greatest at Lautoka where producer vendo rs live on average about 30km away from the market compared to llkm for permanent vendors. On the whole, differen ces in distance travelled by producer and non­ producer vendors are greater in the Western Division than elsewhere .

Travel time and fares paid refl ect the relatively short distance between vendors ' residen ce and market , and for mo st vendors do not add greatly to marketing costs. Sixt een per cent of vendors live within ten minutes of the market and two-thirds within half an hour . Only 9 per cent live mo re than two hours aw ay . Of the vendors who pay person al travel fares , 81 per cent pay less than $1 and one-half of these have fares under 20�; only one in twelve pays more than $2 .

The mo st connnon means of transport used by vendors to go to market are buses and 'carriers ' - trucks that collect paying passengers; together they are used by three-quarters of the vendors (Tab le 2.7) . Boat and private vehicles are Tab le 2.6

Re sidence-to-market distance (km) of market vendors

Vendor tyEe Vendor race

Non - All vendors Producer producer Permanent Pacific Indian Other vendors vendors vendors islander

All markets 15 .8 19.8 8.5 n.a. 22 . 7 11 .1 4.2 Suva 17 .2 2 4.8 8.9 9.0 10 . 7 . . a 14 3 3 4 Raiwaqa 21. 6 2.6 2 . 6 2.7 4.2 b Nausori 32 .4 15 . l 7 . 9 n.a. 16 .3 b 9.5 6.9 Korovou 6 . 9 n.a . n.a. n.a. 5 . 8 12 .9 Vaileka 16. 4 25. 4 4 . 5 4 . 8 34 . 0 6 . 8 Tavua 20. 4 20 . 3 8 . 1 7 . 2 27.4 7 . 9 Ba 15 . 8 13 . 5 6 . 3 6 . 8 17 . 7 9 .7 Lautoka 11. 9 30. 1 10 . 9 10 . 6 50 . 7 14 . 3 Nadi 13 ,5 9,7 8 ,9 8 . 9 18 . 7 9 , 3 4 . 8 Sigatoka 6,5 17 . 5 10 ,0 9 . 7 18 ,5 11 ,3 Navua 11 . 3 17 ,2 2 . 3 4 . 7 17 . 5 9 . 0 Levuka 7 ,1 16 ,0 7 . 1 Waiyevo 11 ,4 19 , 8 27,7 2 0 . 3 Savusavu 11 . 9 n.a. 8.5 1 . 6 23 . 5 3 . 2 Labasa 9 ,2 18 . 8 9 . 0 8 . 7 2 3 . 2 10 . 8 6 . 4

a ' - ' indicates category inappropriate. ' b ' n.a. indicates category appropriate but data unavailable.

Source : Market surveys , January 1976. 73

Tab le 2.7 a Means of vendor transEort to market

b Private Foot Bus Carrier c Boat Other vehicle d All markets 6. 7 59 .7 16.8 6. 7 10.1 n.a. e Suva 11. 6 73. 3 11 . 6 2.0 1. 7

Raiwaqa 5.6 66. 7 27.8

Nausori 3. 0 70. 8 9.0 12 .0 1. 9 2.3

Korovou 6.7 84 .4 6. 7

Vaileka 1. 8 80 .7 14 .0 3.5

Tavua 5.6 33. 9 14 .5 2.4 30 .6 12 . 9

Ba 2.3 53.2 26.6 14.5 1. 2 1. 8 Lautoka 7.2 55 .9 19 .8 2.9 2.4 9.4

Nadi 4.1 78.4 5.3 11. 1 1.2

Sigatoka 5.6 66. 7 16.7 5.6 4.2

Navua 2.6 59 .0 32 .5 5.1 0.9

Levuka 73. 3 15.6 4.4 2.2 4.4

Waiyevo 12 .5 25.0 4.2 54 .2 4.2

Savusavu 16. 3 18.6 65 .1 Lab as a 13. 8 23.0 32 .2 2.9 21. 3 6.9

Producer vendo rs 2.8 58.2 20.2 3.2 11. 7 6.6

Non-producer vendors 12 .3 70.5 7.2 8. 8 0. 5 0.7

a Figures are percentages of vendo rs ' declared means of travel . b ' Carrier ' includes hired taxis .

c ' Private vehicle ' is non-commercial transport (usually in vehicle belonging to vendor ' s family) . d ' n.a. ' indicates not available. e ' - ' indicates no transport recorded in the category .

Source : Market surveys , January 1976. 74 other means of transport whilst feet are an alternative for ven do rs who are primarily permanent . Buses are important because bus systems are well developed and producers are often ab le to transport the ir pro duce without charge in the outside compartments. Produce brought on to carriers by paying passengers is charged a nominal rate on the basis of number of bundles etc. Where several vendors from one area are going to market at the same time, especially if they have a considerable amount of produce and do not live on a direct bus route to the market, they are likely to use a carrier. Apart from their flexible routes and schedules an d the amol.lllt of produce that can be carried, carriers are also preferred because vendors are ab le to go to market on credit , the transport er collect ing the money either in the market after a few sales have been made or on the return journey .

The cost of transport of produce is not great . Almost three-quarters (71 per cent) of vendors who bring produce to market do not pay transport charges in excess of personal fares . Of vendors who pay cartage fees , for 95 per cent the charge is less than $ 1 and for almost two-thirds it is lOc or less. In addition, about one-quarter of vendors are non-producers who buy at the market goods for resale and so do not pay directly for transport.

Permanent vendors (see also Ap pendix 1) . Permanent daily sellers purchase stocks at the market from producers or less connnonly from non-producer wholesale-assemb lers; imported produce is bought from wholesalers in the market town . Local pro duce is often purchased from regular sources, often relatives, rather than a variety of producers and on the basis of regular 'orders'. Some vendors who sell two or three times a week also purchase food at the market for resale , but like irregular vendors they produce much of wha t they sell.

There are two basic types of market vendors : the per­ manent, non-producers ('middlemen ') and the non-permanent producers who sell their own produce. These two types of vendor have much in common . Selling me thods are similar: transaction s are in cash , bargaining is uncommon as prices are generally not negotiable , and vendors use passive sales techniques. In other respects, however, the operations of permanent and non-permanent vendors differ markedly .

Permanent vendors are distinguished from other vendors by a range of characteristics , an obvious one being race : 75 almost all are Indian or Chinese, although there is a small and apparently in creasing number of Fij ian permanent vendors in the larger markets. Permanent vendors are more likely than other vendors to be male , married and heads of house­ hold. Mo re of ten they have non-agricultural backgrounds than do non-permanent vendors and have had mo re formal education . Although the families of permanent vendors are mo re dependen t on market sales than are those of non-permanent vendors , only one-fifth of permanent vendors do not have an alternative income source . In contrast , 37 per cent of weekly vendors and 27 per cent of irregular vendors are without alternative sources of income .

Permanent vendors live closer than other vendors to the market at which they sell , have shorter travel times and consequently spend less on fares : 89 per cent of permanent vendors live within twenty minutes of the market compared to 50 per cent of other vendors. Thirteen per cent walk to market. Contrary to expectations, however, permanent vendors sell mo re frequently than non-permanent vendors at other markets. Permanent vendors also , mo re frequently , have selling assistance (26 per cent compared to 13 per cent of non-permanent vendors) and are more likely to sell other than tmprepared foods and a greater variety of products than other vendors , even in cases where the permanent vendors show marked product specializat ion . Most producer vendors sell no more than perhaps three differen t items wh ilst permanent vendors frequently offer ten or mo re .

The operation of permanent vendors , particularly those dealing in yaqona, rice, potatoes, on ions and garlic - a connnon combination of goods - is in some ways similar to that of small shopkeepers . They purchase supplies at the market and occupy two or three contiguous stalls that are rented on a monthly basis. Often family members help in the market . Many such vendors have been selling for five years or longer from the same stalls . It is not unusual for a market stall to be but one collIDlercial interest of a permanent ven dor : he may also operate a general retail store, import groceries and mixed goods , or assemb le and forward produce orders.

Some permanent vendors have telephones in their stalls , and in a few markets some operate fully-fledged importing and exporting businesses and even hold government supply contracts. 76

Stalls are well-stocked and vendors may have addit ional stock stored at their homes. Most permanent vendors sell produce by we ight , each having a set of scales in his stall. Prices of goods are marked, and rice an d yaqona are pre-packed, usually in paper bags stamped with the vendor's name , in popular purchase sizes . Permanent vendors dealing in produce other than yaqona, potatoes, rice , and associated goods rarely operate on such a scale or have as many commercial side­ interests. Nevertheless , their operations differ more in scale than form from the larger-scale vendors .

Non-permanent vendors. Most non-permanent producer vendors are farmers , so 'non-permanent ' and 'producer' vendors are used synonymously. Most live close to the market at wh ich they sell. Some come to market more or less regu­ larly once or twice a week , usually on Friday and Saturday , but many sell irregularly and less frequently . One-fifth of vendors describe their selling frequency as 'irregular'; the remainder, after discounting permanent vendors , sell one to three times a week, usually at weekends. The frequency of market visits by non-permanent vendors depends largely on their having goods to sell, another reason to visit the market town or a need for money an d no pressing commitments at home . Most sell irregularly, bringing their small stocks on buses or carriers. They sell the three or four different products bro ught on one trip in tmweighed tmits for prices that are multiples of five cent s. Prices are rarely marked but , as with permanent vendors, goods are usually wrapped in newspaper upon sale .

Non-permanent producer vendors rarely occupy more than one selling space and mo st stay in the market for only a day at a time . At Ba and Labasa, for instance , the average market stay of producers durin g one week in Novemb er 1976 was 1. 3 days; for non-producers it was 4.8 (Ba) and 5.4 (Labasa) days . As they grow the products they sell , producer vendors can participate in marketing with minimal investmen t needing only sufficient to pay fares and market fees: non­ producer vendors do not require much working capital , but they do need to purchase stock and pay fees up to one month in advance .

Non-perman ent vendors usually plan to stay in the market for one or two days at a time , depending on the amotmt of produce they have for sale, the distance travelled, other things to do in town and availab ility of return transport . 77

All try to complete trading by mi d-Saturday when markets . close as few wan t to remain to sell surp lus produce the followin g Monday .

A number of strategies are followed in dealing with tmsold produce . If it is highly perishable - like leaf ve getables and ripe tomatoes - prices are reduced to get rid of the produce. If the unsold produce is not highly perish­ able , prices are mo re stable. Most local produce , however, has a very limited post-harvest life. Producer vendors know from past experience the amoun t of produce they can expect to move in one day , so that few miscalculate greatly enough to necessitate either ext ra days at mar�et or sales at prices much below those expect ed . Un sold food is rarely taken home . It is us ually given to friends or relatives who live in town , probably the ones with whom the vendor spends the night wh ile marketing. Very few vendors actually throw away unspoilt food they cannot sell.

Permanent vendors , generally , are not un der the same pressures to move produce as non-permanent vendors . They are not in a hurry to sell stock to cat ch tran sport home and many deal in less perishable items like potat oes an d yaqona. All vendors who stock highly perishable it ems like leaf vege­ tables an d tomatoes are faced with some urgen cy in moving stock and towards the end of a day's trading there can be sharp price reductions.

Weekly market patterns. The character of market s changes noticeably in the course of a week in response to the involve­ ment of producer and non-producer vendors . There are at least four di stinct trends. One is that in terms of share of trade, permanent vendors become pro gressively less important during the week. At Ba and Labasa markets in November 19 76, for example, 64 and 82 per cent of vendors from Monday to Thursday were permanent, compared to 52 per cent on Friday and 35 and 39 per cent on Saturday . Fij ian vendors become more numerous as the week progresses , although their share of the total number of vendors may not increase as much as the number of small, independen t Indian farmers . A third trend is that market hinterlands - as measured by the distance travelled by vendors - expand during the week. This trend is mo st marked in the Western Division but occurs in all market s. Finally , me asured by either the number of vendors or the amotmt of produce sold , Friday and Saturday trading domin ates the markets. Moreover, as many products 78 are handled by particular races , the relat ive av ailability of foods changes over the week .

Pro ducer vendors can be from one of three basic mo des of product ion which are distinguishable in terms of source an d form of employment . Th ese are village , small independent , or wage-labour employing. The village farmin g system, based on un paid family and community labour an d using commonly­ owned village land, is confined to Fij ians and is largely subsistence . Small independent farmers work either leasehold, which in cludes Fij ians on formally-leased mataqali land, or freehold land with the help of f arnily labour and employed hands as required seasonally . Apart from some Chinese ginger and dale farmers in Naitasiri, farming based on wage-labour employment is uncommon in vegetable production for the local market but does represent the mo st 'commercial ' form of pro­ duction . As it has been suggested that the mode of production affects parti cipation in marketing (see The Agricultural System in Chapter 1) , it is interesting to examine producer­ vendors from this perspect ive (Tab les 2.8 and 2.9).

At Ba and Labasa market s, the only markets wh ere producer vendors were asked about their mo de of production , almost two-thirds of Fij ian producer ven dors are villagers . Mo st Fij ians who are independent farmers are leasing mataqali land, although at Ba some have leasehold land on government-organized land development schemes . Indian producer vendors are almost all small in dependent farmers on leased land. Large-scale farmers employing wage labour were not en countered selling in the market s. From acquaintance with Suva market , however, it appears that such farmers wholesale their produce to market vendors rather than retail it themselves.

If villagers were primarily end-of-week , irregular vendors , they would be mo re active during Friday and Saturday trading. This was so at Ba, where 29 per cent of Fij ian vendors present from Monday to Thursday were villagers , com­ pared to 84 per cent on Friday and Saturday . At Labasa, however, the proportion over the two periods was about equal (62 and 65 per cent) . The situat ion at Labasa is probably accounted for by the fact that the lon ger distances travelled by village producers in the Labasa area en courage mid-week selling.

Returns to vendors

Returns to vendors are in fluenced by such a variety of Tab le 2.8 a Produc tion systems of producer-vendors at Ba and Labasa markets

Independent farmers Villagers Fij ian on Lessee on Other mataqali develop . lease­ Freeholder Other leaseb scheme holder

------· ---· -·------. Fij ians: Ba (106) 61 . 3 10 .4 26 .4 2.8 Lab as a (65) 63 .1 24 .6 9.2 3.1

Indians: Ba (156) 83 .3 14 .1 2.6 Labasa (139 ) 89 .9 4.3 5.8 a Figures in brackets are percentages of valid cases . b Villagers farm mataqali. land but un like independent farmers do no t have access to land through individual , forma l lease arrangements. A producer status on the questionnaire was 'farmer with wage-labour ', there were no responses in this category al though many farmers , especially other than vi llagers , periodically employ short-term labour . Source: Market surveys , Ba and Labasa , November 1976 . Table 2.9

Product ion system of Fij ian producer-vendors at Ba and Labasa markets

Independent farmers Villagers Leased Lessee on Other mataqali develop . lease­ Freehold Other land scheme hold

Monday-Thursday Ba (38) 28.9 7.9 52.6 7.9 2.6 Labasa (34) 61 . 7 29 .4 8.8

Friday-Saturday Ba (64) 84.4 12.5 12 .5 Labasa (31) 64.5 19 .4 9.7 6.5

Note: Figures in brackets are percentages of valid cases.

Source : Market surveys, Ba and Labasa, Novemb er 1976 . 81

factors that estimat es of trading performan ce are re adily open to contradiction . Nonetheless, by delineat ing the chief variables affecting returns an d by drawing information from two studies conducted at Suva market , it is possible to get a general impression of returns .

Producer vendors. The -main direct costs faced by pro­ ducer-vendors are transport to and from market for themselves and their produce , market fees and food costs wh ile at market . Accommodation is rarely a significant cost because if vendors s tay at the market overnight , they sleep either on foo�paths around the market or wi th relatives .

To gain in formation on the form and profitability of Fij ian producer involvemen t in marketing, a study was con­ ducted at Suva market. Attention was focused on the effects of res idence distance on frequency of participat ion and market ing ret urns. 22 From each of the three study areas , records were kept of three to seven women vendors who visited the market for Friday and Saturday trading more or less regularly over two months . The study areas are briefly described in Table 2.10.

Ap art from two male small independent farmers (gala la) ,, the vendors were married women . All regarded the market as a good source of money and indeed they , or their husbands, had no other equally reliable or regular source of income . Ap art from the need for money and availability of produce to sell , a maj or factor in fluencing involvement in marketing was the presence of grown-up children , especially daughters, who could look after the family while the mo ther was at the market . This was particularly impo rtant in mo re distant areas where the seller could be away for three or four days on one trip. Women who were unable to visit the market becaus e of household responsib ilities , frequently gave small amounts of produce to other women who would sell it and return all the in come to the owner.

An important motive for market participat ion by the surveyed producer vendors was that basic foodstuffs and fuel supplies were cheaper in the market towns than in their village stores. For some vendors , it appeared that marketing was a means to the chief obj ective : the low cost purchasing of necessities.

Vendors have a reasonable idea of wh at they will receive 00 N Table 2 .10

Participation and in come of Fij ian producer-vendors at Suva marke t (Over nine consecutive weeks in Sep temb er and October 1976)

Average Average Costs as gross income Average % Vendors numb er of gross Area net income / surveyed vi sits to income/ visit Transport All costs market vi it f $ a Nakorotubu 7 3.0 40.9 29 .0 19.6 29 .0 b Sawakasa 6 4.8 38.3 23.3 8.2 17.4 c Namara 4 6.0 17.5 11.8 17.o 36 .7 a Vendor villages in Nakorotubu are 88 to 96 km from Suva market, requiring 4 to 6 hours of boat and truck (67 km) transport. Boat and truck fares , both ways, are about $9 per person . b Sawakasa vendors live on a main road about 54 km north of Suva . They travel by either bus or carrier; the return fare is between $2.20 and $6 depending on the amount of produce carried . Average income is relatively high as 2 of 6 vendo rs are active ga lala (independent farmers) rather than villagers . c The four villages in Namara whence the vendors come are about 40 km from Suva , reached by a $1 bus ride or for $2 in a carrier .

Source : Baines 1976 . 83

for the ir products an d they take to market only enough to gain the mon ey they require. Producer-vendors are aware of the consequences of over-supply on prices and are careful not to brin g too mu ch of a particular item. Expected returns are further insured by bringing ab out four different products. As an example, product s brought on one trip by a typical vendor from Nakorotubu, the study area most distant from Suva, comprised thirteen dozen coconuts, forty pawpaws , about 30lb of chillies , about 8lb each of tumeric an d wild ginger an d four l5-20lb baskets of cassava. The vendors themselves sell all the produce they bring to market although they are frequently approached by permanent vendors seeking goods for resale.

Th ere are interesting differences in returns to vendors ' between study areas (Table 2.10) . Vendors from Nakorotubu, who require at least three full days to visit Suva market , come the least frequently but make the highest gross an d net returns per trip. The vendors living nearest to Suva have the lowest returns and the highest share of the gro ss returns accounted for by costs.

If net returns are related to frequency of vi sit , the differences in in come between area are not as marked as in fo rmation on single visits suggests. Calculated to an annual rate, net returns for vendors from the three areas are $409 , $503 and $648. The highest figure is for the vendor sample that in cludes two active galala farmers : villager producer-vendors in the same area make substantially less . If market returns are adjusted for frequency and cost of visit , it appears that rural Fij ian villagers have particular and similar requirements fo r cash that are largely in dependent of residential locat ion and that market involvement is adj usted to meet these needs.

Non-producer vendors. Market sales are as important a source of in come for non-producer vendors as for producer vendors , but forms of participat ion an d returns differ signi­ ficantly. Non-producer vendors purchase goods at the market from farme rs or assemb ler-wholesalers, many of whom are farmers themselves. Goods are purchased by either weight , for example, French beans, or standard units such as sacks of eggplant , cases of tomatoes, bundles of Chinese cabbage , an d so on . Produce is purchased as required , but vendors tend to purchase fruit and vegetables on three- to four-day cycles an d attempt to sell mo st by market closing on Sat urday . 84

Table 2 .11

Markups from purchase to retail price by non-producer-vendors at Suva market

Percentage mark-up Numb er Product cases mean standard devi·ation-�-

Eggplant 231. 4 94. 6 12 Tomatoes 135.4 83.0 9 Pawpaw 14 7. 2 84.3 3 Rourou 120. 1 82 . 3 8 Coconut 109 .2 37. 8 3 Cucumber 102 .6 67.0 2 Chillies 97.2 64.1 6 Dhania 87.2 68. 7 9 Mango 79.6 123. 3 2 Butter beans 75.0 35.4 2 Cassava 74.5 5.7 2 French beans 73. 6 44.8 6 English cabbage 72 . 8 69.8 44 Radish 59 .3 36. 7 2 Bora beans 52 . 8 45.8 5 Amaranth 50. 0 o.o 2 Ke re la 45.9 29.5 2 Chinese cabbage 21. 8 42 .6 7 Long beans 12 .5 17.6 2 Rice 9.2 4.4 3

Source : Surveys, Suva market , 7-11 Septemb er 1976. 85

Popular opinion holds that non-producer vendors set excessive markups and generally operate in a non-competitive fashion. An attempt was made to monitor non-producer trade . It was found that at least amon g permanent vendors selling fruit and vegetables in Suva market there is marked competition amon g vendors , with prices reflecting supply and demand con­ ditions , and that whilst some vendors make relat ively good incomes , many do not. Like the producer-vendors , permanent vendors have set opinions on wh at the price of any item should be. They have the purchase price as a starting point in setting prices . However, on the whole they are realis­ tically flexib le in adj usting prices to meet supply an d demand relativities.

Data collected in Suva market in September 1976 revealed that price markups on twenty products handled by non-producer­ vendors ran ged from over 200 per cent to under 10 per cent , the average being 82 .9 per cent (Table 2.11) . 23 Net income is considerab ly less than these markups. One vendor, for instance, made only 30 per cent over expenditure , excluding market fees, from six days ' trading with gross daily turnovers from $6 to $7 on Tues day and Wednesday to $20 to $22 on Friday and Sat urday (Table 2.12) .

Daily net in come of twenty of the surveyed vendors aver­ ages $13.1, which implies giving a six-day weekly rate of $78.6 {Table 2.13) . There are marked differences in trad ing profitability by race of vendor . The average net daily in come of Indian vendors ($16.1) is over double that of Fij ian vendors ($7.6) . At an annual rate, average incomes are $4087 for all vendors , $502 3 for Indians and $2371 for Fij ians, from which an average of $300 should be deducted to cover stall fees .

Income extrapolation s for market vendors should be treated with caut ion because of the range of possible in flu­ encing variab les such as income wh ich varies greatly with season , size of stall and location in market . Nevertheless they do suggest the general range an d racial differences in market income for permanent vendors. Such annual incomes , particularly as they are generally untaxed, are high for semi-skilled employment in Fiji. Calculated on an hourly basis , returns are not so attractive . Assuming vendors work about 60 hours per week and that each vendor works alone, net hourly in come of Indian permanent vendors is 27 cents and of Fij ian permanent vendors is 13 cents. Differences between Indian and Fij ian incomes are partly due to Indians 86

Tab le 2.12

Trading activity of a Fij ian and an Indian non-producer­ vendor at Suva market , 7-11 Septemb er 1976

Fij ian vendor

Purchases Sales Un sold Net income Product Un it quantity cost ($) ($) quantity ($)

Watermelon single 3 3.00 6.00 3.00 Mango heap 110 23.16 27.57 66 4.41 Eggplant heap 90 7. 38 16 . 60 13 9.22 Chinese cabbage bundle 24 5.00 3.60 -1.40 Cucumber single 103 3. 32 8. 30 20 4.98 Rourou blID.dle 24 3. 00 9.60 6. 60 Banana heap n.a. 5. 00 10. 00 5.00 Tomatoes 'tin ' 3 9.00 14 .40 5.40 Chillies heap 12 1.00 1.20 0.20

Total 59 .86 97 .27 37.41

Indian vendor

Purchases Sales Un sold Net income Product Un it quantity co st ($) ($) quantity ($)

Eggplant heap 20 0. 84 2.80 6 1.96 Butter bean lb 22 1.10 2.22 1.12 Amaran th bundle 12 0.60 1. 80 1.20 Tomato lb 125 10 . 00 18 .75 8. 75 Okra lb 16 2.20 3.85 5 1.65 En glish cabbage lb 90 6.91 12 .00 15 5.09 Dhan ia bundle 24 1.06 1. 60 8 0.54 Chinese cabbage bundle 30 4.75 4.75 Ginger lb 5 0. 30 0.60 0. 30 Rice lb 300 36 .66 40. 00 50 3.34 Ch illies lb 10 1.25 3.00 5 1. 75 Be le bundle 14 n.a. 1. 75 (?) 1. 75 Coconut single 3 0.12 0. 30 0.18 French bean lb 20 0.80 1.10 10 0. 30 Onion lb so 2.75 4.50 25 1. 75 Irish potato lb 50 2.50 3. 75 25 1.25 Sultanas lb 1 0. 60 o. 70 0.10

Total 72 .44 103.47 31.03

Note : Of the food un sold at close-of-trading on Saturday , only mangoes , rice , potatoes , onions, English cabbage an d some cucumber were kept to sell during the following week.

Source : Surveys, Suva market , 7-11 September 1976. Table 2.13

Ne t profits of non-producer-vendors at Suva marke t

Net daily profit ($) Weekly net Vendors Cases pro fit ($) from standard mean maximum minimum daily rate deviation All 20 13.1 12 .7 42 .0 1.6 78.6

Indian 13 16 .1 14 .2 42 .0 3.1 96 . 6

Fij ian 7 7.6 7.0 19 .l 1.6 45 .6

No te : Data are for consecutive days in the same week - in ten cases for 5 days , six cases for 4 days and four cases for 3 days . Profits are net of purchase cost only; other chief fixed costs are stall fees of $1 to $2 per day depending on amount of selling space occupied .

Source : Surveys , Suva market, 7-11 Septemb er 1976 . 88

having better stall locat ions . Also important is the fact that each deals in differen t items , due as much to the In dian vendors ' better access to supply and their ability to purchase goods on credit from known suppliers , as to the conservatism of Fij ian vendors who deal mainly in locally-grown Fij ian staples .

Produce sold

The distinct seasonality of most local foods is reflected in market activity. In the hotter and wetter January-July period practically no locally-grown temperate climate vege­ tables are available. Eggplant, okra, ota androuro u are the main green vegetables in supply throughout the year. In the off season , the amount of trade in seasonal vegetables by permanent vendors decreases greatly as the vendors do not swit ch broad product lines: those dealing in vegetables and fruit sell what is in season and do not change to non-seasonal yaqona, rice, potatoes, etc. , or even to native vegetables, during the vegetable off season. The number of producer­ vendors also declines in the off season , particularly among Indian vendors who are more likely than Fij ians to grow and sell temperate vegetables .

Notwithstanding seasonality of supply, some general features of produce trade can be delineated. 24 The main groups of food sold in all markets are nat ive staples , native vegetables, Indian and introduced vegetables and fruit , excludin g imported fruit (Table 2.14) . 25 These foods account for ab out 90 per cent of produce in al l markets. The balance consists of Indian staples - such as Irish potatoes, onions, garlic, rice , pulses and dry spices - imported fruit , yaqona and tobacco and oils . Indian staples, imported fruit , yaqona and tobacco were not recorded at Levuka, Waiyevo or Savusavu markets but were present in all other market s.

There is considerable range in the relative importance of different foodstuffs between markets, reflecting the culture of the consumers and the production of the hinterland. Native staples and native vegetables, for in stance , are poorly represented in the Western visDi ion , particuarly at Ba , Laut oka , Nadi and Sigatoka markets wh ere they account for less than 10 per cent of produce. In markets out side cane­ producing areas , an d especially in the smaller towns where the purchasing population is small and not as culturally diverse as elsewhere, native produce can account for more than two-thirds of foodstuffs for sale. Table 2 .14

a Composit ion of foodstuffs traded b}': market

Native Native Indian Indian In troduced Imported Yaqona, % surveyed Fruit Oils Total staples vegetables staples ve getables vegetables fruit tobacco food at mkt

b Suva 10.8 0.1 0. 5 29 .2 16.1 37.3 1.9 3. 6 0.1 100.0 53.2 c Nausori 13.6 1. 7 1. 7 37.9 14 .1 29 .0 1. 6 0.2 100.0 5.3 Vaileka 17.8 4.1 1.2 34 .5 13.1 19 .8 0.6 8.6 0.2 100.0 1. 7 Tavua 15.9 0. 8 13.6 8.1 5.8 48.9 0.2 6.4 0.3 100.0 3.6 Ba 8.6 0.4 5. 1 41 .2 14 .2 27.5 0.1 2.9 0.1 100.0 4.5 Lautoka 7.4 1.2 4.4 25.0 25.0 33 .9 0.2 2.8 0.1 100 .0 9.6 Na di 6.9 0.4 4.3 17.6 34 .1 34 .6 0.1 2.0 0.2 100.0 4.0 Sigatoka 3.5 0.3 2.0 22.7 20.0 43.0 0. 3 8.3 0.1 100.0 5.1 Navua 12 .7 1.0 0.1 29.5 3.8 52 .6 0.2 100.0 4.5 Levuka 86.6 2.8 1.0 3.0 6.6 100.0 0.1 Waiyevo 65.6 0. 3 9.5 1.6 21.9 1.1 100.0 0.2 Savusavu 29.3 2.3 13.3 54 .5 0.6 100.0 0.4 Lab asa 5.1 0. 3 4.6 37.1 6.9 29 .6 10.7 5.5 0.1 100.0 7.7

All markets 11.4 1.1 2.5 17.3 18.6 42 .5 2.2 4.3 0.1 100.0 100.0 a Inadequate data are available for Korovou and Raiwaqa markets. b All food values were converted to a standard un it (lbs) from estimates of the mean size of the selling un its in which foodstuff quantities were recorded. Informat ion was also ob tained on marine products (molluscs , crustaceans, beche-de-mer, seaweeds , dried and smoked fish and turtles), but as the amoun ts were relat ively small and because of difficulty in converting them to standard weights, the informat ion is not in cluded. c ' -' indicates no record of produce in the relevant category .

Source : Market surveys, January 1976. 90

Around one-half of foodstuffs traded are sold on Friday an d Saturday (Table 2 . 15) . Exceptions to this sales pattern occur only wh ere weekday sales , especially Monday to Thursday sales , are minimal , as at Levuka , Waiyevo and Korovou markets. At all other markets, excluding Raiwaqa because of inadequate data, ab out one-half of the week 's trade occurs on Friday and Sat urday . Variat ion in the availability of items over the week is also caused by differen tial handling of produce according to type - producer, non-producer - and race of vendor. Native staples and native vegetables, for in stance, constitute a larger share of market produce on Fridays and Saturdays than at other times .

Data on produce sold and the vendor's race reveal two important features of market operat ion . One is that the share of trade handled by vendors of one race may differ from that race's representation among vendors (Tables 2.16, 2.17) . Wh ilst 43 .7 per cent of vendors are Fij ian or other Pacific Islanders , for example , they trade only 21.2 per cent of total market produce; Indians account for 54 .8 per cent of vendors and ndleha almost three-quarters of all produce; an d prac­ tically all the balance is with the 1.1 per cent of vendors who are Ch inese. Share of trade by race varies by market, but overall Fij ian vendors handle considerably less than their numbers would suggest.

Second, each race specializes to some extent in the type of product handled (Table 2.18) . Between markets there are differences in racial involvement in the trade of individual items , but at the products group level , broad patterns of specializat ion remain .

Fij ian vendors sell mainly nat ive staples (56 per cent of Fij ian trade) , fruit (11.8 per cent), animal products (11.9 per cent ), native vegetables (6.2 per cent) and yaqona and tobacco (6.2 per cent) . Trade in other foods such as Indian staples and vegetables, introduced vegetables, imported fruit and oils , comprises less than one-tenth of total Fij ian trade , but accounts for almost one-half of Indian market trade . Chin ese ven do rs sell mainly Indian staples , introduced vegetables and imported fruit ; together these products make up over 80 per cent of Chinese trade .

Trade in nat ive staples an d vegetables an� to a lesser extent, fruit and animal products is dominated by Fij ians . Indian vendors handle a large share of the trade of Indian staples and vegetables, yaqona and tobacco and oils. Chinese Tab le 2 .15 ------a Food trade variation by day in selected markets

Suva Nausori Sigatoka Vaileka Savusavu Labasa

M-Th F-S M-Th F-S M-Th F-S M-Th F-S M-Th F-S M-Th F-S

All foods 57.6 42 .4 48.3 51. 7 59.5 40.5 56.4 43.6 42 .3 57.7 61.1 38. 9

Native staples 43.9 56.1 25.9 74.1 57.1 42 .9 40. 5 59 .5 39 . 6 60.4 43.5 56.5

Native ve getables 30.6 69 .4 13.6 86.4 11. 7 88. 3 18. 6 81. 4 25.0 75.0 20. 9 79.1 b c Indian staples 74. 4 25.6 82 . 3 17.7 98.0 2.0 40.0 60.0 n.a. n.a.

Indian vegetables 54.9 45 .1 56.6 43. 4 61 . 5 38. 5 66.1 33.9 72 .5 27 .5 54 .8 45.2

In troduced vegetables 55 .1 44.9 47.3 52 .7 58.0 42 .0 58.4 41.6 33. 3 66.7 68. 5 31. 5

Fruit 60. 3 39 .7 50.0 50.0 60.0 40.0 56 .4 43.6 37 .0 63.0 62 .7 37. 3

Imported fruit 68. 0 32 .0 n.a. n.a. 95.9 4. 1 81 . 8 18.2 62 .o 38.0

Yaqona, tobacco 67.4 32 .6 42 .0 58.0 57.0 43.0 65.2 34 .8 75.7 24.3

Oils 75.0 25.0 49 .3 50.7 n.a. n.a. 66.2 33 .8 55.0 45.0 65 .9 34 .1

a Figures are percentages of quantity of food category recorded over week that were present on specified days. b ' - ' indicates no product recorded for the particular period. c ' n.a. ' indicates product sold but further in formation is not available.

Source : Market surveys, January 1976. 92

Table 2.16

a Share of trade in main foodstuff cat�_i!Il_d�� rke!_ ·  Native staples Indian staples_ Indian vegs In tro . vegs b c d F I c F c F c F c F I c

e All marke ts 62 .8 36 .2 0. 8 68. 6 30 .9 0.4 1.1 94. 5 4.0 8.1 91 .5 0.2 10 .0 86.0 1.8

Suva 63.8 34 .8 1.1 71 .2 29 .9 0.6 1.1 91.4 6. 8 9.8 89 .6 0.4 12 .l 82 .2 3.1

Nausori 46. 6 53.2 0.2 83. 0 16 .5 0.5 _f 100 8.8 91.2 9.2 89.1 1.7

Vaileka 76.4 24.6 75 .0 25.0 100 0. 3 99.7 1. 3 98. 7

Tavua 53. 7 46.3 74 .4 25.6 0.2 99 .8 34 .0 64 .5 17 . 7 82 . 3

Ba 40.9 59.1 14 .4 85 .6 1.5 98. 5 0.2 99 .8 0.2 99.8

Lautoka 56. 3 43. 7 44.1 55 .9 0.9 99 .1 7.3 92 .7 6.6 93.4

Nadi 35 .9 62 .6 1.5 66 .7 33. 3 100 4.0 95.9 0.1 11 .6 88 .1 0.3

Sigatoka 88.1 11 . 9 71.9 28. 1 1.0 99.0 7.0 93.0 7.7 92 .3

Navua 76 .6 23.4 87 .4 12 .6 100 16.0 84.0 23.6 76.4 g Waiyevo 90.0 10 .0 100 n.f.s. 72 .1 27.9 50 .0 50.0

Savusavu 96 . 8 3.2 83. 7 16 .3 n. f. s. 24.4 75 .6 100

Lab as a 76.2 21. 5 2.3 57 .7 39 .7 99 .4 0. 6 0. 7 99 .3 0.2 99.8

Imported fruit Yagona, tobacco All food

F c F c F c F c F c

All markets 23. 9 74 .6 1. 3 0. 6 81 .2 18. 2 4.2 95.6 1.6 98.4 21.2 73. 7 4.6

Suva 26.4 71 .5 2.0 1.1 77. 7 21.2 4.6 95.4 2.4 97.4 22 .7 55.1 21.1

Nausori 39 .0 61 .0 100 100 7.0 93. 0 29 .3 70.6 0.1

Vaileka 24.4 75 .6 100 35 .4 64.6 100 46. 3 53. 7

Tavua 42 .3 57. 7 100 1. 7 98. 3 1.0 99.0 14 .7 85.3

Ba 13 .5 86.5 100 100 100 10 .3 89.7

Lautoka 21. 3 76. 3 2.4 100 100 100 17 .3 82 .7 0.1

Nadi 20.0 79 .7 0. 3 89 .5 10.5 2.0 98.0 100 8.4 91.1 0.5

Sigatoka 6. 7 93. 3 100 6.2 93. 8 5.3 94 .7 16.9 83. 1

Navua 21.4 78.6 n. f. s. 76.2 23.8 n.f. s. 57.6 42 .3

Waiyevo 88 . 7 11 . 3 100 n. f. s. n. f. s. 85 .3 14.7

Savusavu 78 .5 21.5 n. f. s. n. f. s. 28.0 72 .0 85 .6 14 .4

Lab as a 1.9 98.1 84 .1 15.9 100 100 13.9 85 .4 0.6

a Korovou and Raiwaqa markets are excluded because of incomplete data; Levuka market is excluded because data were recorded for Fij ian vendors only . b 'F' stands for Fij ian and other Pacific Islanders . c 'I' stands for Indian . d 'C' stands for Chinese vendo rs . group . Where percentages e Figures are percentages of total quantity recorded of each food gely part-Europeans) . do not sum to 100 , the balance is comprised of 'other' vendors (lar f '-' means not recorded for the particul ar racial group . g 'n. f.s. ' indicates item was not recorded .

Source : Market surveys , January 1976. 93

Table 2.17 a Trade in main produce lines by race (all markets) Fi hin- Fij i hin j �- Indian C Other - Indian C -Other an ese an ese Native staples Fruit dalo 70.1 29.1 0.5 pawpaw 45 .8 48.2 6.0 dalo-ni-tana 96.6 3.4 banana 40.9 57.8 1. 3 cassava 95.0 5.0 lemon 27.9 68.9 2.8 0.4 sweet potato 100.0 orange 34 .1 41. 9 7.8 yams 87.8 12 .2 ivi 92 .9 7.1 breadfruit 65 .5 34 . 5 watermelon 2.9 97. 1 plantain 85.1 12 .6 2.3 pineapple 32 .4 64 .6 3.0 coconut 53.3 44.4 2.3 Imported fruit Native vegs apple 3.4 40.1 47. 5 ota 87.4 0. 7 11. 9 pear 0. 3 44 .5 55.2 rourou 70. 7 29 .1 0.2 grapes 10 .1 65 .7 24 .2 be le 82 . 8 17.0 0.2 orange 28. 7 71. 3 Indian staples Yaqona, rice 1. 3 89 .5 7.1 2.0 tobacco jackfruit 100.0 tobacco 1. 8 98.2 pumpkin 3. 9 95.5 0. 3 0. 3 yaqona 6. 7 93.3 Irish potato 1. 5 88.5 9.6 0.4 Oils onion 1.2 88 .2 10 .6 coconut oil 3.7 96. 3 garlic 0.4 88.9 10.0 0. 7 ghee 1.6 98.4 carrot 8. 5 57.8 33. 7 'oil' (tmspec.) 100.0 dhal 100.0 pulses 100.0 Animal spices (dry) 100.0 products fowls 23.5 76.5 Indian vegs fowl eggs 28. 7 66.9 4.4 okra 0.1 99 .9 mussels 100.0 eggplant 10. 1 89 .7 0.2 other shellf. 89 .9 8.1 2.0 ginger 9.1 83 .8 7.1 dried/ kerela 34 .2 65 .8 smoked fish 100.0 chillies 15.9 83. 8 0.3 prawns 100.0 beans (not Fr. ) 2.5 93.4 4.1 seaweed 100.0 gourd 72 .3 27.7 sea cucumb er 100. 0 amaranth 0.2 99 .8 landcrabs 100.0 coriander 100.0 100.0 tamarind 1.5 92 .5 6.0 other crabs Intro . vegs tomato 11. 7 62 . 3 26.0 Ch. cabbage 4.6 85 .0 2.8 7.5 lettuce 29.5 62 .7 7.8 French beans 4.6 85 .1 2.8 7.5 cucumber 7.7 83 .0 8.2 1.1 a Only main foods are included; in mo st cases this is taken as foods for which data are available fo r mo re than 1000 lbs . Numbers are percentage of each food handled by vendors of indicated races . b '-' in dicates item not recorded . Source : Market surveys , January 1976. \.0 � Table 2.18

Trade composit ion by race of vendor (all markets)

Fij ian/ Food group Indian Chinese Other Pacific Islander Native staples 56.0 7.1 3.1 7.7 Native vegetables 6.2 0.5 0.5 0.4 Indian staples 1.1 25.7 35 .1 25.9 Indian vegetables 3.4 12 .6 4.0 5.8 Introduced vegetables 2.4 5.1 24.6 57.4 Fruit 11.8 6.2 4.5 0.6 Imported fruit 0. 5 1. 8 24.7 a Yaqona, tobacco 6.2 33.2 Edible oil 0.3 3.4 1. 5 Animal products 11.9 4.0 3.5 0.7 Other and unknown 0.2 0.4 Total 100.0 100.0 100.0 100.0

a ' - ' indicates items not recorded. Source : Market surveys, January 1976. 95

ven dors do not dominate any particular food group but with Indian vendors account for mo st of the trade in in troduced vegetables and imported fruit .

However , wh ilst vendo rs of one race may dominate trade of a general pro duce category , resource access and differen ces in agricultural production by race cause considerab le vari­ at ion in the share of trade in in dividual items . For instance , while nat ive staples are largely in the hands of Fij ians, Indian vendors account for at least 29 per cent of the dalo and breadfruit an d nearly one-half of coconut s traded. Similarly , Indians handle almost one-third of rourou traded . Fij ian and Indian participat ion in fruit is roughly even . Their shares in pawpaw and oran ges are about equal , but Fij ians account for almost all ivi and Indians fo r lemon , watermelon and pineapple. The only animal products , fresh fish excluded, in wh ich Indians have anything mo re than a minimal share are fowls and fowl eggs.

Indian dominat ion of Indian staples and ve getables, imported fruit , yaqona and tobacco , an d oils is broken only in a few pro ducts. Carrots are the only food within the first catego ry whe re Indians do not account for at least 88 per cent of trade . Kerela and gourd are the only maj or Indian ve getables in which non-Indian vendors have a signi­ ficant trade . In the category of introduced vegetables, Chinese traders have a considerab le business in lettuce and tomatoes , an d they also handle much of the trade in imported fruit . Indian vendors virtually monopolize the trade in tobacco (98.2 per cent), yaqona (93.3 per cent) an d oil (96.3 to 100 per cent , depending on variety) .

Market supply

Hinterlands . There are three basic forms of market hinterland 1n Fij i. One is an exclusive hinterland where the market is the centre of a discrete area an d is the only significant co1Ill11ercial outlet for the produce of the hinterland. Levuka, Waiyevo , Savusavu and Labasa markets have exclus ive hinterlands . A second form is the hinterland wh ere produce sells primarily through the nearby outlet but some is also taken by producers to larger regional markets. Korovou, Navua, Vaileka, Sigatoka, Nadi and Ba markets have hinterlands of this nature. The third type is the hinterland which comprises the area contiguous to the market , off-shore islands an d smaller markets and their hinterlands . Suva an d Lautoka market s have this type of hinterland (Fig.13) . 96

Whether produce is sold at local or regional markets depends on a variety of factors , including the quantity available , expected demand at each market , which is related to the day of the week , and ease of access. Trading at all markets reaches a peak on Friday and Saturday and on these days hinterlands expand as vendors travel further to the larger central markets. Trade at the smaller local markets does not decrease on Friday and Saturday , but is proportion­ ately less of the total regional trade than that handled from Monday to Thursday.

In teraction with markets is most intense in the contiguous zone and beyond this in the area connected by road. Market participat ion declines rapidly. Th ere are areas on Viti Levu and Vanua Levu, as well as much of Lomaiviti and Lau , that , save for occasional shipments of high-value produce like yaqona, are beyond the range of the produce markets.

Even some areas with semi-regular market involvement have difficulty of access. Villagers from Ba for instance regularly spend seventy-two hours and even longer on trips to sell at Suva market s. The only off-shore islands with regular and constant market activity are Beqa and some of the Yasawas .

Participation of pro ducers in marketing is by no means uniform within market hinterlands. Apparent spat ial anomalies in participation are readily apparent (Fig. 14) . Moreover, the relationship developed in economic locat ion theory between production and distance from market centre is not easily identifiable in data on market-residence distance and crops produced (Tab le 2.19) . 26

Correlations between produce quantity and the residence­ market distance of the producer are not definitive (Tab le 2.20) . Most marked direct correlat ions between distance and quanti�y occur for yaqona and tobacco (0. 271) and native vegetables (0. 181) . Oils show a tendency to inverse correlation (-0.194) . Native staples, for wh ich a high inverse correlation could be expected, score 0.044. Correlations for some indi­ vidual items within the broad product categories vary markedly from the group figure. Among native staples, for instance , correlation for dalo is negative (-0.098) but is posit ive (0.149) for dry coconut s. Foods with a direct correlat ion of more than 0.2 are okra, Chinese cabbage , orange and guava . Curry leaf , lime, sugar cane , coconut oil and ghee, all have marked inverse correlations. D Native staples � Native vegetables - Indian staples § Indian vegetables . LO IVITI - Introduced vegetables �� Native fruits e D [>I Imported fruits j:::::::::::I Yaqona. Tobacco � Oils - - - - Province boundary

· ·· · ··· · ·· Tikina boundary numbers indicate per cent of market food from area CiTAILEVU 2.2

Bau 4.2

�lrnpocted Moala 0 EOA 0.1 � 73 �� 10 km 50 (_Y0.8

Figure 14 Origin of food at Suva market (Source : Market surveys) \() Table 2.19 ----�·---- 00 a Distance (km) of ven,dors ' residence from_ !_llarket for sel�_c t�_!oducts sold by pro�ucer__ y endors

tJ Ctl ;:I I'll •14 Ctl Ctl � •14 > .i.J H � � 0 Ctl Ctl Ctl (I) 0 (I) Ctl 0 .i.J c1j § I'll I'll ..!G Ctl I'll .-! ;:I .i.J ·14 c1j ;:I (I) ;:I Ctl .-! H > ;:I •14 > ;:I "O 00 > �> > > ,..Cl ;:I Ctl Ctl Ctl Ctl c1j Ctl •14 Ctl (I) Ctl Ctl Ctl U) � � E-l r:Q ....:I g; U) z ....:I E-l U) ....:I �� b dalo 26.9 23.2 19 .5 35.3 65.5 17 .7 82 .8 19 .5 22 .5 19 .6 25.1 21. 9 26.4 cassava 23. 7 16.3 15 .1 29.9 29.0 15 .8 52 .0 17 .5 16 .7 12 .2 15.0 breadfruit 16.6 11. 6 12 .1 21. 4 14.7 25.8 14 .5 17 .9 13.2 11 . 9 16.7 coconut 22 .9 34 .8 16 .1 24.5 50.2 42 .2 5.3 14.3 16.7 14 .0 12 .9 banana 21. 3 20.0 14.2 7.6 27 .4 35 .7 17 .5 49.1 12 .2 16 .9 24.3 14 .7 rourou 23.2 28.5 32 .5 32 .8 14.3 24.2 14.5 16 .1 22 .7 14.2 22 .7 11.4 okra 19 .2 46.4 14 .3 7.7 8.4 27.9 10.0 18. 0 eggplant 14 .5 17.7 10.8 9.2 11. 3 10 .9 20. 4 10. 6 12 .2 18.3 25. 1 14.2 chillies 17.4 27.0 14 .2 19 .8 23. 2 17 .5 18.4 10.3 10.1 20.0 20. 1 14.5 Chinese cabbage 13.2 9.3 14 .5 9.8 16 .9 16.1 21. 7 13.5 cucumber 14.0 20.8 14.5 6.9 13.0 12 .4 14.2 10.6 16.1 12 .9 pawpaw 15.3 19 .2 19.5 12 .9 15.6 32 .2 17.1 9.7 6.6 19.6 12 .1 12 .4 lemon 19 .3 11. 6 15.9 11. 3 27.4 12 .7 39 .4 10.9 9.7 13 .7 15 .0 16.1 orange 22.5 22.2 21. 3 17.7 49 .9 67. 3 11. 9 20. 9 16.1 pineapple 18.5 16.6 13.2 26.2 17 .1 27.9 20.1 24.2 22 .4 24.3 13.8 coconut oil 10. 3 4.8 14. 7 8.5 8.2 12 .9 3. 7 16. 4 land crabs 45.9 66.8 51. 4 16 .1 a Data calcu1ated for all foods that have twen ty or more valid responses (travel distance of producer-vendors) . I b - I indicates product not recorded or recorded less than twenty times. c Raiwaqa and Korovou markets are excluded because of inadequate returns.

Source : Market surveys, January 1977. 99

Table 2. 20

Correlat ion between quantity of produce brought to market b� Eroducers and Eroducers ' residential distance from marketa

Numb er Numb er Correl. Correl . markets markets coeff. coeff. recorded recorded

Native staEles 0.044 13 Introduced ----- vegetables dalo -0 . 098 10 0.093 11

cassava 0.237 10 tomato 0.077 6

breadfruit -0. 041 9 Ch . cabbage o. 32 5 7

plantain 0.071 6 cucumb er -0 . 051 8

coconut 0.149 8 Fruit 0.029 13 cooking banana 0.211 5 banana -0 . 182 11

Native vegetables 0.181 12 lime -0. 321 8

rourou 0.168 11 lemon 0.117 11

be le 0.052 7 orange 0.202 7

pawpaw -0. 013 9 Indian staEles -0. 008 12 guava 0.505 4 pumpkin -0.077 7 V- apple -0. 047 4 Indian vegetables 0.022 12 ivi -0. 020 5 okra 0.484 5 pineapple -0.123 8 eggplant -0. 036 11 sugar cane -0.283 4 amaran th 0.050 4 Yagona, tobacco 0.271 3 kerela 0.009 5

chillies 0.062 10 Oils -0. 194 7

'spinach ' -0. 072 4 coconut oil -0. 209 6

curry leaf -0. 491 3 ghee -0. 205 3

roselle 0.030 4

beans -0. 018 10

gourd 0.025 4 a Calculated for all producer-sold foods with residential distance data available for twenty or mo re producers and with data available for three or mo re of possib le thirteen markets. b Raiwaqa and Korovou markets excluded because of insufficient in formation.

Source : Market surveys , January 1976. 100

There are a number of possible reasons for the ab sence of strong correlat ion between distance an d production for market . One is that hinterlands are not sufficien tly expansive to encourage zonation of production by distance from market . Another is that while there is spatial differentiation , it is not a product of distance alone : equally significant are the race of producers and sett lemen t patterns by race. Fij ian producer-vendors (Tab le 2.6) tend to live further from market than Indian or Chinese sellers so it is to be expected that products associated with Fij ian growers come from greater distances . The trend ought to be particularly apparent in markets with hinterlands that exceed the immediate cont iguous area. As long as production of particular products remains related to the race of producer, it is unlikely that there will be significant changes in correlation indices. Environmental requiremen ts of plants, product perishability and price structures that support high-cost transfers are other factors affect ing the apparent lack of correlation , but these also need to be seen in the context of spatial feat ures of settlement an dproduction by race . The influen ce of these factors on supply distance could possibly be deter­ mined by treating Indian and Fij ian products separately. This has not been ·done with our data.

Vertical trade . For many years the produce trade in Fij i was handled almo st exclusively by the producers them­ selves, even after formal markets had been established. Since World War II both non-producer-vendors and produce assemb ler­ whol esalers have become in creasingly connnon . Two measures of the extent of vertical trade within markets are available : the numb er of vendors who are non-producers , and the degree to which trade in particular items is handled by non-producer­ vendors.

Non-producer-vendors account for a greater share of trade than their numb er alone would suggest . At all markets except Savusavu, Levuka, and Navua , wh ere almost all vendors are producers , the one-quarter to one-third of vendors who are non-producers handle more than one-half the produce traded. At Vaileka , for instance, the 25 per cent of different vendors present over one week who are non-producers handle 58 per cent of produce . Another indication of the dominance of non-pro ducer permanent vendors is that in November 1976 in Ba marke� on a Saturday when producer-vendor attendance is at its highest for the week, the 35 per cent of vendors who were non-producers occupied 51 per cent of the selling spaces. 101

Some products are mo re subj ect to vertical transfer than others (Tab les 2.21, 2.22) . Ap art from imported fruit , Indian staples are the product over which non-producers have mo st control , only 1.1 per cent being sold by producers . Yaqona and tobacco (96.5 per cent sold by non-producers) are the local products in which there has been greatest develop­ ment of vertical trade ; nat ive staples (69.1 per cent with producers) andnat ive vegetables (75.1 per cent) are least involved in vertical trade . Between one-quarter and one­ half the trade of other product categories is with non­ producers . There is considerable range in vertical trade development among individual foods comprising the broad product groups.

An important cause of the different ial development of vertical trade is the perishability of products. For instance while 30 .9 per cent of nat ive staples are handled by non­ pro ducers , almo st double this proportion of the long-lasting coconut is sold by non-producer vendors. Other factors con­ tributing to the difference are the race of producers and access of races to resources. Products grown largely by Fij ians are less involved in vertical transactions than Indian-grown items . This occurs with all produce but is most marked with seafood,- excluding fish. The perishability of the product and Fij ian control over the productive environment results in there being practically no development of vertical trade in non-fish seafood. 27

Producer-trade is chiefly in native staples and vegetables and fruit (Table 2.23) . In the smaller Fij ian dominated markets these are the only products sold by producers . Where Indian producer-vendors are numerous, as at Ba market , intro­ duced vegetables and Indian vegetables account for a larger share of producer-sold food. Almost 60 per cent of producer­ sold food is nat ive staples or native vegetables and fruit and the balance largely comprises other vegetables. The trade of non-producer vendors is concentrated in introduced and Indian vegetables and Indian staples, which together account for a little mo re than one-half of non-producer trade . The balance is fruit and yaqona and tobacco . Between market s there is a wide range in the composit ion of non-producer trade , but the general pattern of the overall trade composit ion found in all markets combined is identifiable in each.

Specialized trade involvement in individual product s by non-producers and by race has significant ramifications. I-' 0 Table 2 .21 N a ProEortion of food sold by Eroducer

Native Native Indian Indian Introduced b Yaqona , F ru1t. Oils staples vegetables staples vegetables vegetables tobacco

All ma rkets 69.1 75 .1 1.1 43.9 24.4 31 .6 3.5 33.2 Suva 65.4 78.0 1.2 23. 8 34 .3 26 .9 4.0 29.7 c Nausori 81. 7 94 .0 15 .5 43. 9 49 .2 4 7.8 32.5 Vaileka 83.4 98.2 7.7 11.8 23.3 57 .8 10 .7 Tavua 48.4 85 .1 0.2 54 .7 3.9 41 .8 0.1 Ba 16.7 75 .0 1.8 13.8 55 .7 22 .5 0.7 41.3 Lautoka 64.6 88.6 1.4 37.4 43 .5 32 .8 5.7 79 .2 Na di 13 .4 21 .5 0.5 5.8 4.9 5.3 47.8 Sigatoka 83.8 84 .4 8.0 40.8 16 .5 26 .7 Navua 97 .4 94.9 18.0 23.6 23.4 77. 6 Levuka 100.0 100 .0 100 .0 100.0 100 .0 Waiyevo 100.0 100.0 100.0 100 .0 100.0 Savusavu 70 .9 83 .7 3.3 66.7 64 .9 Labasa 79 .0 100.0 11 .4 2.4 12 .8 5.5 50.5 a Figures are percentages of indicated products sold by producers of the foodstuffs at each market. b 'Imp orted fruit' is excluded because none is sold by producer. c ' - ' indicates no sales recorded.

Source : Market surveys, January 1976 . 103

Tab le 2 .22

Producer-vendor participation in sales of different foods (all markets) Native staples 69 .l Introduced vegetab les 24.4

dalo 78.4 lett uce 9.9 dalo-ni-tana 98.8 tomato 11 .1 cassava 70.5 French beans 6.1 sweet potato 97.6 Chinese cabbage 51. 6 yam (all varieties) 100.0 cucumber 35.4 breadfruit 77. 5 plantain 91. 0 Fruit 31 . 6 coconut 46.3 pawpaw 46.2 banana 13. 7 Native ve getables 75.1 lemon 25 .8 ota 93.9 orange 20.9 rourou 77.0 ivi 74.5 be le 61. 4 water melon 34 .9 pineapple 52 .8 Indian staples 1.1 Imported fruit 0 potato 0 onion 0 Yaqona and tobacco 3.5 garlic 0 rice 2.2 yaqona 2.7 pumpkin 16.3 tobacco 5.9 dhal 0.4 pulses 12 .6 Oil 33. 2 spices (dried) 3.5 coconut oil 48.0 jackfruit 20.6 salad and soya oil 0 'oil' (unspec.) 3. 7 Indian vegetables 43.9 ghee 15.9 okra 26.6 eggplant 58.1 Animal products ginger 12 .1 fowls 100.0 kerela 81.3 fowl eggs 40.3 chillies 63.2 land crabs 100. 0 beans (all varieties other crab s 97.7 except French) 38.2 mus sels 100.0 gourd 22.1 other shell fish 100.0 tamarind 18.8 dried f ish 91.9 amaranth 67.3 prawn 100.0 coriander 26.3 seaweed 100.0 sea cucumb er 100.0

Note : Figures are percent age of sales of each food handled by producer­ vendors .

Source : Market surveys, January 1976. Tab le 2.23 I-' 0 � a Composit io� of Eroducer and non-Eroducer trade

C/J ro ;j .µ r-1 ro ro � 0 > • 0 al ,.... � � ro :> ro ro � 0 al ro 0 .µ ro � al C/J C/J ,.... ro C/J r-1 ;j ·r-1 ro ;j ;j >.. ;j ro .µ 't:I r-1 ro > ;j :> ;j bO :> :> ·r-1 :> ,.0 •r-1 ro r-1 I:: ;j ro ro ro ro ro • ro al ro ro ro � C/) z � E::::! g:;i ...J z U) z H ::s: U) H Producer trade

Native staples 22 .7 22 .6 23.0 38. 7 23. 6 6. 6 13.2 19 .4 14 .6 38.5 90.9 64.9 35 .4 32 .4 b Native vegetables 2.7 2.5 2.9 10.5 2.2 1.1 2.9 1.0 2.9 4.2 2.7 Indian staples 0.2 0.2 0. 6 0.1 Indian vegetables 22.9 23.2 34 .0 10.5 12 .0 25.8 25.0 19 .4 9.0 16.5 10.8 33. 5 Introduced vegetab les 18.0 18.0 14 .1 8.1 0.4 37 .4 29.7 30. 6 40.7 2.9 2.7 1.1 Fruit 32 .9 32 . 7 25.5 29.8 61. 8 29.1 28. 8 33.3 34 .7 38. 5 9.1 21. 6 60.4 27 .7 Yaqona, tobacco 0.5 0.5 2.4 0.4 0.7 2.1 Oils 0.1 0.1 0.5

Non-Eroducer trade

Native staples 4.7 4.9 5.3 5.1 12 . 6 8. 8 4.4 7.8 0.8 0.5 21.2 1.2 Native vegetables 0.2 0.3 0.2 0.3 Indian staples 7.5 6.4 3.0 2.0 21.1 6.3 6.9 5.0 2.4 5.5 Indian vegetables 30 .5 30.5 44.3 49.5 4.9 46.2 25.4 16.2 26.2 35.8 30. 3 39 .4 Introduced vegetables 14.2 14 .2 15.0 16.2 7.2 8. 1 23. 3 34.3 14.9 4.3 0.8 Fruit 35 .7 36.2 28. 3 13.6 43.6 27.1 35 .7 34.1 44.8 59.5 48. 5 28.9 Imported fruit 2.3 2.3 1.0 0.2 0.4 0.4 10.6 Yaqona, tobacco 5.0 4. 8 4.0 12 .6 9.9 3.6 3.7 2.3 10.4 6.2 Oils 0.1 0.1 0.2 0.4 0.1

a Data for Korovou and Raiwaqa markets are incomplete and are therefore excluded. Data are percentage of total producer and non-producer trade in market that is in indicated produce . ' ' b - indicates no units recorded in the particular category. Source : Market surveys , January 1976. 105

Any action or policy affect ing the ma rket involvemen t of a particular race, or of producer or non-producer vendors , will not have uniform consequences for products sold. Moves to limit non-producer involvement in market trade , for in stance , would affect mo st adversely Indian vendors and their products.

Mark�t tran sport

Vendors in Fij i have always travelled to market mainly by foot , bus or at some centres , boat . Where vendors are primarily pro ducers selling small quantit ies of produce that they periodically bring to market , these means of transport may be sat isfactory. But with increas ing trade vo lume , mo re specialized agriculture and the vertical development of trade , various new methods of transport are becoming popular .

There have been a number of changes in market transport . One has been the extension of 'carrier ' 28 services to areas accessible by road but inadequately served by bus. Carriers not only fill gaps in bus schedules, but they have the additional advantage of being ab le to double as goods vehicles and transport large quantit ies of produce . Another form of transport/supply change has been where assembler-wholesalers use vehicles to collect produce purchased in rural areas . They also pick up goods from intermediate assemb ly points, which may include warehouses of wholesalers who deal in the imported goods retailed in markets . A third change is that some producers use their own vehicles to bring their produce to markets. None of these 'new' means of transport necessarily involve increased specializat ion in market transport . But they do suggest the evolut ion of supply systems more appro­ priate to changing producer/consumer demands . A review of non-bus transport of vendors and producers to markets puts this important means of market supply in to perspective .

While at the maj or markets assemb ler-wholesalers perform a very important role in market supply , transport to market is otherwise not specialized. Non-bus vehicle transport is largely in the hands of farmers or individuals who have trucks used for general transport . The distance travelled by these vehicles is sometimes great , just as the amount of produce carried can be remarkable - like six tonnes of coconut s brought to Ba market by four producers from the Ra coast. But non-bus transport should be seen more as a convenient substitute for traditional transport than an indication of transport or supply specializat ion . 106

Non-bus transport was mon itored over one week at Suva , Nausori, Navua , Ba and Labasa market sc29 The surveys confirmed that the 'tradit ional ' modes of transport remain bas ic in the supply of markets. Stock requirements of non-producer vendors are handled by a small number of specialized assemb ler­ wholesalers operat ing their own vehicles; the one-fifth of producer vendors who travel by carriers do so not because of greater market specialization but because of greater convenience. 30

Almost all non-bus vehicles bringing people an d produce to market are made in Japan . Most weigh between one and two tonnes; few weigh more than four . Two-thirds are 'carriers '. Practically all are owned by individuals - rather than by companies, government organizations or Fij ian groups such as village clubs - and largely by Indians; only 10 per cent (Labasa) to 22 per cent (Suva) are owned by Fij ians. The owner or a close relat ive usually drives ; very few drivers are paid employees. The maj ority of vehicles made only one trip to market during the survey week (Tab le 2.24) . Vehicles are most connnonly hired by an individual or a group of pro­ ducers (Table 2.2 5) : relat ively few collect people and goods along a route. The number of people bringing produce on 'carriers ' averages two to three, but ranges up to five or six. Passengers without produce are somet imes carried . Most vehicles are based in rural areas at their owner's house; those based in towns usually operate out of the market .

The markets covered in the transport survey have similar proportions of vendors who are non-producers (Table 2.3)� but the frequen cy with which vehicles are used to transport non-growers ' produce suggests two different forms of non­ producer supply . In one, supplies are brought to market by growers themselves who wholesale all or part of their produce to non-producer vendors. This is connnon at Navua an d Labasa. At Navua no cases were recorded of transport bringing purchased produce and at Labasa only 7 per cent of mon itored trips were with purchased produce .

The other form of non-producer supply involves purchase of produce in rural areas ; there is practically no horizontal movement of produce between market s. Rural assembler-whole­ saler involvemen t is most common at Ba where 37 per cent of journeys were of people with produce they had purchased. Suva and Nausori are· more . like Labasa than Ba in respect of food transport by non-growers. At Suva , 14 per cent of journeys were with non-grower food and at Nausori, 12 per 107

Tab le 2 .24

Vehicles transporting produce to market : selected characteristicsa

Own ership form

Market Individual Company Government Fij ian group

Suva (198) 80 .8 9.1 7.1 3.0

Nausori (45) 84.4 4.4 11.4

Ba (57) 80 .7 15 .8 1.8 1 . 8

Labasa (21) 100.0

Race of owner

Market Indian Fij ian Chine se

Suva (102) 67.6 21. 6 10.8

Nausori (35) 82 . 9 17.6

Ba (57) 84 .2 14 .0 1.8

Lab as a (21) 90.5 9.5

Number of journe�s during week

Market 1 2 3 4 >4 journeys

Suva (203) 59.6 22 .7 7.9 3.4 6.4

Nausori (46) 84 .8 13.0 2.2

Ba (57) 66 .7 24.6 5.3 1.8

Lab as a (21) 85 .7 14 .3

a Figures are percen tages of valid cases (in brackets) at each market.

Source : Market transport surveys , Jtme and Novemb er 1976. Table 2.25

Type of journey made by people bringing food to market in non-bus vehicular transport

Marke t Journey type Suva Nausori Navua Ba Lab as a

Vehicle hired 61 . 7 46.0 17.2 28.9 13.7

Vehicle collects people/produce 3.5 26.5 48.2 4.8 51. 7

Vehicle used by owner , other 34 .8 27.5 34 .6 67.3 34 .4

Note : Figures are percentages of all trips mon itored at each market .

Source : Marke t transport surveys , June and November 1976 . 109

cent . In formation from Suva gives an pressim ion of wh ere food is purchased. Of fifty-four trips involving purchased goods, three-quarters were of produce bought from growers on farms ; half the remainder were with produce purchased at other markets, and the balance comprised goods from importers , such as potatoes, on ions, fruit , etc. , and rice mills.

The length and timin g of journeys reveals in teresting marketing patterns. The average distance travelled is not great , ranging from 24km at Nausori to 42km at Labasa (Tab le 2.26) . The longest journeys are made by Fij ian-owned vehicles (average 37km) . Trips by Ch inese-own ed vehicles , recorded at Suva and Nausori, averaged 13krn and by Indian vehicles, 30km. These trip distances are substantially greater than those made by all vendors of each race .

Th ere is a tendency at all markets, though less so at Labasa, for the longest trips to take place on Monday and Tuesday and towards the end of the week. Mid-week and Saturday trips are relatively short . On Mon day and Tuesday , produce is brought by growers or non-growers to sell to non-producer vendors to form the basis of their weekly stocks . The Thursday-Friday journeys are largely of producer vendors corning for the peak end-of-week trade. Journeys made by growers , and trips in individually-owned vehicles, are mo st connnon from Thursday to Saturday, when about 70 per cent of all these trips were made. In the Central Divis ion markets, over three-quarters of journeys are completed by llam, with more trucks arriving before 6am (10 per cent) than after 3prn (8 per cent) . Time of arrival did not change markedly by day except for .Saturday when all inward journeys were completed by llarn, since there is limited trading on Saturday afternoon . The fewest journeys to all markets were made mid-week , especially on Wednesdays.

Race and marketing

The race of producers is a significan t variable in market operations. This can be deduced from the discussion of market vendors , the produce they sell an d their financial returns. It is useful to examine further some of the racial differences in market involvement and some implications of these differences for the market system.

Markets are readily classified by the race of vendors. Only at Suva, Nausori and Vaileka are Indian and Fij ian vendors present in approximately equal proportion to their Table 2.26

Distance (km) of journeys ma de by trucks an d 'carriers ' transporting produce to selected markets

Suva Nausori Navua Ba Lab as a a All days 35 .6 (350) 24 .1 (68) 27.2 (4) 27.2 (82) 42 .8 (24) Monday 36 . 7 (42) 86 . 1 (4) 25. 7 (1) 48. 1 (8) 12 . 9 (4) Tuesday 35 . 9 (48) 22 . 5 (4) 4 . 8 (1) 25 . 4 (6) 43. 1 (4) Wednesday 33 . 6 (32) 11 . 3 (1) 112 . 7 (1) 15 . 8 (2) 29.8 (2) h Thursday 37 .2 (71) 15 . 0 (16) 69 . 2 (6) Friday 36 . 2 (98) 19 . 5 (32) 39 . 6 (22) 2 7 . 5 (5) Saturday 27. 0 (58) 21 . 2 (27) 24 . 1 (1) 19 . 8 (28) 63. 3 (3) a Figures in brackets are number of valid cases from which mean journey distance is calcu- lated. b '-' indicates no record in relevant category . Boat trips are not recorded . The mean dis tance of the 24 monitored boat journeys at Navua (1 on Monday , 8 on Thursday , 14 on Friday and 1 on Saturday) was 23 km; and at Labasa, the mean distance of 5 recorded boat trips on Saturday was 66 km.

Source : Market transport surveys , June and Novemb er 1976 . 111 representat ion among vendors nationally (Fig. 11) . At Ba, Nadi , Lautoka and Labasa, vendors are predominantly Indian : Raiwaqa, Tavua , Levuka , Navua and Savusavu markets have greater Fij ian partic ipation . Vendors are mo st racially diverse at the larger markets. Ch inese are mo st numerous at Suva; a few are present at Nausori, Nadi, Labasa an d Lautoka . They do not sell in other markets. However, it should be noted that if instead of looking at racial involve­ ment in terms of individual vendors , the numb er of ' vendor­ days ' by race over one week is examined, the picture of racial involvement changes markedly . At Labasa, for in stance , 86 per cent of vendors present in one week are Indians, 31 compared with approximately 59 per cent of different vendors .

Personal characteristics of vendors vary between races. Th ere are no significant differences by race in vendor age or marital status but there are in the representat ion of the sexes . Fij ian vendors are predominantly female (65 per cent) wh ilst Indians are male (73 per cent) . Indian vendors have had less formal educat ion than Fij ians : one-third have had one year or less formal education compared to one-tenth of Fij ians . As suggested by the occupation of their fathers , Fij ian vendors come from mo re rural backgrounds than In dians . Nearly 90 per cent of the fathers of Fij ian vendors had agricultural occupations compared to 70 per cent of Indian vendors . Few vendors ' fathers themselves were described as vendors , although many of each race would have sold at time s in the market. However, mo re Indians than Fij ians come from families where the father was a market vendor (7 compared to 2 per cent) . Amongst Indians, the few Guj erati and Punj ab i vendors are mo st likely to be from commercial non-agricultural backgrounds. 32 The fathers of Chinese vendors were equally in agriculture and market selling (40 per cent each) , shop­ keeping being the other important occupation .

Wh ilst there is little difference between races in the proportion of vendors who are self-employed, there is in the use of assistance in selling. For instance , 57 per cent of Chinese vendors have help in selling, compared with 22 per cent of Indians an d 13 per cent of Fij ians. Ch inese vendors are mo st dependent on the market as a source of income, over 90 per cent claiming all family income is derived from market sales. The corresponding figures for Fij ians an d Indians are 73 and 69 per cent respect ively : for both, agriculture is the chief other income source. Th ese differences in assistance and in come dependency rates reflect differential 112 involvement of races in marketing as perman ent or non­ permanent vendors. The Ch inese are most likely of all vendors to be permanent : 22 per cent are non-perman ent producer sellers . In co ntrast, about one-half of Indian vendors and 14 per cent of Fij ian vendors are permanent vendors . Involve­ ment in non-producer sales is reflected in the type of product sold . Few Indian or Fij ian vendors (15 and 17 per cent respectively) sell an ything but unprepared foodstuffs , but almost two-thirds of Chinese vendors sell other goods .

Fij ian vendors , living an average of 23km from the market at wh ich they sell, live furthest from market and in more isolated areas than other vendors. Indian vendors live llkm away and Chinese 4km. Th ere are considerable differences between markets in the distance travelled by vendors bu t overall the racial relativit ies of distance travelled are ma intained . The average distance for Fij ians ranges from 2.5km for the few vendors at Raiwaqa to 50km in Lautoka. Th e range for Indians is 3.Skm (Savusavu) to 20km (Waiyevo ). Travel distance variation is reflected in method s of transport . Indians use buses or private vehicles , such as taxis, more than Fijians , who travel more in boats and 'carriers ', thus suggesting the relative remoteness from market of the areas in which Fij ians live .

Vendors of each race have characteristic forms of market involvement . For in stance, Indian producer vendors are mainly men , deal largely in introduced vegetables and come from areas near enough to markets for day trips to be possible. Fij ian producer vendors are mo re likely to be women , deal in native vegetables and staples, live considerable distances from ma rket and spend at least one night at market. It is also likely that market sales are a relatively mo re important source of in come for Fij ian than for Indian producer vendors.

Fij ian vendors are just as delib erate in their involvement in market sales as Indians, although few have the conmitment to permanent or intensive market activity characteristic of many Indian vendors. The semi-regular, episodic, marketing involvement of Fij ians is an established mode of participation - as is the dominance of women and emphasis on 'native ' produce (Calvert 1870:82). Similarly , the deeper involvement , broader range of activity and seemingly greater in itiative of Indian vendors are long-established traits. An important trend in Fij ian participation , however, is the development of non­ traditional forms of involvement , especially in a non-producer permanent vendor capacity. 11 3

Fij ian expansion into new market roles is not easy . Cultural obstacles to Fij ian involvement in connnercial exchange both restrict entry to permanent market activity an d mould the pattern of involvement among the small numb er of Fij ian non-producer vendors. Asked why there were not many Fij ian permanen t vendors at Suva market one of the few replied : 33

I think most Fij ians are ashamed to do this kind of job. When they see me selling , Fij ian relat ives have very mixed feelings . Some are happy to see me doing such a business while others are surprised at my courage and determination in taking such a course . I used to tell my friends and relat ives that selling in the market is not only interesting but is also an important and worthwh ile occupation for Wlemployed Fij ian housewives and for Fij ians in general . Selling in the market is most profitable money-wise as I have been proving. More Fij ians should take the initiative to sell in the market as it is indeed a rewarding way of sat isfying our financial needs. It should also be a good an d useful stepping stone to the world of business.

Another felt that there were few perman ent Fij ian vendors because

mo st of us Fij ians could not see the value of selling in the market because most of us have not done this kind of job, to sell in order to earn mon ey . Also, selling is a new activity of Fij ian culture. Some people therefore are not prepared to connnit them­ selves to doing it. To a lot of Fij ians it is a shameful kind of job. Some Fij i<1Ils also still feel that only the men have to earn mon ey for the family and all women have to do is look after the home and children . I think that we women can contribute a lot towards the rtnlning of our family , the educat ion of our children and so on if we do such jobs as selling in the market to earn the money that is becoming increasingly valuab le to us today .

Other obstacles to involvement of Fij ians in marketing are structural. The supply of many items handled by Indian permanent vendors is 'tied up ' with in formal supply and credit arrangements between producers or assemb ler-wholesalers an d vendors . It is difficult for Fij ians to gain regular access 114

to these products . On the other hand , as few other permanent vendors deal in native vegetables and fruit , Fij ians have had little difficulty in establishing supplies of these product s.34 In moving in to permanent selling of root crops and native vegetables , Fij ians have concentrated on products that were previously largely ignored. Because of this , they have had little difficulty in gainin g a foothold and in many cases they have received considerable advice and assistance from other vendors. It remains to be seen if such goodwill will continue if Fij ians attemp t to trade in goods at present largely handled by Indians , and vice versa.

Wh ile the attitudinal , motivat ional and structural impedimen ts to Fij ian en try are considerable , Fij ian permanent vendors invar iably stress the ease of their actual entry and the advantages of market selling over other occupat ions. Apart from being ab le to occupy a vacant niche , a chief reason for the ease of en try is that some traditional obstacles to successful Fij ian economic enterprise are not present in the market.

The market is a public commercial arena where all tran s­ actions are in cash . The demands, of family and other rela­ tives , which can debilitate Fij ian stores are largely avoided : even when relat ives do come to market , the public/cash character of transactions makes it easy to deflect or lessen their expectations and demands. As produce must be purchased daily , losses are readily apparent and business can caase without significan t debts being accumulated . Even more advant­ ageous is the fact_ that markets are centres of commerce. Through competition for supplies and customers , Fij ian vendors are ab le to learn much more ab out the functionin g of commerce than would be possible operat ing a typical isolated Fij ian general shop . Wh ere shopkeeping tends to reinforce the commercial isolat ion of Fij ians , selling in the market , especially but not only as permanent non-producer vendors , draws them into the ma instream of commercial activity. As well as being a pro fitable end in itself, market selling is probably the mo st effective training in connnercial activity currently av ailable to Fij ians - and it is at no direct cost to the participants, the public or the government .

The National Marketing Authority

The National Marketing Authority was established by the 115

Marketing Act of 18 March 1971 'to make the best possible arrangements to facilit ate an d develop internal an d export markets for the products of Fiji' . Specific obj ect ives defined by the Authority 's first board included provision of a guaranteed market for particular products, mainten ance of supplies to urban markets - the pricing of wh ich would help dampen inflation ary tren ds - close liaison and co-ordin­ at ion with rural development , agricultural extension and established marketing agencies , and general market research for local produce. 35 The NMA was not meant to be profit making but was expected to break even in trading. Trading figures for 1971-75 are presented in Table 2.27.

The NMA operates as a produce wholesaler and retailer , buyirii.g in both rural areas and at its urb an depots. Sales within Fij i are almo st all retail. External sales are primarily wholesale consignment transaction s to New Zealand. Purchases are made in response to perceived demand. Exten sion staff of the Departmen t of Agriculture supply in format ion on produce availability and co-ordinate harvesting with visits by NMA purchase teams . Payment is on a farm-gate, cash basis . Un solicited produce may be taken to NMA depots, but unless the item is in short supply, such produce is rarely accepted. The NMA does not make contracts with producers although a few farmers receive favourab le treatment if they have estab­ lished a reputat ion for reliable delivery of supplies ; these people are ab le to sell the Autho rity a large proportion of their produce . Suppliers are predominantly producers but produce is also accepted from middlemen . Overall , the NMA handles only a small part of what farmers could sell .

Most produce an d seafood handled by the NMA either goes to government supply contracts and orders from institut ional and large-scale purchasers like hotels , or is exported. Retail facilities are maintained at NMA depots and in some markets, but very little of its trade is direct to consumers . The NMA has become increasingly involved in sales of vegetables, mainly dalo , and fruit . Exports were negligible until 1974 but in 1975 comprised more than on e-third of sales .

NMA headquarters are in the Suva suburb of Vatuwaqa wh ere an 11 , 000 sq. ft . warehouse/office was opened in January 1974. At Vatuwaqa there are six fifteen-ton capacity coolers for produce storage . Two-thirds of the forty to fifty employees are casual wo rkers engaged in receiving and preparing produce for sale . A second warehouse with freezer facilities has recently been built at Lautoka , the centre of Western 118

Tab le 2.28

Origin of NMA purchases by division and province

1971 1972 1975 1976

Central Divis ion 0.6 7.0 42 .3 48.1 Naitas iri 4.3 31 . 2 29.0 Rew a 0.4 2.2 3.4 Tailevu 0.6 1.1 5.4 4.7 Serua-Namosi 1. 2 0. 9 1.0 Suva market 2.6 10.0

Western Division 44. 2 24. 6 19.2 Ba 0.3 18.7 10.1 Nadroga and Navosa 43.9 5.0 5.3 Ra 0.9 3.8

Northern Division 67.3 28.0 15 .9 11. 7 Cakaudrove 67 .3 27.7 11. 1 6.8 Macuata 0.3 4.8 4.9 Bua

Eastern Divis ion 32 . 0 20. 7 17.2 20.8 Kadavu 0.8 9.0 10.7 10 .7 Lomaiviti 31.2 5.0 1. 7 3 .3 Lau 4.7 4.4 4.2 Rotuma 2.0 0.4 2.6

Note: Figures are percentage of total purchases for each year .

Source : Purchase Analysis , 1972-1976, NMA.

There have also been changes in the type of produce purchased ( Table 2.29) . A main factor behind government involvemen t in marketing in the late 1960s was concern over mi ddleman involvement in vaqona marketing. One mot ive for establishing the NMA was to attemp t to counter perceived profiteering by middlmen . Consequently , in the first years on NMA activity , yaqona comprised a large share of trade : in 1971 it accounted for 94.3 per cent of purchases . Once 119

Table 2.29

Product composit ion of NMA purchases , 1971-76

Product 1971 1972 1973 1974 1975 1976

Root vegetables dalo 3.2 32 .0 27 .6 34 . 3 37. 6 60.3 cassava 0. 8 3.5 4.4 2.5 sweet potato 1.0 3.6 2.0 1. 7 yam 1.1 1.6 0. 3 1.1 dalo-ni-tana 0.2 1. 2 0.2 0.4 Irish potato (o ' seas) 7.6 3.6 Irish potato (local) 42 .6 22 .6

sub-total 4.3 76.6 67.7 44 .8 43.3 60.3

94.3 18. 6 9.2 3.0 7.1 4.3

Other fruit and vegetables banana 3. 7 4.8 2.6 watermelon 0.3 0. 5 1.0 citrus 2.4 0.6 coconut 1. 7 tomato 1. 3 1.2 cucumber 1.4 0.6 eggplant 1.0 1.0 pumpkin 0.6 0.1 cabbage 0.9 1. 5 lettuce 1. 3 0.3 rice 1. 3 other fruit , ve getables 10.5 4.2 18. 6 peanut 0. 3

sub-total 4.0 21. 8 16.4 21.5 10.7

Animal products meat , eggs 0. 7 2.9 0.9 fish 0.1 32 .2 27 .0 24.7

Non-foods voivoi (pandanus leaf) 1.3

Note : Figures are percentages of total purchases in each year.

Source : NMA, Purchase Analysis ( 1971 , 1972 , 1975, 1976) , Annual Report ( 1973) and Commodity Trading Results (1974) . 120 the Authority became better established , however, yaqona lost mu ch of its dominance and trade emphasis shifted to dalo . In 1976, 60 .3 per cent of NMA purchases was dalo . Another maj or change in coIIllllodity trading occurred with the 1974 takeover from the Fisheries Department of trade in fish and other marine product s: these products comprised one-quarter of all NMA purchases in 1976. Involvement with dalo and fish has resulted in the trade in other vegetables and fruit falling from about 20 per cent of purchases in 1973-75 to 10 per cent in 1976 .

.. Over the period for wh ich trade results are available (1971-75 , Table 2.27) , the NMA became in creasingly dependent on government grant s. One reason for this has been the with­ drawal of un charged governmen t services - staff and facilit ies - as the Authority became better established, but a maj or reason is the rising cost of administration and marketing. Administ rative co sts of $38, 200 (11.5 per cent of total sales) in the first full year of operation rose to $370, 700 (40 .9 per cent) by 1975 . A large part of the increased costs can be attributed to involvement in export trade , especially as most overseas sales are airfreighted.

Price markups and pro fit margin s in produce marketing, as well as in other areas of commerce, are a contentious issue in Fij i and , as noted above , were a factor leading to the establishment of the NMA . Actual markups on produce applied by the NMA are not available , but an idea of the range of these comes from data on gross trading margins and also from actual markups on marine products.

Gross margins , that is , sale price less purchase price, on over fifty products in 1974 ranged from 37.9 per cent profit on cassava to 127 .7 per cent losses on mangoes . Gross profit as a percentage of sales for all items was 28. 7 per cent, including profits of 31. 6 per cent on root crops and 36 .4 per cent on other vegetables; losses were 16 per cent and 32 .9 per cent on fruit an d marine products. As for marine products, the average retail markup in November 1974 was 61 per cent and ranged between 33 per cent (lobster) and 100 per cent (clams) . The markup on fish was between 35 per cent (Grade 1) and 90 per cent (Grade 4) .

CoIIllllodity trading in 1975 , at least in the Suva NMA branch , was more successful than in 1974. Overall gross trading profit was 56 per cent; losses were recorded for only one coIIllllodity, cabbage . Gross margins on other products 121

ranged from 7.1 per cent (yaqona) to 90 per cent (breadfruit ). Dalo had a 51.2 per cent gross ma rgin . Data on wastage due to spoilage are not available. As wastage is often consid­ erable , it is likely that mo st produce markups are cons iderably greater than the gross margins. How NMA markups compare with in dividual traders (see Tab le 2.11) is unknown ; the data quoted here , however , suggest that they are unlikely to be substantially less even though all NMA costs are not recovered through sales.

Financial results are not necessarily a val id ind ication of the NMA' s success. As established, the Au thority had a variety of obj ectives, ranging from maintaining a steady flow of reasonably priced, high quality produce to market, to creat ing a marketing infrast ructure wh ere none was present (see footnote 35) . Leaving aside the question of financial performance, the feeling that it has not been a great success is quite widespread amon g producers an d consumers acquainted with the NMA.

The NMA has suffered from unrealistic expectat ions from many sides . It is difficult to determine the actual nature of these, but among the most connnon were that the NMA would provide a guaranteed outlet for most , if not all , local produce - particularly produce grown at the instigation of agricultural extension officers - would 'save the consumer' by driving the 'parasitic middlemen ' from the market system and would become the chief internal marketing mechanism.

Few of the official objectives (wh ich are somewhat dif­ ferent to these popular assumptions) have been achieved. There is no �uaranteed market for specified primary produce at predetermined prices for different product ion zones ', an d 'a steady flow of reasonably priced, high quality produce to [internal public] markets' has not been maintained, so a contribut ion 'to combat in flation by dampening upward spirals in produce prices ' is dubious . 36 The Authority has provided markets in some areas previously without access to connnercial outlets, but co-ordination with rural development agencies , ext ens ion services and other marketing bodies an d organ izations has been weak , even though the Authority relies heavily on such bodies for supply access.

Three chief factors contribute to the limited achievemen ts by the NMA. Only one - the service of areas with difficult market access - is the product of officially-stated policy, but all have a fundamental influen ce on NMA operations. 122

The commitment to serve regions with limited alternative commercial outlets , wh ich is a maj or policy position (Fij i, Central Planning Office 1975:67), has resulted in priority being given to Kadavu , Lomaiviti, Cakaudrove and the Lau islands of Totoya, Matuku and Moala. As produce an d transport become available , the NMA arranges purchases , buys on the beach and transports produce to Suva in vessel-space it hires at commercial rates. Comparable effort is not spent on pur­ chases from other areas . Farmers around Suva may be actively pursued for produce , especially dalo, when supplies are short, but otherwise it is felt that producers on Viti Levu - especially non-Fij ian producers - have market outlets that they can utiliz e on their own initiative .

The provision of market access for producers in isolated areas is des irable in an equity sense , but it does highlight two key issues in the NMA' s activities . One is the incom­ patibility of the Authority 's obj ectives of serving both producer and con sumer. If retail prices are based directly on production and tran sport costs, consumers are probably most cheaply supplied from nearby , accessible areas. The second is the way in wh ich the Authority is used to support and promote connnercial production by Fij ian farmers.

Fij ians are the chief suppliers of produce to the NMA (Table 2.30) . Fifty-five per cent of purchases made by the Suva NMA branch over six alternate mon ths in 1976 was from Fij ians; Indian growers contributed 32 .7 per cent and Chinese 9.5 per cent . Purchases of dalo are even more dominat ed by Fij ians - Indian and Chinese growers together contribute only one-third of dalo purchased by the NMA. Some Fij ian dominance is to be expected as dalo is the .main food crop grown by Fij ians, but there is extensive Indian and Chinese dalo farming in Naitasiri Province that could meet a large part of NMA requirements.

Lobbies representing the largely Indian producers in areas near Suva, such as the Central Farmers ' As sociation of Tailevu and Naitasiri, find ready ground for criticism in the apparent spatial and racial bias of purchase origin. Recently, an increasing amount of produce has in fact been purchased from areas near Suva , but this has occurred because of a failure to obtain predictable supplies from more remote areas rather than because of deliberate policy reorientat ion . Eventually a decision will have to be made whether the mo re flexible and cheaper , but largely non-Fij ian, sources of dalo near Suva are to be encouraged at the expense of island (Fij ian) sources. 123

Table 2.30 a NMA purchases by race of seller

b c Seller Produce Dalo

In dian 32 . 7 25. 6

Fij ian , Ro tuman 43. 8 43.4

Chinese 9.5 8 . 4

Other 1.8

Race not known 1. 1 0.4 d NMA, Agricultural Department 11 . l 22 .2 a Figures are percentages of produce (including dalo) and dalo purchases (by value) made at the Suva NMA depot from sellers of the indicated race . Ra ce is derived from name of seller . b 'All produce ' data are based on purchase records of six alternate months in 1976. c 'Dalo ' data are from all dalo purchases made between Decemb er 1975 and November 1976. d Produce under 'NMA, Agricultural Department ' is purchased from growers by agriculture field officers or NMA purchase officers; mo st of the produce is grown by Fij ians.

Source: Purchase Journal , December 1975 to November 1976, NMA, Suva. 124

A second factor contributing to the NMA's weak showing is its thinly disguised hostility to intermediaries, other than itself, in the produce trade . Seemingly no chance is mi ssed to blame the 'middlemen ' for a wide variety of ills. Representatives of a farmers ' lobby seeking greater NMA purchase activity in the Central Division , for in stance , found themselves being assured by the NMA representat ive that mi ddlemen , not the NMA , were the cause of their diffi­ culties and that farmers ou ght to join with the Authority against the middleman . 37 Antagonism towards middlemen is ironic as the NMA is itself the largest produce in termediary organ izat ion in Fij i.

Un reliable deliveries an d in consistent standards regarding the quality of produce are among the mo st pressing problems facing the Authority. Th ese are to be expected when the NMA deals with an almost infinite number of small-s cale producers who are irregularly involved in marketing. As sembler-wholesalers in direct contact with producers could do useful work by assemb ling produce, enforcing quality standards, and supervising delivery schedules suitable to the NMA. The middleman is essentially a commercial being, likely to appreciate the necessity of maintaining quality and punctuality . Moreover , it is eas ier and mo re efficient to operate a business through a finite number of recognized mi ddlemen than through a mult iplicity of producers .

A third factor behind the unfavourable impress ion created by the NMA is the Authority's organizat ion . The administration appears to many producers as in flexible , insufficiently field-oriented and arbitrary. Too much was attemp ted before a solid foundation of either policy or practice was established . Inflexibility is apparent in attitudes to intermediaries and in pricing practices . Few producers dispute that prices must ultimat ely be based on market conditions rather than product ion costs, but many regard NMA prices as unreflective of market realities . Complaints are not so much that NMA prices are consistently lower than those offered by the few other bulk purchasers - or which can be obtained in markets - but that they do not reflect marke L trends. Dalo is a case in point . In all markets, the price of dalo almost doubles towards the end of each year , reflecting lower production in the preceding drier mon ths and greater seasonal demand. Prices paid by the NMA, however, remain virtually lfilchanged throughout the year (Table 2.31) . Similar price inflexibility is apparent in prices paid for other items . Moreover, it is lIDlikely Table 2. 31

NMA dalo bu�ing Erices, 1976 a Buying price (cen ts per lb) Production area Purchase locat ion Jan . Mar . May July Sep . Nov . Naitasiri' 7.6 6.3 6.3 6.3 6. 7 6.7 Vatuwaqa Depot Tailevu 6.0 6.3 6.7 " II Ba (Bucalevu) 6.3 " " Mo toriki 7.0 6.3 6.3 6.3 7.0 " II Karo , Gau 6.5 6.3 6.0 6.0 6.0 6.0 Island loading point Kadavu 5.0 5.5 6.3 6.0 II II " Mo ala 7.0 6.0 6.0 6.0 JI " I I Rotuma 6.0 6.0 " " " Taveun i 5.0 5.5 5.0 6.0 6.0 6.0 " " " a In most location s on ly one price was paid over a particular month . Where more than one price was paid in one month , the most frequent is recorded; the range is rarely mo re than 1 cent. Price variation - in response to produce quality , general dalo price levels and urgency of NMA' s needs - within one mon th is mo st common in Naitasiri. Ab sence of price in formation means no purchases were recorded for that month .

Source : Purchase Journal, 1976, NMA , Suva . 126 that price differentials between into-depot and farm- gate prices adequately reflect transfer costs; island suppliers are subsidized .

Organizat ional shortcomin gs are apparent in a mo st crit ical area - contact with producers. The NMA has virtually no field staff; officers of the Departmen t of Agriculture act as market intelligence sources for both producers and the NMA. Even with this assistance, relationships with suppliers have invariably soured because of a breakdown in ' commitments ' by either the producer or the NMA.

Events in Kadavu in 1975-76 are not un typical of supplier­ NMA relations. In February 1975, an NMA buying agent was sent from Suva to the Nabukalevu District to purchase twen ty tonnes -of dalo that was ·supposed to have been grown an d har­ vested specifically for the NMA . The agent managed to collect only 9-1/2 tonnes, the farmers telling him 'they no longer have faith or trust the representatives of the Agricultural Department due to a lot of false and misleading information '. Within a year, events had come full circle. An NMA agent early in 1976 f otnld himself obliged to purchase thirty-eight tonnes of dalo brought to the ship , rather than the arranged twenty tonnes. The response of the NMA was to announce that in future it would purchase only the agreed-upon quota of dalo ; if it was oversupplied , purchases from Kadavu would be suspended. 38

The NMA never has provided a 'guaranteed market for specified primary produce ' but it has dealt with a wide range of items , the profitability of which is not uniform and in many cases doubtful : highly perishab le leaf vegetables and fruit are a case in point. Because its prices are low and its decisions to purchase apparently capricious , reported as being subject to pressure from growers and variable in standards of quality , the NMA has come to be regarded as a buyer of produce that cannot be sold elsewhere. This has resulted in the NMA finding it difficult to establish maj or domestic. outlets and has not en couraged producers to specialize in production or pay attention to grading. All in all, the NMA has not been recognized as an effect ive or .efficient marketing organiz at ion .

There is no doubt that the NMA has had a profound in fluence on the marketing of produce from more remote regions and has inj ected considerable amotmts of money into rural areas . The Authority has not, however, been as successful 127 in other ways . Instead of becoming in tegrated in to an d in creasing the efficiency of the existing marketing system, the NMA has created a parallel organ ization . This was an inevitable outcome given that a significant mot ive in estab­ lishing the Authority was to 'drive out ' the middlemen . Involvemen t in marketing was not to be co-ordinated with the system that had been developed over seven ty years , but to replace the system. Attempts to establish a foothold in the public marketing system at both the retail or wholesale levels failed. Pro tracted conflict with the established system was avoided because the considerable financial , staffing and power resources of the NMA could be channelled in to a largely unoccupied fun ction � the movement of produce from isolated areas for in ternal contract purchasers and for export .

This choice of action enabled the NMA to gain some expertise in produce trading, but it is potentially dangerous . Expo rt markets for produce are not a guaranteed outlet . If export markets become closed to produce from Fiji, NMA oper­ ations will be dislocated. The Authority will then have to recorrnnence establishing internal out lets. A profitable short-term solution was folllld to the problem of oversupply of produce, but only at the cost of avo iding the critical issue of developing viable links between producers (in par­ ticular isolated producers) and in ternal markets.

Conclus ion

The basic structure of the produce distribution system and the nature and extent of participat ion in it by different racial groups require no more elaboration . To conclude dis­ cussion of the produce distribut ion system, however , a number of issues deriving from the above analysis should be noted.

One is that while produce markets have a relat ively central role in economic and even popular social and poli­ tical life in Fij i, they are not well understood. Many of the fault s attributed to them - such as being dirty , crowded, congested and contributing to congestion - are not inherent to markets but arise because the markets were established in a supply and deman d environment quit e different to that of today . Markets were created to provide producers with an outlet for their produce an d the consumers of the then relat ively small towns with a source of fresh produce . Con­ flicts between producer an d non-producer vendors , food and non-food vendors and vendors and market administrators are 128 all symptoms of the in creasing disharmony between the function of market design and contemporary reality. Given the in­ creas es in demand and numbers of producers, it is an indication of the resilience and strength of the fundamental concept of produce markets that they continue to meet the needs of both producers and consumers relatively efficiently . The success is even more surprising in light of the trend for market s to take on a third function - to act as a maj or source of revenue for local councils .

It was intended that markets would pay their way , but the surplus of revenue over receipts at some markets is well in excess of justifiable reserves for capital improvement s. Even after allowing for the urban-dwelling permanent vendors, it would appear that rural producers are being taxed to support urban services. It mus t be decided whether fees should subsidize urban population or cover only recurrent expenses. Market fees are an efficient form of revenue collection , derived mainly from rural populat ions; if they are to be treated as a tax, the surplus should be invested in proj ects beneficial to producers who sell in urban market s. Improved and subsidized transport services and subsidies on farm inputs would be appropriate areas for investment. The fact that no market administration has substantial vendor - and no producer­ vendor - participation does nothing to counter the feeling of producer-vendors that they are being taxed for services they neither want nor use. 39

A second issue is that while a number of courses have been suggested to resolve Fiji's perceived internal produce market problems , for example, high prices and irregular supply, the basic problem of get ting produce into the market system is rarely acknowledged. The amount of produce entering commercial exchange is conditioned by producers' income expectations and needs . So long as these are either so limit ed or so high that they cannot be met by marketing, 'guaranteed prices ' or other popular remedies such as rotat ion of production zones, provision of storage in rural and urban areas and product grading, will not result in increased or more reliab le supplies . Many farmers have established a level of involvement in market ing that , wh ilst slight an d even irregular , is adequate to meet their financial requi rements. The present urban markets are a satisfactory marketing channel for such a connnercial subsistence producer as long as they can ab sorb the produce he wants, or needs, to sell. He stands to benefit from a national marketing body only if it can dispose of produce he is unable to sell through 129

urb an markets an d at a price not sign ificantly lowe r than local retail prices .

Given this production environment , the volume of produce en tering commercial trade can be increased in two ways. First, producers previously out of market contact can be integrated in to the system. However , their potential involvement is limited because of the subsistence-connnercial nature of pro­ duction of many and their limited financial needs. An alter­ native is to tap the production of commercial farmers who use mechanized product ion methods and wage labour . Given full encouragemen t, these farmers can contribute significantly and reliably to internal demands. This policy may be dis­ tasteful in the Fij ian context because the farmers who are presently concerned are mainly non-Fij ians and their involve­ ment in the market system to the limit of their productive capability could limit involvement of Fij ians as well as other subsistence-commercial producers.

A final issue is that of the 'middleman '. Produce marketing, as indeed is much of the commerce in Fij i, is beset by an anti-middleman attitude . Typical of this att itude is the response of a senior MAFF officer to a request to prohibit the sale of imported food in Ba market : 'A very progressive idea! If it comes to pass I feel certain it would eventually drive out those middlemen who for so long being [sic] the cause of in flation and also hindrance to the marketing of locally grown produce '.40

Anti-middlemen postures are en countered among producers, consumers and, as we have seen , government officials. In part , the att itudes are due to unawareness of the role and funct ion of the in termediaries and are compounded by un real­ istic impress ions of their net margins and the fact that middlemen tend to be racially specialized. Whatever their origin , however, anti-mi ddlemen attitudes have a considerable negative impact on the produce dist ribut ion system. At the broadest level , they have resulted in the government 's role in marketing being seen, by the government it self and others , primarily in terms of confrontat ion and competit ion with the established 'middleman-dominated' market system rather than bein g of assistance to an d improvemen t of a proven system.

The present an d potential contribut ion of in termediaries to produce marketing in Fij i is significant . Given the multitude of producers , the isolat ion and distance from markets of important production area� the disinclination of 130

many producers to retail produce and the level and consistency en cy of demand, it is inconceivable that the marketing system could funct ion without intermediaries. Many present problems in the supply and quality of produce could be overcome by encouraging the development of produce assemb ler-wholesalers and transporters. Professional market vendors , that is , 'middlemen ', contribute sign ificantly to improving supply and price flexib ility of produce, as well as being a vehicle through wh ich quality controls can be enforced. Apart from assembling and grading produce and stabiliz ing supply , middle­ men perform another function : market selling and other forms of intermediary activity are sources of employment and effec­ tive means of conunercial training.

There is , of course, a role in the market system for small-scale producer vendors. The limited post-harvest life of mo st products discourages multiple trade links . Moreover, produce retailing by pro ducers is an effective me ans of supplying immediate cash needs. The dependence of an efficient pro duce distribut ion system on both in termediaries and pro­ ducer-vendors is beyond dispute.

How can Fiji's produce distribution system meet the often conflicting interests of pro ducer and consumer more efficiently wh ilst at the same time being used to promote greater Fij ian involvement in commerce?

One way is to recognize that the two main components of the internal produce marketing system, the urban produce market s and the Nat ional Marketing Authority, are integral parts of one system. Each has a particular and different role. Either to support or discourage one part results in disequilib rium in the total system. Attention needs to be given to the total marketing system and to defining the com­ plementary functions of its parts. This could be done by establishing a produce marketing board as an in dependent arbiter to oversee the broad concerns of internal marketing in both the public and private sectors . Membership of the board should include representat ives of the Department of Agriculture, the Nat ional Marketing Authority, urban market administrations, permanent and non-perman ent market vendors, produce assemb ler-wholesalers , commercial and commercial­ subsistence producers and consumers .

A secon d way is to acknowledge the critical role of urban produce markets in the marketing system. Specifically, this 131 coul d be done by:

(a) put ting markets under the oversight of a national body , such as a produce marketing board (mentioned above , establishing a minimum reinves tmen t rate for market revenue and including producer and non­ producer vendors on market administrat ions; (b) providing facilities that would assist the acti­ vities of intermediaries, such as transport loading and unloading facilities and areas adi acent to markets wh ere retailers could purchase goods wh ole- sale ; (c) developing facilities that en courage producer­ vendor retail part icipat ion : basic amen ities are the provision of sufficient selling space , sleeping, toilet and bathing facilities and accessible bus and 'carrier ' unloading areas .

Because of the central role the Nat ional Marketing Authority has in the government 's scheme of produce marketing and because of its potential in reality, its achievements an d appropriate function s should be realistically reviewed. In particular, consideration should be given to : planning compensatory action against possible dislocat ions of export markets, encouraging rural assemb lers to take over many supply function s and to en force quality standards, developing field staff and a field orientat ion , improving relat ions with pro­ ducers an d the extension staff of the Department of Agricul t ure and focusing trade activity on a small number of profitable lines .

A fourth possible act ion is for the public , including the government , to reconsider the role of intermediaries in produce market ing. Moreover, in addition to fulfilling present roles , in termediaries should be encouraged to take on other productive fun ctions. In particular , they should be encouraged to become further invo lved in the purchase and assembly of produce in rural areas and its subsequent tran s­ port and to act as agents for consignments of pro duce from isolated areas.

Finally , it should be recognized that, mo re than any other field, markets offer a proven avenue for Fij ian involve­ ment and advance in commerce. As a public trading place , traditionally based on cash sales an d requiring daily accounting of income to enable purchase of goods to sell the 132 next day , urban produce markets are an environment that overcome some of the key problems facin g Fij ians in commerce. Fij ian non-producer activity should be en couraged by making Fij ians aware of the profitability of market selling an d its advantages over other occupations and common forms of Fij ian connnercial activity. Minimal capital investmen t would be required; the chief concern might well be to en sure adequate stall space for Fij ians. As well as en couraging Fij ians to become produce retailers, they should also be stimu­ lated to act as consignmen t agents for produce from isolated areas and as rural assemb ler-wholesalers , both funct ions at present being open to easy entry as they are relatively unoccupied. Chapter 3

THE PROCESSED FOODS DISTRIBUTION SYSTEM

Following the analysis of the produce distribut ion system in the previous chapter , attention is now focused on the other main part of Fiji's food distribution system, that of processed foods. I The broad structure of the processed foods distribution system is examined - as is the analysis of the produce system - with particular attention being given to the involvement and influence of goveTilment and different races. The analysis is structured around the role of participants such as indent agents, importers, wholesalers , retailers and customers . This approach is not altogether satisfactory for two reasons . F1rst, although these functions are distinguished by participants, they are not so distinct in reality . Second , the approach focuses on the distribution system as it operates in Fiji. External link­ ages significantly influen ce supply sources and efficien cy and , above all , prices. However, analysis of external connections is logistically difficult and sustained investi­ gation is discouraged by corporate structures. The following analysis of the processed foods distribut ion system in Fij i should be seen within these constraints.

Local considerat ions

As with produce, the distribution of processed foodstuffs (for convenience , the term 'groceries ' will be used) in Fij i is af fected by factors that are to some extent peculiar to the Dominion . One such factor is the influence of the race, culture and religion of the trader and customer on the distribution system. This in fluen ce is particularly apparent at the retail level .

Whilst a large part of the stock of retail grocery stores is commonly consumed foods, actual stock composit ion varies in relation to the ethnicity and religion of clientele. The chief food stocks of shops serving In dians are rice , sharps, dhal, ghee, cooking oil, spices , pot atoes, on ions,

133 134 garlic and canned corned beef or mutton. Shops with a Fij ian clientele stock rice , onions and corned beef, but sell flour rathe r than sharps, have greater stocks of canned mackerel and dripp ing than Indian stores and rarely carry oil, potatoes , garlic , spices (apart from pre-mixed curry powder, which is not used by Indians) , dhal, ghee or corned mutton. The religious beliefs of storekeepers may also be reflected in stock composition : one Moslem storekeeper ceased stocking yeast when he learn t it was used to manu­ facture 'home-brew' rather than for baking. Cattle products , pork (and in deed all meat s) , tea and coffee are foods potentially affected by religious taboos.

Ano ther factor somewhat peculiar to the gro cery trade in Fij i is the marked spatial characteristics of the processed foods distribution system . The chief reason for this is that , especially outside urban centres, areas of Indian an d Fij ian settlement are largely discrete. Most rural shops cater to a clientele of one or other race. Moreover , as all shops in Fij ian villages are owned and operated by Fij ians of that village - or by a co-operative society store - and Indian stores are chiefly located in Indian areas , race is even more a factor in rural trading than settlement patterns alone suggest. Only occasionally is the racial exclusiveness of rural outlets breached by a strategically located Chinese shop serving a predominantly Fij ian clientele, or an Indian shop with Fij ian and Indian custom.

Settlement in urban areas is more in tegrated than in rural areas , and con sequently the stock of urban shops owned by a storekeeper of a particular race is frequently broader than in rural areas. However, there are virtually no Fij ian shops in urban areas ; trade of Fij ian shops is both almost exclusively with Fij ians and in a limited range of items sold largely to Fij ians.

A third factor affecting the processed food trade in Fij i is limited and seasonal demand for grocery items . Demand is low because of limited disposable incomes, especially in rural areas, and because rural residents, as well as even some urban res idents, produce a large part of their food requirements. Home production is mo st marked among Fij ians. Indians do grow some vegetables, but with staple foods of rice and dhal, even in rural areas they depend on purchased food more than Fij ians. Demand for mo st store­ purchased foodstuffs is elastic. Salt , flour , baking-powder 135

and perhaps tea and canned mackerel are 'necessities ', but consump tion of other products like canned milk and corned beef, dry biscuits and (for Fij ians at least) rice is dis­ cret ionary. There is a tendency in urban areas for the dis­ cretionary items (and bread) to become necessities and even for produce to be purchased , although home gardens are an important food source for many urban families.

In both rural and urban areas, demand for consumer items , including food, reaches a peak just before Christmas , but equally important is the seasonality associated with agricultural activities. The period between January an d March is generally characterized by low consumer demand levels. It is used therefore for restocking by wholesalers and retailers alike . Many smaller isolated rural stores are practically moribund during off-peak trading periods .

A final factor affecting the processed food trade is the limited formal connnercial training of traders which contributes to an under-capitalized , pragmatic approach to the business. Few participants in the processed food trade , especially in the locally-owned sector , have had any formal business training. Levels of formal education of retailers , wholesalers and importers alike , are low. Business practices are very much a product of experience. Some local traders have considerable initiative , but with neither the training nor experience equal to that of expatriate firms t mo st are disadvantaged in relation to these firms : it is not surprising that , notwithstanding the energy and initiative of local me rchants, overseas firms are the mo st important source of innovation in the grocery trade , especially in retail activities.

The processed food trade in Fij i has two bas ic forms of organizat ion . In one, business is heavily capitalized and overseas-owned an d oriented. Labour is employed under union or wages councils award systems and is not drawn from the owners ' or management 's families. 'Modern ' or 'Western ' prin ciples of business, such as cost accounting, stock control and formal credit policies, are employed; management is in close liaison with the overseas owners , and is highly trained and skilled , impersonal and 'distant ' from employees ; inves tment in plant and stock is considerable but so is turnover; relations with governmen t are systematized; taxes are formally applied an d, mo st basically, the mo tivation of business is financial profit. 136

In the other form of organizat ion , labour is intensive , family-based and informally employed and rewarded. Skills are largely tradit ional or self-taught , wo rking capital and investment minimal and cash flow often slight and erratic. Business transactions are largely oral and methods are empirical; inventory control and formal costing procedures are not practised. Links with administrative structures are so tenuous that in come taxes are rarely paid, or requested.

Operational goals are as much local or family service, _ social prestige and employment of family memb ers as they are profit-making.

A similar dichotomized commercial organizat ion has been recognized in other developing countries. The different parts of the system have been termed, among other things, 'upper' an d 'lower' circuit s (Santos 1973; McGee 1973) , 'formal ' and 'informal ' sectors (110 1972) and 'firm-centred' and 'bazaar ' economies (Geertz 1963) . Characteristics of each of the forms of organization as they apply to Fij i are listed in Tab le 3 .1.

While businesses operating with either of the two basic operational forms are recogntzable in Fij i, many fall between the extreme s and have characteristics of each . However, although assignment of businesses to a particular category - or the middle ground between them - is in part arbitrary, and the relationship between the forms not fully understood, the concept does contribute to an understanding of business operations in Fij i. One such example is that the broad intermediate 'circuit ' suggests that expanding lower circuit businesses have difficulty in entering the upper circuit , or that upper circuit businesses are expanding into the lower circuit . From the analysis that follows , it appears in fact that the intermediate zone marks the limit of expansion for family-operated businesses.

A second in stance of the ut ility of the concept is its emphasis on the common features of imported and locally­ processed food distribution on the one hand and of the agri­ cultural product ion-marketing system on the other : the three circuits of connnercial organizat ion have obvious correlations with the subsistence/subsistence-commercial/connnercial cont inuum of agricultural production an d market ing wh ich was discussed in Chapter 2.

In the subsequent analysis of the processed food distri­ but ion system, neither the unique factors affecting the system 137

Tab le 3.1

Forms of conunercial organization

Characteristics Upper circuit Lower circuit

En try Difficult , usually overseas . Easy.

Technology Capital intensive , imported. Labour in tensive .

Capital External , ab undant; from Scarce, local ; personal , banks and other institutions. non-institut ional sources .

Fixed costs Considerable . Negligible .

Government contact Easy access to government Negligible regulation or agencies and aid ; consider­ access to government. ab le regulat ion .

Skills Specialized; from formal Non-specialized; from non­ training (incl. overseas) . formal local sources .

Workforce Non-family; regular; mainly Family; irregular and part­ full time ; award rates; time unpaid help ; wages irre­ recruited by advertisement; gular and non-award, incl . pay­ formally hired. ment in kind ; no advertising.

Hours Strict schedule . Irregular .

Organization and Bureaucrat ic; registered Family . ownership companies ; high degree o'seas ownership.

Business methods Established written proced­ Empirical ; oral; little stock ures; regular audits and monitoring. stock control .

Mo tive Primarily profit . Service (incl. employment) ; social prestige ; profit .

Inventories Extens ive range, high value . Limited , low value.

Source of stock Direct import and local Local importers and wholesalers ; suppliers; access to bulk­ in manufacturing, reuse . purchase discounts.

Prices Fixed. Fixed , but room for negotiation .

Credit policy Formal; little to individuals , In formal ; important for important for big customers. individual customers. Customer relations Formal , distant , indirect . Direct , in formal , social relationships .

Sales techniques Aggressive , media advertising , Passive , no advertising. draw-it ems and sales. Markets Internal and external ; aided Internal , irregular, uncon­ by tariffs, quotas , licences. trolled .

Customers Institut ions (incl . supply Mainly poorer individuals . contracts); upper and middle class individuals .

Links with commer­ Extensive (e.g. Chamb ers of None . cial organizat ions Conunerce me mbership)

Location Primarily urban . Urban and rural .

External links Strong. Negl igible.

Vertical development Extensive , particularly with Minimal , domestic. overseas i�ter��ts. Turnover Can be considerable . Generally limited; low returns relative to time invested .

Source : !LO (1972 :6) and McGee (1973 :7), adapted for Fij i. 138

in Fij i nor the concept of upper an d lower trade circuits are dominant themes . They are raised here , however, to give greater meaning to our analysis .

Government influence

The governmen t has a largely pass ive influence on Fiji's processed food distribut ion syst e m. Apart from a general regulatory function which affects all participants to much the same extent - as , say , policies on import i ng, business licences and employment conditions - government involvement in the grocery trade is limited to a small number of sharply­ def in ed fields where policy is directed to specific ends. The Prices an d Incomes Board, De partment of Co-operatives and the Fij ian Bus in ess Opportunity and Management Advisory Service are the chief bodies with specific involvement in connnerce . The Ministry of Connnerce , Industries and Co­ operatives has little influence on or interest in connnerce, apart from its involvement in the establishment of import­ substitut ion industries and the co-operative movement and its general con cern in consumer protection and foreign participation in the economy . 2 The main areas of government influence on the processed food distribution system are examined below.

Foodstuff imports

The government policy of en couragin g self-sufficiency in food supplies has specific practical consequences for internal and external trade . Influence on the nature and level of foodstuff import s is exercised through both import duties and the issue of import licen ces for particular commodities . Import licences are used to prot ect newly­ established food processing industries. At present they are required for tea, wheat and wheat flour imports. Licences are granted only wh ere demand cannot be met by local production . For instance, early in 1977 when the Dominion 's on e flour mill ceased production for a short time to expand facilit ies , import licences for specific amounts of flour were granted to some bakers and wholesalers . Once the newly-established industries currently protected by import licences ha ve gained a foothold in the local market , it is likely that local production will be protected by import duties alone.

Import duties are used to raise revenue and to encourage local food production . Basic foods that are not produced 139

locally or where local production accounts for on ly a very small share of demand, are either admitted free (for example, solid milk, ghee, spices , tea , wheat flour , cereals, oil seeds and edible oils, margarine, live animals and canned non-luxury fish) or with 5 per cent duty (for example, all me at , uncanned fish , uncanned vegetables an d fruit , fresh milk and cream, cheese, curd, cocoa and non-wheat flour) . Foods for wh ich there is some local production or manufacture and wh ich are protected from import competit ion have 30-40 per cent tariff rates: eggs , honey , peanuts, sugar an d molasses , fowls and ducks, bread , biscuits and cake s are foods so protected . Rates for non-locally produced 'luxury ' foods such as pasta products, jams , jellies , soups, sauces, chocolates and sugar confectionery range between 15 and 40 per cent .

It is probable that import duties will continue to support reasonable attempts to establish local food processing or product ion , and in deed are a flexible and effective means of support . Apart from import dut ies an d licences required for the few protected items , there is minimal restriction of en try into the import ing trade .

In addition to support through import tariff and licence structures, local food processing is assisted in other specific ways . Industries that result in import-subst itut ion and use considerable local labour an d local raw materials are particularly favoured. Limitations may be placed on impo rts of competitive product s; in addition tax and investment in centives , guaranteed markets through government supply preferences an d monopoly production rights for specified periods may be offered . New industries established un der such condit ions to date have been chiefly of non-food products. Among foodstuffs, only tea packing an d flour milling have had substantial direct governmental support : some others , e.g. , biscuit manufacture , are supported indirectly through tariff structures while others , such as meat cann in g, receive no apparent direct support .

Business licence and employment regulations

Business licence and employment regulat ions are an other in fluence on the processed food distribut ion system.

Any business activity in Fij i is liable to an annual licence fee payable to a city or town council , township 140 board or the central goveniment , depending on the location of business. Fees for retail shops are in the range $20- $80pa and for wholesale-retail establishments, $75-$300. Licence fees within one area of jurisdiction are uniform for a particular funct ion : rates vary in response to the assumed prosperity of areas of jurisdiction . Such licensing is not to be confused with registration of a company under the Companies Ordinance, which is uncommon among part icipants in gro cery distribut ion .

The cost of retail licences does not appear to discourage entry into retailing. Moreover, methods of policing are sufficiently slack to enable small periodic retailers , especially those operating from their homes in rural and peri-urban areas , to avoid paying licence fees . In contrast, the differential between retail and wholesale-retail licen ce fees - there is no 'wholesale' category - is a disincentive for merchants with small, an d some not-so-small , wholesale businesses to purchase the appropriate licence.

A considerable body of legislat ion deals with the regu­ lat ion of employment but much of this has little practical effect for many of the businesses involved in the gro cery trade in Fij i, either because it applies only to businesses in urban areas or to wage-employees. The making of legis­ lat ion contingent upon wage-employment , and in some cases town-locat ion , emphasizes the dichotomy in business organ­ izat ion between family and non-family bus inesses. Under the Shop (Regulation of Hours and Employment) Ordinance and the Wages Councils Ordinance, for instance, conditions of employment and wage rates are set for the wholesale and retail trade . Such regulations , however, only apply to employment 'for hire or reward ', which excludes employment in most family-businesses, an d in gazetted cities, towns and townships. Other legislat ion of potential significance to wholesale and retail trade employees, such as that governing workmen 's compensat ion and the Fij i National Provident F1llld, is also of little consequence as it applies only to wage employment. Even basic employment legislat ion like the Employment Ordinance and Factories Ordinance has only an indirect relevance to the maj ority of retail and wholesale workers. In towns the hours of operation for business may be limited by local goveniment , but beyond this the in fluence of goveniment regul at ion is hardly felt by businesses without wage labour . 141

Department of Co-operatives

Within the government, the Department of Co-operat ive s of the Ministry of Connnerce, In dustry and Co-operatives has the most sustained concern in the retail and wh olesale trade . Apart from being important because they represent the most direct involvement of the government in commerce , co-operatives are sign ificant as they are seen as the chief official mean s of involving Fij ians in connnerce.

Co-operat ive societies have operated in Fij i under the Co-operative Societies Ordinance since 1947, having been established an d maintained as the spearhead of Fij ian economic advance. 3 In 1974 , the most recent year for which trading results are available, there were 1095 societies with approx­ imately 35, 000 members , most of whom were Fij ian . Eight out of ten co-operat ive societies are consumer or consume r­ marketing societies that operate a store selling groceries and basic consumer goods. Turnover of the stores in 1974 was $8. 76 million and of marketing activities (mainly copra) , $3.24 million . 4 In addit ion , there are some 160 thrift and credit societies and six farmers' input supply and market ing societies . The recurrin g budget of the Department of Co­ operatives in 1977 was $606 ,000.

The retail-wholesale facilities of the Department of Co-operatives comprise a three-tier structure . Primary society stores are theoreti cally supplied by a numb er of regional wh olesale associat ions that in turn draw supplies exclus ively from the apex organizat ion , the Suva-based Fij i Co-operative As sociation Ltd.

There are some twenty regional wholesale associations (Table 3.2). Although it is not anticipated that the number of wholesale associations will increase significantly , some of those in the islands will be developed into central loading points with basic port an d warehouse facilities to stimul ate two-way trade. The Fij i Co-operative Association is being encouraged and assisted to expand and improve services to consumer societies. Whereas primary societ ies are entirely self-supporting, regional associations receive government grants to help me et both capital , e.g . , storage facilities , and recurring expenses , in cluding management costs. For the period 1976-80 , $935 , 000 has been reserved for these ends . All societies receive , without charge , con siderable adminis­ trative assistance , including audit ing of trading records Tab le 3.2 ...... i::-. N Trade act ivity of regional co-operative wholesale associations

Wholesale tradins All trading activit ies Ave rage Average As sociation Gross profit Net profit Gross prof it Net profit Member Account monthly mon thly as % sales as % sales as % sales as % sales societies period sales ($) sales ( $)

Western Division Ba 11, 406 8.0 3. 1 n.a. n.a. n.a. 72 1/6/74-31/3/75 Nadroga/Navosa 5,914 8.3 1.2 n.a. n.a. n.a. 68 16/2/ 75-31/12/75 Vat ulele 3, 717 5.0 0. 0 8, 671 12 .3 o. o n.a. 17/8/75-13/8/76 Ra 5,083 3.9 0. 0 6,600 11. 9 1. 5 36 1/9/74-30/9/75 Central Division Wainimala 5,941 -1. 6 -7 .2 5, 964 -1.2 -8.4 n.a. 1/9/75-30/11/76 Central Fij ian 10 ,917 4.0 -5.2 11 , 186 6.3 -8 .8 n.a. 1/8/75-31/7 /76 Navua Valley 440 9.8 -16.5 n.a. n.a. n.a. n.a. 11/76-10/77 Eastern Division Ovalau-Moturiki 11 , 703 5.9 1. 4 12 ,199 6.5 1. 7 n.a. 1975-76? Koro 8,521 5.7 2.1 n.a. n.a. n.a. n.a. 31/5/76-13/1/77 Gau 10 , 001 9.6 6.3 10, 196 11 . 3 4. 3 16 13/3/75-31/5/76 Kadavu 8,969 8.8 6.2 n.a. n.a. n.a. 57 7 /6/75-30/4/76 Lakeba 12 , 920 13.2 8.1 n.a. n.a. n.a. n.a. 1/9/74-31/8/75 Vanua Balevu 6,348 10.4 4.6 6,869 17 .2 3.2 n.a. 23/1/76-25/1/77 Cicia 2,748 o.o -30.8 7,886 36.2 11 .3 n.a. l/6/7 5-31/ 5/76 Northern Division Bua n.a. 4.7 1. 3 n.a. 6.4 2.4 n.a. 1975-76? Dogotuki 2,047 9.1 5.5 12 ,294 9.1 6. 7 n.a. 1/2/77-30/4/77

Note: Data not available for Macuata, Cakaudrove , Beqa and Moala As sociat ions .

Source : Department of Co-operatives, Suva. 143 and business training classes for society personnel at the Department of Co-operatives' trainin g centre near S�va .

In addition to the advice , training and audit facilities noted above , co-operative societies have further advantages over other retail and wh olesale outlets. They have , for instance , a five-year tax holiday from the date of regis­ trat ion , and after that time only their reserves arc liable to taxation . Bonus payments to members are liab le to taxation as part of the member's personal income , but because the income of memb ers is often below the min imum rate for in come tax , the payments are usually untaxed . Each year , 25 per cent of a society 's net profit must be put in a reserve account which is designed as a source of funds for capital imp rovements for the society . In general practice, however, maximum allowable bonuses, i.e. , 75 per cent of net profits ) are distributed and there is little subsequent developmen t of capital assets.

Consumer co-operative societ ies are established both to provide a locally-controlled constant source of consume r goods in a commun ity and to be a commercial training ground for Fij ians. However, few develop beyond this service function to become more dynamic trading ventures. Additionally , the minimum cap ital generated by the societies makes develop­ ment contingent upon considerable financial inputs, even for recurring costs, from the Department of Co-operatives. The consumer and consumer-marketing societies have generally succeeded in satisfying the limited local demand for basic consumer goods, but have not been so successful in their role as a spearhead of Fij ian economic advance.

Prices and Incomes Board

Government policy relating to foodstuff distribution , which in cludes locally produced, non-pro cessed food, is in part aimed at reducing the numb er of int ermediaries between local or overseas producer an d consumer . Two mechanisms operate to this end: minimum administrative restrictions on importing, and price controls set and enforced by the Prices and Incomes Board. The PIB of the Ministry of Finance has become an important means of government intervention in the structure of the grocery trade in Fij i. The first mechanism is examined in the section 'Importing' below and the PIB's price controls here. 144

On the twelve essential commodit ies under price control (the controls do not apply to products manufactured within Fiji), there is a three-stage level of allowable price markup (Tab le 3.3 ). Import and wholesale markups on food are in the range of 1-5 per cent of landed and into-store cost; retail markups are in general from 10 to 12 per cent. No trader can claim wholesale and retail markups on the same item. The type of markup taken depends on the type of sale : as defined in the Counter-Inflation Act 1973, 'whole­ sale ' is sale for the purpose of resale or use in trade or business and 'retail ' is sale to a consumer. As in tended, the permissible wholesale markup does not leave room for a second wholesaler. The impo rt percentage can be charged by an porterim whether the goods are sold retail or wholesale .

The PIB is rmequivocal ab out the intention of this markup structure :

The purpose of this provision [allowable markups for imported items ] is to allow only a nominal markup to any secondary whole saler who handles the goods once they have been imported into Fij i, and to allow an in creased retail markup to a retailer who has himself imported the goods. Retailers in rural areas will need to arrange their buying direct from wholesaler/importers in order to prevent their in to-store costs being inflated by more than one wholesale markup . The co-operative movement is obviously provided with a tremendous incentive to prove its worth in channeling goods to outlying co-operatives at the lowest possible prices (PIB n.d. :26) .

The PIB's in tention to eliminate secondary wholesalers , who are known locally as 're-wholesalers ', has been successful . Only where there is limited wholesaler access to products such as sugar and some canned mackerel are items regularly re-wholesaled between importer or local manufacturer and consumer .

Wh ile price controls have achieved the obj ective of limitin g re-wholesaling, there has not been equal success with a second obj ective - to counter apparent monopolistic conditions in trade (Fij i, Central Planning Office 1975 : 163) . In fact , one consequence of the price markup structure has been to strengthen the position of larger importers and wholesalers. Many smaller wholesalers have ceased dealing 145

Table 3.3 Import , wholesale and retail markups on price­ controlled items (under Order of 1 Ap ril 1975)

Import Wholesale Re tail markup (%) markup (%) markup (%)

Bread Nil Nil 11 Butter 3 2 7 Edible oils 5 2 12 Japanese mackerel 5 2.5 12 Other fish 5 2 12 Flour ( local ) Nil 5 10 Flour ( imported) 4 1 5 Light ing kerosene 20 1 11 Margarine 5 2 13 Corned beef and mut ton 3 2 10 Baby milk 5 2 13 Other milk 5 2 10 Rice ( imported brown ) Nil 5 10 Rice ( imported white) 4 1 5 Sharps (local ) Nil 5 10 Sharps ( imported) 4 1 5 Tea ( local ) Nil 4 10 Tea ( imported) 3 2 10

Note: Percentage markups are on the in to-store cost of goods for importer , wholesaler and retailer.

Source : Prices and Incomes Board , n.d . :27. 146 in commodities under price control, particularly heavier an d bulkier items like flour, sharps an d rice as they claim the allowable markup does not cover transport , storage and handling costs. The who lesalers continuing to deal in these items are generally larger operators , who claim they handle these particular goods only as a service to customers and to en sure patronage of other wholesale lines. In sum, price markup structures have shortened the distribu tion chain , but at the cost of increased specialization and fewer outlets.

Banking and managemen t services

The Fij i Development Bank (FDB) and the Fij ian Business Opportunity and Managemen t Advisory Service run programs to help Fij ians in connnerce . FDB was established in 1966 as a government-owned financing institution . The obj ective was to promote the Dominion 's economic developmen t by providing financial assistance to enterprises operating in Fiji. Loans were to be provided to assist the expansion or modernization of proj ects in agriculture , fishery , manufacturing and pro­ cessing industries, transport and communication , tourism, mining and construction industries. The role of agriculture was emphasized in the establishing ordinance. On 31 June 1976, $9.2 million in loans was outstanding, of which slightly more than $250,000 was in connnerce.

The Bank did not lend for activities in commerce un til May 1975, when the Commercial and Industrial Loans to Fij ians Scheme was started 'to give tho se of Fij ian race a better opportunity of acquiring a share in Fiji's commercial sector ' through the provision of 'medium and long-term loans for any useful purpose', including the provision of working capital on 'special terms ' (Fij i Development Bank n.d. :1) . To June 1976, $997, 000 in 130 loans had been lent un der the scheme , although many of the loans were in transport and industrial fields that may have been approved without the scheme. Interest rates and repayment schedules under the sclfeme are more favourable and attitudes to security requirements more flexible than for other FDB loans. Fields to benefit from the scheme are tran sport (trucks , taxis, boats ), shops an d existing manufacturing companies. Actual investment in shops has comprised a very small share of investment . Mo reover, as the bank admits, adequate advice and supervision of borrowers has not been available (Fij i Developmen t Bank 1976 :8) ; thus the success of the scheme is unknown . The Fij ian Business Opportunity and Managemen t Advisory Service works closely with the Developmen t Bank 's Connnercial and In dustrial Loans to Fij ians Scheme . It is a un it estab­ lished in 197 5 within the Ministry of Fij ian Af fairs to assist Fij ian en try into commerce and industry . Fij ian connnercial involvemen t is en couraged in three ways : (a) identifying opportunit ies for Fij ian investmen t in industry an d connnerce an d, wh ere practicable , ensuring Fij ian involvemen t in training an d manage­ ment in the field of investmen t; (b ) offering assistan ce in the form of locating loans sources , undertaking marke t research and compiling accounts and business records ; (c) providing bas ic bus iness training.

The Advisory Service conducts preliminary evaluations of loan applicat ions to the FDB un der the Commercial and In dustrial Loans to Fij ians Scheme , in addition to other proj ects it becomes aware of independently. A Fij ian loan proposal needs support from the Service to be considered by the bank for a loan . Achievements of the Advisory Service have been limited and piecemeal . The reasons for short­ comings appear to be a lack of suitable personnel to identify and work with potential en trepreneurs and to organize effective training courses . 5

Both the Fij i Development Bank and the Fij ian Business Opportunity and Management Advisory Service have had slight effect on commerce an d negligible consequences for the increased involvement of Fij ians in the grocery trade . There are no programs that specifically encourage or assist the en try of other races in to commerce .

Foodstuff importing into Fij i is characterized by ease­ of-entry, min imal governmen t in tervent ion , entrepreneurial in itiative , and distinctive upper/lower circuit and racial roles . A large number of importers and in dent agents dealing in small individual transactions are juxtaposed with a small number of larger firms , some with their own inden t departments , which account fo r a considerable share of food imports.

Mo st importing is by wholesalers who sell directly to retailers , or by retailers themselves. Apart from potatoes , 148 onions, garlic an d a couple of brands handled by exclusive distributors , there is negligible re-wholesaling of imported foodlines; the re-wholesaling of locally-processed products is equally unconnnon . The structure of permissible markups on items under price control, the ease of entry in to importillg, the relatively con fined region within which distri­ but ion takes place, the open trade - as indicated by the preponderance of 'connnon lines ' - and the competitive nature of the grocery trade all discourage additional marketing stages .

A key figure in im porting is the 'indent agent ', who is also known in Fij i as a 'commiss ion agent' or 'man ufac­ turer 's representative '. The fllll ction of the indent agent is to place orders from local me rchants with overseas suppliers (Fig. 15) . The sign ificance of the indent agen t to the importing trade is suggested by the fact that few suppliers accept orders save through indent agents, an d often on ly agents reconunended by conunercial banks in Fij i.

Indent agents obtain business with a greater or lesser degree of active pursuit of potential customers : after a secretary, the most common em ployee of an in dent agent is an order-taker who solicits business. Under normal circum­ stances, the indent agent does not take possession of, or assume legal liability for, goods ordered through him. Goods ordered via an indent agent are forwarded in the name of the client, who pays by a sight-draft or letter of credit as the goods are received in Fij i. No deposit is paid at the time of placing an order. Payment is made only at the time of actual acceptance of goods . If goods are not accepted by the merchant. who placed the order or paymen t is not forth­ coming for any reason , the indent agent has a moral , but not legal, responsibility to find an alternative buyer. Rej ected deliveries reflect poorly on the indent agent and can lead to a break with the affected supplier, so it is in the interest of the indent agent to assess carefully his clients' credit . The agent receives from the supplier a connnission of up to about 5 per cent of the value of orders. Im porters do not pay indent agents for their services.

In dent agents are among the most speculative and entre­ preneurial participants in foodstuff distribution in Fi j i. . The chief requiremen ts for en try into indenting are familiarity with local demands and overseas supply sources, a certain amollllt of aggressiveness to win orders and resilience to survive in a very competit i ve field. Not only is competition FIJI OVERSEAS

Letter of credit.

_____...... • · •• • •••• •••••• • .. •� .____s�i g�h _t _d r_aft ...... • _

------7 ------1 - .____c _us�rn__ Ms_ __, : I 1- • I �--�---� ------·-·

Island traider/ ...... , ••• Buying house lI Wholesaler, I I Wholesaler-Retailer, Indent Agent or Retailer (Commission Agent)

(Manufacturer's Rep.) · ·- - · - Island Trader, 'Agent'

------� �--____y-�· ·-··-· -··-· -··-· -··-·-··-··_··· _-· -··_· _· ·_·_·-· ··_· _·_· ·_·_··_· _· ·_· _··_·_··_··· _-·+· · _· _··_· _· ·_·_· _· ·-·�1 Companfs Bu�ng House Wholesale Compa n or Agent Retailer Indent Office Company's CUSTOMS Customs Agent order payment goods comm1ss1on ...,__ direction of transfer

Figure 15 Grocery im porting channels 150 between agents in tense, but it can be as difficult for an agent to receive connnissions from a supplier as it is for him to ensure that traders placing orders have the means to make payment.

The only formal en try requiremen t for indenting is pay­ men t of an annual business licence fee. 6 Capital investmen t is slight . A telephon e and a typewriter are the basic equipment requirements. With groceries, no display stock is required although it can be helpful. Training comes largely from experience working for other indentors. Some agents operate from their homes and have no employees , but most have a rented office and two or three employees. Mo re prosperous firms have a telex machine, a business vehicle and a larger , more specialized and frequently non-family staff. It is indicative of the in security of indenting that thirteen of sixty-three in dent agents listed in the 1976-77 Suva telephone directory were uncontactable an d apparently not operating in March 1977.

Indent agents are located chiefly in Suva and Lautoka , where there are thirteen , ten of whom have either a branch or head office in Suva. 7 On ly 22 agents in Suva are indent agents in the strict sense of people who merely place orders on behalf of importers . There are an other twenty-four agents who describe themselves as 'indentors ' but who occas ionally import goods with their own resources. These goods are generally for sale through their own wholesale-retail outlets. Product specializat ion occurs among in dent agents but not so much in foodstuffs as in other goods . Four of forty-six 'indentors ' in Suva deal exclus ively with foodstuffs , twenty­ one deal with non-food items on ly and twenty-one with both food and non-foodstuffs. Agents often say they will 'deal in anything that makes a profit ', a proviso that apparently limits exclus ive involvement in food. Of the twen ty-five agents in Suva who indent foodstuffs, six or so appear to handle the bulk of the trade .

In terms of number of agencies , in denting is dominated by locally-owned in denting firms . These locally-owned firms are almost always operated by the owner an d are licensed bus inesses , although not registered companies . Of the forty­ three agencies in Suva for wh ich the informat ion is available, thirty of the owne rs are In dians, in cluding nineteen Guj eratis , and one is Chinese . Twelve inden t agencies are European firms ; only two of these are locally-owned. An equal number of locally- and overseas-owned agencies have a second branch 151

office (in either Lautoka or Suva none has mo re than two offices) : half the overseas-owned firms have bran ches , compared to one-in-eight local agencies .

The ownership of the European indent agencies is in teresting for what it suggests of the sign ifican ce of vertical linkages in the food supply sy stem. At least six of the European firms are owned by established island trader companies in New Zealand and Australia, in cluding one agency that is a wholly-owned subsidiary of Burns Philp . The remainder are owned by overseas manufacturers of non-food products. The staff of the overseas-owned agencies is gener­ ally all local. A consequence of overseas ownership on indent agencies is that orders for goods from the home country of the proprietary firm must usually be placed via the paren t company.

Although mo st locally-owned indent agencies consist of lone independen t operators , some do have formal links with other local firms . Among agents who in dent foodstuffs, operation of a wholesale-retail outlet by other family members is the mo st connnon form of linkage. At least seven of the owners of locally-owned indent agencies own or share ownership of other firms , and it is likely that there are other such linkages.

The ab sence of restrictions on foodstuff importing , the aggressive business habits of indent agents and the ease of en try into inden ting all contribute to con siderable compe­ tit ion in foodstuff importing . A number of factors , however , counteract potential instability in the trade. One is the back-order character of food imports. Whether for lack of capital , predictable and limited demand for foodstuffs or other reasons, foodstuffs are imported primarily in response to established demand. Most retailers and wholesale-retailers who import do so as stocks require replenishing an d then they only order goods for which they have assured outlets. They ra,ely import untried products. In the past year or so, a couple of firms in Suva have taken advantage of container freight facilities and have begun importing on speculation foodstuffs and general household items . Their impact on the food trade has been slight . Inden t agents frequently import new non-food lines or 'good deals ' on established lines at their own expense and subsequently wholesale the goods - but this rarely occurs with foodstuffs. 152

Another conservative force in foodstuff imports is exclusive brand representation by in dent agents , and to a lesser extent by importers, here called 'distributors '. Al though it is an ambition of most indent agents to gain exclusive representation rights to a product, exclusive agencies or franchise rights are far less common among food­ stuffs than other goods , and anyway do not apply to most imported staples. Food brands that are 'exclusive ' to indent agents include Kraft , Quaker Oats, Sunshine (milk) , SPC (canned fruit ), Distillers Co . (yeast), Lion (baking powder) , Golden Circle (canned fruit), Cadbury (cocoa) , Zorn an d some brands of canned fish . Orders for products with a residen t representative are accepted only if they come through that agent. Mo st of the exclusive representation rights are held by overseas-owned indent agencies , and mo st overseas­ owned agencies dealing in foodstuffs have some exclusive brand rights.8

Access to particular brands may be restricted by means other than assigning exclusive rights to an indent agent. Some manufacturers accept orders only through appointed agents in the producing country. Products of the New Zealand Dairy Board , such as butter and An chor milk, for instance , are exported only via official Dairy Board agents in New Zealand. Wh ile some manufacturers appoint in dent agents to handle orders for their products, others appoint an importer/ wholesaler as exclus ive distributor. This practice has , however , practically ceased in food-lines. The one maj or exception is Ocean mackerel, wh ich is imported only by the Carpenter group . Another related form of control occurs where a supplier agrees to do business on ly with particular firms . It has been suggested (but not substantiated) , for in stance, that a maj or Australian source of rice has agreed to supply only three mills in Fiji. Wh ile exclusive repres­ en tation can be justified economically by supplier , agent and purchaser , it can create monopolistic trade practices . On the whole , however , such access restriction has a far smaller role in the trade of foodstuffs than other products .

In conclusion , three general points can be made about processed food importing. First, the system is predominantly

a back-order supply system characterized by the underdevelop­ ment of warehousing and few importer-distributors (rather than importer-wholesalers). This structure contributes to the periodic short-supply of many connnodities. Second, the influence of overseas links in the in ternal marketin g system is readily apparent. Not only do the links orientate trade 153 to particul ar countries , bu t even the source of supply within overseas countries is often determined by the dist ribution policies of suppliers and overseas agents. Mo reover , over­ seas-own ed indenting firms appear to handle a disproportion·· ately large share of trade. The dominance may be due partly to better overseas links 1 also important is the leve l of capitalization , familiarity with overseas business me thods and exclusive rights to some key products.

Finally, of all levels of trade in processed foods , importing operates with least government involvement . On the one hand , this may be justified by the belief that minimal restriction will en courage compet ition among importers , which will ultimately benefit consumers . On the other hand, the policy is perhaps questionab le in the light of the critical consequences of importing costs on in ternal price levels.

Wholesaling

There are two basic forms of processed food wholesaling in Fij i, the wholesaler-retailer and the non-retailing whole­ saler. By estima ted share-of-trade and numb er of outlets , wholesaler-retailers are the more important.

Wholesaler-retailers import or purchase from local food processors the items they wholesale ; apart from wh olesalers in isolated locations who may re-wholesale, the chief re­ wholesaling organ izat ion is the Fij i Co-operative Association Ltd . Excluding the FCA, wholesaler- retailers are of two basic types : the overseas-owned general trading and commercial corporation which has some involvement in the grocery trade, e.g . Mo rris Hedstrom Ltd an d Burns Philp (South Sea) Co . Lt d7 and the local In dian or Chinese family-owned and operated concern that has usually developed from a family retail business. The scale and longevity of the latter varies . It is not unusual for such businesses to withdraw from importing and wholesaling to concentrate on retailing during economically depressed periods; for some , wholesaling has an in significant role in their total small trade , but for others it can account for up to 90 per cent of estimated annual gross turnover of up up to $1.5 million , around three-quarters of wh ich is in groceries.

Non-retailing wholesalers are almost exclusively food processors/manufacturers that wholesale their products to 154 distributors or retailers. Nine main food products (flour , sharps, sugar, dry biscuits, oils , sugar , rice , tea and dairy produce) are handled largely by such wholesalers .

Before analys ing the operations of these forms of wholesaler, two points about wholesaling in Fij i need to be kept in mind . First , while they do have strong links with retailing and importing, wholesalers rarely expand into manufacturing and inden ting. Only one case came to notice during our work in Fij i where a locally-own ed retailer­ wholesaler also operated as an indentor; only one in stance of a wholesaler-retailer involved in manufacturing was noted.

Second , the processed foods wholesaling system in Fij i is characterized by both relative rigidity and flexibility. The rigidity in the system comes in large part from government price control measures that discourage wholesaling inter­ mediaries . Some consequences of the PIB for the food distribution system have been examined above .

On the other hand, the ab sence of sign ificant entry restrict ions to foodstuff importing and wholesaling contributes to an extremely flexible and competitive wholesale environment. Competition is primarily through price, with service , in particular delivery service., a second form of incent ive . Competition is such that in areas wh ere it is most intense, for in stance the Suva-Nausori region , where there are over -· 120 companies supplying retailers with groceries, 9 it is not llllconnnon for wholesaler-retailers to operate below cost on key items in order to capture a share of the trade. On items that are not llllder price control , markups on a wholesale price are arolllld 15 per cent over landed cost ; retail markups are usually a further 10 per cent. A commonly reported net pro fit rate for general wholesaling in Fij i is 10 per cent, but it is likely that the net prof it on groceries is less than this . 10

It is indicative of the degree of competition among wholesalers that at least five of the fifteen or so main wholesaler-retailers operating out of Suva either ceased or sharply curta iled their wholesale activities over 1976-77. While compet it ion may help maintain internal prices at a low level, it contributes to supply inefficiencies by encouraging tmder-capitalizat ion , limited sto cks and periodic shortages. Competit ion has not changed the basic fllll ction of the whole­ saler, the forwarding of supplies between producer and 155

re tailer , although the responsibility of actual transfer falls in creasingly on the retailer. Practically no products are repacked by wholesalers , the retailer breaking bulk from the man ufacturer's sale unit to that purchased by the consumer.

Wholesaler-retailers

The re-wholesaler. Re-wholesalers or 'jobbers ' are not common in Fij i, both because the limited allowab le markups on price controlled items discourage intermediary trade links and because of the limited distances between importer/manu­ facturer and retailer; mo st areas with a considerable grocery deman d are within a couple of hours ' road journey from the main supply centres. Nevertheless, they are an interesting and significant part of the processed foods distribut ion system, not least because although the governmen t discourages re-wholesaling (for example, through the PIB) , its mo st extensive development has occurred un der spon sorship of the Department of Co-operatives in the form of co-operative wholesale associations.

The re-wholesaler purchases supplies from manufacturers in Fij i or, where goods are produced overseas , from importers. The re-wholesaler, as distinct from the wholesaler an d importer who may re-wholesale some goods , does not import . Re-wholesaling is the least common form of wholesaling an d occurs primarily in areas wh ere importing is discouraged by limited demand and difficult connnun ication s with Suva or Lautoka , the chief ports of entry . Moreover , limited and seasonal trade often means the re-wholesaler is without the financial resources to pay for imports upon receipt , wh ich is the standard practice in Fij i.

Some wholesalers in the smaller towns of Viti Levu (like Ko rovou) , and most wholesaler-retailers elsewhere do in fact re-wholesale some products. More commonly, re-wholesalers operate out of isolated shops in rural areas when ce wholesale activities are directed to even mo re remote stores . Many shops in strategic rural locations, such as breaks-of-transport , function as wholesale outlets.

The re-wholesaler is invariably an Indian or Chinese trader. His store rarely appears larger or more prosperous than most In dian or Chinese rural general stores. There is little structural or functional division of wholesale and retail facilit ies, apart perhaps from more exten s ive storage 156 facilities than in retail outlets. Goods , however , are just as likely to be stocked in the shop area as in specialized storage space. Stock consists of grocery and general house­ hold items purchased from urban-based wholesalers in bulk - such as 100 or 224 lb sacks of flour , rice or sugar , and 48 or 72 unit cartons of canned goo ds. Stock is usually re­ wholesaled in smaller units but is not repacked. Half-sack me asures or half-dozen lots of items like packets of tea or canned goods are connnon re-wholesale units . Re-wholesalers are generally pass ive in their wholesale activities : they do not solicit trade or deliver purchases.

Mo st re-wholesalers do far mo re business as retailers than as wholesalers because re-wholesalers are primarily emergen cy supply sources. When urgent supplies run short in retail stores , or when the retailer has insufficient ready cash, or too poor a credit rating, to warran t a trip to a mo re distant supply source where goods are cheaper and available in a greater ran ge and quantity, the re-wholesaler is used.

The re-wholesaler has only a small margin in which to establish a re-wholesale price as the grocery lines that are the bas is of his trade are under price control . In fact , sales to retailers are often made at retail rates, although there may be some discount if payment is in cash. Even where a retailer is supplied at a special re-wholesale price , the resultant retail price is usually higher than for goods pur­ chased directly from a wholesaler. The re-wholesaler survives in the face of these pricing disadvantages because of his greater accessibility for some retailers , lower direct tran s­ port delivery costs to the retailer and at times by offering more liberal credit than wholesalers. Mo reover, as mo st rural retail stores have relatively static patronage and serve customers in semi-monopolistic conditions , the higher prices necessitated through servicing by re-wholesalers can be passed on to consumers without the consequences of losing cus tom that would be expected in more competitive circumstances .

The number and location of re-wholesalers cannot be readily determin ed from business licen ce records . Since for most re-wholesalers, whol esaling is erratic and largely supplemen tary to retailing, the additional cost of a whole­ sale-retail licence may not be warranted; and re-wholesalers located in rural areas can ofte n operate undetected by authorities . However, a suggestion of the extent of this 157

fo rm of tradin g became apparent during a survey of licensed retail outlets in the Central Division : of 352 'retail ' stores surveyed in urban an d rural areas , twen ty­ two supplied other stores.

Re-wholesaling is mo st fully developed in the co-operative movement. As the middle level of the co-operative con sumer society organizat ion , regional wholesale associat ions have been established largely since 1972 to supply goods at wholesale prices to member primary societies . The regional associations are supplied by the Suva-based apex organizat ion , the Fij i Co-operative As sociation Ltd.11 The co-operative movement has become involved in wholesaling in order to provide a more efficient service to member societ ies and to enable it to retain a share of trade that was go ing to sup­ pliers outside the movement . It was thought there would be greater efficiency if the apex organ izat ion operated as a central buying house for societ ies , and the associations operated as regional warehouses . These would meet society demands more efficiently , and be ab le to pass on bulk purchase discounts. The present regional associations can potentially serve mo st primary soc ieties and it is not envisaged that the number will increase great ly . The associations receive government grants for capital costs of shop and storage development and to subsidize management expenses.

The co-operative wholesale system has not been successful financially (Table 3.2) . Except perhaps in the mo st isolated island areas , society loyalty to co-operative wholesale suppliers is weak. Re gional wholesale associat ions have had least success on Viti Levu. Over the most recent accounts period, three of seven associations there including one on nearby Vatulele island made a net loss on wholesale trading; two broke even and two made a net profit .

The experience of the Nadroga-Navo sa Association suggests the difficulties facing wholesale associat ions on Viti Levu. The As sociat ion ceased operat ion s in January 1977 after two years activity during which the numb er of purchasing societies fell from over seventy to less than ten . The Association charged for delive-ry and in sisted on · part-paymen t at deliver¥. In contrast) other wholesalers in Sigatoka where the As sociat ion purchased stock for re-wholesale, imported or purchased directly from local food processors and could offer lower wholesale prices than the Association . Other wholesalers could also use free deliveries and a flexible credit policy to attract custom. In addition to this basic 158 supply problem, the Associat ion was sometimes poorly stocked when loans from the Departmen t of Co-operat ives to purchase stock were delayed.

Where societies do not have ready access to alternative suppliers , as in Lau and Lomaiviti, wholesale association s have been more successful . Only one of seven associations in the Eastern Division had a net trading loss over the mo st recent accounts period. Co-operative 'disloyalty', however, is apparent in those areas also in that regional associat ions connnonly purchase stock from importers an d wholesalers other than the FCA.

Society and associat ion 'disloyalty' can be attributed in part to economic connnon sense. Supplies for societies at regional associations, and even at the FCA are frequently more expensive than elsewhere owing to re-wholesaling markups, poor buys and bad pricing practices. Because of this , regional associations cannot expect the custom of societies that have access to direct importer-wholesalers . Most societies on road links to Labasa or the main towns of Viti Levu are in this category, as are societies in the Yasawas and the Lomaiviti and Lau islands, wh ich have regular shipping links with Suva.

Compounding the diseconomies of supply by re-wholesalers are questions of the efficiency of the co-operative wh olesaling organizat ions. The FCA has a reputation of being an expensive and unreliab le grocery wholesaler. Trading losses are countered by across-the-board price increases that only result in further diminut ion of trade an d decreas.ed profit ·and con sequent addit ional price hikes . 12

Management of the FCA has recently been taken in hand by an overseas aid team and reorganization is 1IDder way to curb the downward trend in sales volume and profit rates. Over the past two years , sales have decl ined by 14.5 per cent, although grocery sales fell by a little less than this . For the year ending 30 Septemb er 1976, total sales were slightly more than $1 million , just under half of which was in groceries; fuel sales comprise an equal share of trade . Of grocery sales, 77 per cent were wholesale transact ions made at a net loss of 17.5 per cent : only one department of the Association , a retail outlet catering primarily to FCA staff in Suva, made a net profit . 159

The FCA imports mo st overseas supplies direct, the orders usually going from its indent section to a 'buy ing house' in New Zealand that acts as agent and consolidates orders. Stocks produced locally are purchased direct from manufacturers. Orders from member associat ions are proces sed in Suva . By supplying only regional associat ions rather than primary societies, it should be possible to overcome much of the problem of bad credit to societies ; credit accounts of a small number of high-volume purchasers are easy to monitor.

In theory , supply runs link regional wholesale associ­ ations with the Suva FCA headquarters about every two weeks . In fact, supplies are very irregular on both the sea runs to Kadavu , Lomaiviti, Lau and Taveuni, Vanua Levu and Ro tuma and to societies on the circum-Viti Levu road . The supply run that operates with most regularity is the liquor wholesaler 's route around Viti Levu. The Association ships supplies on private vessels or on government ship s at commercial rates. Links between regional wholesale associ­ ations and societies on more isolated islands are even more tenuous than between the FCA and regional associations. It is not unusual for some islands to be out of foodstuffs other than that produced on the island, for weeks and even months.

The declared obj ective of the FCA is a monopoly of wholesale supply to co-operative societies in at least the Eastern Division and Cakaudrove . It remains to be seen whether the advantages from centralized importing and distri­ but ion are sufficien tly attractive to win society and associ­ ation custom. The activities of the FCA provide an interesting in stance of an effort to structure a supply system more to meet service obj ectives than to operate within parameters appropriate to the physical and commercial environment.

Importing wholesaler-retailers . The mo st connnon type of wholesaler-retailer is the trader who buys locally-processed foods, such as sugar , flour , sharps, rice, tea and dry biscuits from a manufacturer and who imports other stock directly . Price control regulations encourage merchants to import directly ; in addition to wholesalers , importing is also done by some larger and specialized retailers , such as supermarkets, Indian and Chinese food specialists and the few fresh fruit and vegetable specialists. Although the require­ ment of payment upon delivery for imported goods means there is a relatively small number of consistent and established 160

importers , there are also numerous speculative seasonal impo rters. Canned tomatoes from Taiwan , for instance , have recently been widely imported by entrepreneurs ranging from small general retailers to duty-free goods merchants. Canned fish is another speculat ive import field. Most importers , includin g the transitory , seasonal operators , retail as well as wholesale .

Two basic forms of the import ing wholesaler-retailer are identifiable. One is the 'integrated wholesaler-retailer ', the merchant who wholesales and retails without structural or functional distinction between these activities. An office and a bulk store for wholesale trade may be set aside from the retail shop , but separate wholesale and retail accounts are not maintained , a connnon staff is employed, and stock is sold wholesale or retail .

Other wholesaler-retailers maintain distinct wholesale and retail functions, facilities , supplies, staff and accounts; examples are Morris Hedstrom and Burns Philp , although in some of their branches the separation of function is not maintained. The operational separation of wholesale and retail activities may be a product of the volume qf trade , but the division also coincides with racial differences in ownership and methods of business operation .

Wholesaler-retailers who import and purchase locally are urban-based and concentrated in the larger towns. In the Central Division , for instance , there are approximately twenty-five such wholesaler-retailers in Suva, ten in Nausori, three in Navua , one in Korovou and only two or three are in rural areas . Similar urban/rural and large/small town differences are found elsewhere (Tables 3.4, 3.5) . Easy access to both suppliers and customers within Fij i, and through ports-of-entry to external suppliers, predicates central urban locations. Integrated wholesaler-retailers are almost all either Indians or Chinese: only two wholesale­ retail businesses out of more than 100 are Fij ian . 13

There are no formal entry requirements to grocery whole­ saling other than paymen t of the appropriate business licence fees . Although a small numb er of businesses are registered private companies , mo st are a family general retail store from which wholesale trade is conducted with minimal struc­ tural modification , apart perhaps from the construction or leasing of a bulk-store . Businesses are operated by the owner and management is by the owner or his family. If 161

Table 3.4

Locat ion and ownershiE of retail and wholesale-retail grocerl'. outlets Fi 'i a Locat ion OwnershiE (Divis ion ) b Indian Chinese Fij ian Co-op . European Other Unknown Total

Retail

Central - ruralc 173 25 262 119 4 12 595 urband 96 71 4 6 4 3 3 187 Western - rural 458 7 315 126 18 16 940 urban 163 28 6 7 2 17 223 Northern - rural 159 9 140 113 14 4 439 urban 44 6 1 2 53 Eastern - rural 4 1 84 111 5 205 urban 9 6 1 16

To tal 1106 153 811 483 4 49 52 2658

Wholesale-retail

Central - rural 5 1 2 1 9 urban 24 10 2 3 1 40 Western - rural 2 1 3 urban 35 2 3 8 48 Northern - rural 3 1 4 urban 11 11 1 8 31 Eastern - rural 1 1 urban 2 1 1 2 6

Total 82 25 2 11 21 1 142 a Race of owner is derived from name of storeowner and/or store as listed in business licence records . In more isolated areas in particular , licence records are not complete or up to date. Data on stores in Kadavu are not available : if included, the number of retail Fij ian and co­ operative society stores in rural Eastern Division would in crease significantly . b Includes other Pacific Islanders and stores operated by Fij ian groups (e. g. women ' s clubs) other than co-operative societies . c It is assumed that all shops in rural areas sell food (with the exception of those around Suva and Nausori, many of which were visited to ascertain funct ion) . d The function of all stores in urban areas was checked individually : non food shops are excluded from the table.

Source: Business licence registers, various offices, lst quarter 1976 for the Central Division and 4th quarter 1976 for elsewhere . 162

Table 3. 5 ---- Locat ion an d ownership of retail and wholesale-retail grocery outlets , Central Divis ion ------

Location Total b Co-op . Euro- Indian Ch inese Fij ian Other Unknown stores SOC. pean

Urban areas Suva retail 140 56 65 4 6 4 2 3 c w/retail 28 15 9 1 2 1 Nausori retail 21 18 3 w/retail 10 7 1 1 1 Navua re tail 14 13 1 w/retail 2 2 Korovou retail 12 9 2

Total - -retail 187 96 71 4 6 4 . 3 3 w/retail 40 24 10 2 3 1

Rural areas Suva retail 151 38 20 70 13 4 6 w/retail 4 2 1 1 Nausori retail 241 100 2 92 45 2 w/retail 4 3 1 Navua retail 75 22 35 17 1 w/retail 1 1 Korovou retail 77 3 3 40 30 1 Vun idawa retail 51 10 25 14 2

Total ---reiail 595 173 25 262 119 4 12 w/retail 9 5 1 2 1

Urban and rural total re tail 782 269 96 266 12 5 4 7 15 w/retail 49 29 11 2 3 3 1

a Race of owner is derived from name of storeowner an d/or store as listed in business licence records . b Includes other Pacific Islanders and stores operated by Fij ian groups (e. g. women ' s clubs) other than co-operative societies. c Excluded from 'w/retail ' are manufacturers and processors holding a licence to wholesale their products.

Source : Business licence registers , various offices , Cen tral Division , lst quarter 1976. 163

vo lume is sufficient, wage labour may be hired. Larger operations employ twelve or fifteen people and managemen t, retail attendants, bulk-store labour , salesman an d drivers are all distinguishable roles. Facilities are better and trade vo lume is larger than among re-wholesalers, but the function of the two forms of operation is essentially the same : to collect and redistribute goods with minimal repacking.

In addition to groceries, wholesaler-retailers deal in the connnon hous eh old items that are basic to the trade of the general stores they service but there is min imal handling of other non-food lines . Just as retail outlets carry a broad stock to attract cus tom, so do wholesaler­ retailers need to supply the basic goods handled by retailers . Nonetheless , there is some specializat ion , particularly amon g the smaller wholesaler-retailers operating in urban and peri-urban areas , a number of whom have ceased dealing in locally manufactured goods , particularly flour � sharps, rice and sugar. One cause of this specializat ion is that allowable wholesale markups on items under price control discourage trade in these items because of the costs of stock, labour, storage and tran sport. When these products came under price control in 1973, wholesalers ceased free delivery of them ; many now have stopped dealing in the. items altogether .

The supply of these basic goods is further constricted by the fact that the producers of some that are processed locally have minimum purchase policies or policies restricting wholesale access, both of which discourage the involvement of traders in wholesaling. Wholesaler-retailer firms that deal in these basic goods are increasingly limited to the larger regional suppliers, including some integrated whole­ saler-retailers, as well as Morris Hedstrom and FCA. The move towards wholesale specialization in small volume/high­ bulk items is especially apparent among some Guj erati merchants in Suva who now deal largely in items they import

directly such as canned foods , dhal , In dian spices , edible oils , potatoes, garlic and onions.

Integrated wholesaler-retailers actively seek business. Their salesmen visit retailers to take orders within specific territories. The few retailers with telephones may telephone orders. Each retail store has a number of more or less regular suppliers who either visit or are in telephone contact. Retailers are very cost-conscious of stock and do 164

not hesitate to in form regular suppliers if they can purchase cheaper elsewhere . Moreover, mercurial sales representat ives of other wholesalers - and of indent agents encouraging retailers to import direct - appear almost daily to push their products.

Wholesaler-retailers operate in broadly-defined but not exclus ive territories (Fig.16) . Botm.daries of operations are more or less demarcated between towns, although larger suppliers ofte n deal chiefly in their own and neighbouring towns , ignoring the in tervening rural areas . Two Suva firms , for instance, send salesmen to the island towns of Levuka , Savusavu and even Waiyevo : salesmen visit these places once every three or four weeks and goods are shipped soon after. In areas accessible by road , a wholesaler-retailer attempts to serve retail shops once a week. In urban and peri-urban areas arotm.d the wholesaler 's base , the salesman takes orders in. the morning and returns with goods in the afternoon of the same day. In mo re distant areas , orders are taken one day and deliveries are made the next , or perhaps two days later.

Mo st towns have two or three chief importing wholesaler­ retailers, and a multitude of less-active or regular im­ porters. The chief importing merchants usually have a maj or share of wholesale supply to the town'shinter land, but rural areas are frequently left to smaller wholesalers. Hence , a maj or importing who lesaler in Tavua may do considerable business in Tavua and Ba and even as far away as Sigatoka , but not actively encourage much closer rural trade as the trade vo lume and credit diff iculties are not considered worth the risk.

Store owners with vehicl es generally use travelling salesmen only to replenish items that are tm.expectedly out of stock , or for regular orders from an established supplier. The owners of larger businesses visit the town depot of their regular wholesaler to collect goods . Deliveries of orders placed through salesmen are often made without an explicit transport charge , wh ich is possible partly because the zone within wh ich the salesmen operate is usually limited - rarely more than ten or fifteen miles from the wholesaler's shop - and few deal in the generally bulky price-controlled goods . Orders placed by retailers when they visit the wholesaler 's base are delivered without tran sport charge if the order is large or delivery can be in tegrated in to a standard supply rtm.. Supply source k Suva ::J / f'::::'.'>I Navua (\\ ... \ ( j \ . � Sigatoka ( \. , ,.., I)( . � . � Rakiraki / ' '\ -1 r--"F. ·�,.· �"-:...-..'. '- -... ,, ' I ....._ _,, . · ·"'0..� ..... \ - ' ,:.:_. LJ Nausori I �� .

·

.. . · .  . ! · [[l . � [ . . rlill � / q1· & l:; . . IIIill Nausori local suppliers / �) ::::::�';:�OMO rni: : D r! I r- · - . -r1 · i�:::�·:�'�""'' ( . r<'i1·1 [t! II . �! Q1: 1 !il l . � •J] ·-... '-) · ..... ";. \ :;." · I"'- '- ...... I'_,..._.. -...... _ / I ...... , / I ./ / ( ' \ L - --.. I / I I · ·· ·· I . ·.·.'-'··"'-:, ·. ····· �'�· ���8&��. ��' km 25 Figure 16 Supply areas of retail grocery stores in the Central Division (Source: Fieldwork) 166

It is rarely profitable for wholesaler-retailers to offer free goods delivery or credit , although such services may be required to establish and maintain custom. When a firm feels it has established a reliable clientele or has decided to cut back wholesaling expansion , credit becomes tight and delivery of goods is discouraged . The withdrawal of these services must be carefully timed as a debtor can easily defect to another supplier and the debt become irrecoverable.

One particularly aggressive but not very common form of wholesaling is where wholesalers sell direct from a truck loaded with supplies. A couple of such merchants operate from rural bases in the Western Division . Six-tonne trucks are loaded with basic grocery supplies in cluding price- con trolled commodities andretailers - usually rural stores - are approached for on-the-spot cash or credit sales. These travelling wholesale merchants reportedly become more sedentary and less speculative once they have established retailers ' patronage through service, attractive prices and debtor­ independence. This form of wholesaling is more common in non-foodstuff trade than the general grocery trade .

Unless a grocery wholesaler-retailer uses competitive pricing, service and credit to gain a larger share of the market, bus iness expansion generally entails movement in to ano ther function or region. The latter is more common . Few integrated wholesalers-retailers expand either vertically into indenting or food manufacture or horizontally into non- f ood trade. Once a company feels that limits of viable expans ion in the local market have been reached and that further attempts to capture a larger share will provoke ultimately uneconomic competition , a branch may be established in an other centre. A number of wholesale-retail outlets opened in Suva over the past few years are expansion outlets for Tavua and Labasa businesses . Expansion outlets have identical functions to the parent shops, although buying may be centralized in one out let . Branches become increasingly independent however, and the family 's en trepreneurial effort is likely to be concen trated in the more active and successful outlet. Moreover, the reliance on family for skilled an d professional input creates a limit to expansion ; the largest number of retail and wholesale-retail general grocery outlets operated by an imme diate family en countered in the survey was three and even these operated largely in dependently of one another . As the food market will not bear much competition , expansion generally entails considerable relocation wh ich in 167

tuTI1 encourages branch autonomy .

TuTI1over of in tegrated wholesaler-retailers varies con siderably. Urban-oriented outlets in Suva probably have gross annual turnovers of $100 ,000 to $200 ,000. Well-stocked suppliers serving rural areas eaTI1 arolll1.d the upper limits of this range. Wholesaler-retailers based in Suva serving smaller towns and island centres have gross turnovers of $1-1. 5 million . Turnovers of the similar class of wholesaler­ retailers in Lautoka and Labasa are probably in the same range. Turnover of wholesalers elsewhere would be consider­ ably less.

In a second form of importing wholesaler-retailer, as distinct from the integrated wholesaler-retailer , wholesale and retail operations are administratively and even spatially dis crete. Burns Philp (South Seas) Co . Ltd,14 Morris Hedstrom Ltal5 and the Fij i Co-operat ive As sociation are the primary examples of this type of wholesaler. These firms differ from other wholesaler-retailers in fundamental ways in addition to scale of operation . Mo tivation , organizat ion and business methods are 'upper circuit ', so reflecting the expatriate­ European ownership or management . The wholesaling activit ies of the FCA have been examined above . In this section , attention will focus on Mo rris Hedstrom and Burns Philp .

At one level, Mo rris Hedstrom and Burns Philp have much in common . Both have integrated vertical structures that enable them to handle goods from the overseas purchase stage through to retailing with minimal non-company involvement. Through trading volume , each has access to quantity discounts that are rarely available to other importers. Their grocery wholesaling activities are largely passive . Another feature in common is the 'service ' aspect of their wholesaling: their stock includes price-controlled essential commodit ies shunned by many wholesalers and they carry some customers over prolonged periods of financial difficulty. The extent and role of philanthropy in such service is difficult to determine, but the broader, partly non-economic, context of the firms ' activities is un deniable .

The importance of grocery trading to Burns Philp and Mo rris Hedstrom differs. Burns Philp is more retailer than wholesaler : only between on e-quarter and on e-third of its grocery sales , in cluding liquor, are wholesale. Those foods that are sold wholesale are mainly imported. Price controls 168 on basic items and competit ion among who lesalers have led to a corporate policy that focuses grocery wholesaling on high value imported items with minimal involvemen t in products man ufactured in Fij i. Where locally-processed products are wholesaled, it is as a service to large customers , such as hotels, and to en sure continued wh olesale patronage of non­ food lines. Groceries are wholesaled mainly through the Suva , Lautoka , Sigatoka and Labasa outlets. Wholesale service is available elsewhere (Levuka , Savusavu and Tavetmi) but is confined to liquor and some other supplies to hotels and clubs. Company managemen t views wholesale activities of the smaller branches as primarily a commtmity service .

Burns Philp 's emphasis on retailing is part of an attempt to rationalize merchandising and to break into the urban­ centred mass market for groceries and general homeware. The orientat ion is a departure from the company's traditional emphasis on grocery and consumer goods merchandising for the expatriate 'carriage-trade ', which has now lost much of its former significance. It is based on western-style, media-oriented , self-service department stores managed largely by young, recently recruited expatriat e staff. Wholesaling has taken a back-seat .

Locally-processed grocery requirements are purchased as required by Burns Philp branch stores for either retail or wholesale trade. Overseas buying is centralized in Suva, whence orders are channelled through Burns Philp off ices overseas (in Aus tralia, New Zealand, Hong Kong and the UK and USA) or through an agent where the company does not have an office (as in Taiwan) . Imported goods are stored in the company's Suva warehouse and distributed to branches or wholesaled to retailers and other wholesalers. Foods that are imported by Burns Philp in clude salt, condensed and evaporated milk, canned corned beef and 'luxury ' items like jams , spreads and miscellaneous canned goods . The company does not hold any exclusive distributive rights to food lines.

Burns Philp has the financial resources and supplier contacts to import all its overseas requirements but local importers are frequently used when stocks run out . As pur­ chases are primarily for retail sale , rather than re-wholesale, the company can use local importers and so reduce importing, storage an d inventory costs. Fruit and vegetables in cluding dhal , onions , garlic and potatoes are the chief items purchased from importers, the products' perishability being for Burns Philp an additional disincentive to importing. 169

Mo rris Hedstrom Ltd and Woolworths Ltd, the me rchandising outlets in Fij i of W. R. Carpenter Holdings Ltd, are mo re involved than Burns Philp in both wholesaling and retailing groceries and basic consumer items . Morris Hedstrom operates ten general wholesale-retail outlets. Wh ere trade volume will support all-cash retailing of foodstuffs and homeware , these activities are hived-off to be handled by in dependen t Woolworths out lets. There are two Woolworths stores in Suva and it is possible that others will be established in Lautoka and Nausori in the near future.

Morris Hedstrom is a maj or foodstuff and general whole­ saler. The significance of wholesaling activit ies varies between branches : in some, such as in Lautoka , wholesaling can account for 80 per cent of trade . Foodstuffs and associ­ ated household non-food requirements like soap account for 50 to 60 per cent of Morris Hedstrom merchandise trade by value .

Morris Hedstrom branch stores have considerab le autonomy in trading activities . Each branch is respon sible for retail and wholesale trading within the contiguous region . Branches purchase autonomously most locally produced supplies but all importing is done through the Suva head office . For some items , such as canned mackerel, demand of all Morris Hedstrom branches is amalgamated in one order so volume discounts are obtained; individual branches are responsible for assessing demand and subsequently selling their share of the split­ shipment. Orders for local products on wh ich volume discounts apply are also amalgamated and subsequently broken in to indi­ vidual branch requirements. The Lautoka branch acts as a clearing house for Western Division and Labasa bulk-volume orders.

Wholesaling of foodstuffs by Mo rris Hedstrom should be seen in the context of the firm's role as a general whole­ saler, which en tails carrying the low earning price-controlled basic commodities spurned by many wholesaler-retailers as well as more lucrat ive grocery and non-food lines. Wholesaling of basic grocery items is largely passive : minimum services are provided to clients, so limiting costs on goods wh ich return little profit . Salesmen do not solicit orders and retailers who stock from Morris Hedstrom are encouraged to collect their goods . Delivery is made without charge only if significant purchases of higher-profit goods are made . In contras t, trade in liquor and non-food items is ac'tively pursued by salesmen . In spite of the passive attitude to 170

wholesaling basic foods , these commodit ies serve as the bas is of Morris Hedstrom' s general wholesale trade and play an im portant part in promoting retail trade . It is the range of goods carried, as much as volume of trade , that sets Morris Hedstrom apart from other wholesaler-retailers .

In addition to a long-established connnitment to broad­ based wholesaling, the economic advantages coming from the company 's vertical links an d the amal gamation of branch requirements to qual ify for volume discounts, another factor contribut ing to Morris Hedstrom' s dominance of general food wholesaling is the firm's control over the importing of Ocean mackerel . This is the most popular variety of canned fish in a brand-conscious country . Trade figures for Ocean mackerel relat ive to other varieties are unknown , but there is no doubt that it dominates canned fish sales. The exclusive import access to Ocean mackerel is a strong draw item for retailers and for other wholesalers serving a Fij ian clien­ tele. In addition to Ocean mackerel, other wholesalers purchase from Morris Hedstrom a limited range of imported items , as they also do from Burns Philp. Morris Hedstrom does not hold exclus ive distribution rights to other signi­ ficant food lines.

As the recent creation of Woolworths ' outlets suggests, the grocery trade is increasin gly specialized and the structure of Morris Hedstrom wholesale activities is far from static. With road development in the Western Division , for instance, wholesale activities are becoming centralized in Lautoka. The time is not very far off when Morris Hedstrom wholesale services in the Western Division will be offered only out of Laut oka . Wholesale activity in Nausori may like­ wise give way to Suva, and it is likely that Labasa will service Savusavu once the two areas are linked by road.

Non-retailing wholesalers

Food processors and manufacturers. Food processors and manuf acturers are the chief non-retailing foodstuff whole­ salers ; the main products involved are flour and sharps (one miller) , sugar (one miller with four mi lls ), dry biscuit s (two man ufacturers) , edible oil (two importers/packers and processors) , rice (three chief millers), tea (three packers) and dairy pro duce (one central dairy) . Apart from three sugar mills, all plants are located in Lautoka or the Suva­ Nausori area. All the firms are primarily processors and manufacturers , but they also wholesale their products. Some 171

manufacturers have expanded into manufacturing of both food and non-food lines , but generally processors are owned and operated independen tly of one another. 16 The ownership of processors is in dependent of ownership of overseas grocery suppliers or importer/wholesalers based in Fij i.

The processors in fluen ce food marketing in two main ways. One is through the ex-factory purchase unit which usually remains unaltered tmtil the item is retailed; the other is company policies towards wholesale access. At one ext reme , manufacturers are largely passive in the marketing of their product s and restrict wholesaler access. Some companies attempt to struct ure wholesaling by establishing price structures with volume purchase discounts; such attempts to regulate distributive systems seem to be the subj ect of informal agreements between manufacturers of similar products. Other manufacturers , particularly those producing in compet ition with other producers, have open access and flexible sales policies , with a number even employing a small active sales staff . An examination of the marketing policies of a small numb er of local foodstuff processors gives an impression of the range of their involvement in marketing.

The sugar in dustry and its involvement in wholesaling its product offers an example of restrictive merchant-access policies. Sugar is milled and marketed by the government Fij i Sugar Corporation (FSC) . About 20, 000 tonnes are ear­ marked annually for the internal market , in cluding markets in neighbouring Pacific countries . So precise is the allo­ cation for the local market that a slight delay in the start of a season's crushing can result in widespread tmavailability of sugar an d price dislocat ions - as happened when the 1977 crushing season was delayed one week through industrial disputes. Until the 1977 season , raw sugar was wholesaled from FSC's four mills to preselect ed wholesalers, a relat ively unchangin1 group chosen on the basis of past purchase history. l Each wholesaler has an annual purchase quota of at least 100 tonnes .. Purchases ar e··made by the tonne , in sacks of approximately 108 kg weight , 18 in minimum lots of five tonnes; there axe no quantity discount s. All sales are cash, ex-mill. Wholesalers collect sugar at the mills and transport it to their bulk stores whence it is sold to retailers without repacking.

From the 1977 season , FSC has limited internal sugar supply to Penang mill at Vaileka, on the grounds that the sugar produced there is free-flowing and so more suit ed to 172

consumption in its raw state. The concentration of internal supply at one outlet has admin istrative advantages for FSC, although the cost to the consumer has not been calculated.

The other maj or mon opoly food producer in Fij i, Flour Mills of Fij i Ltd (FMF) , has an open and flexible in ternal marketing system. FMF is an overseas-owned company estab­ lished in 1973 to mill flour , sharps and other products from imported wh eat . Government in fluence on the establishmen t of the mill and its operations has been considerable. It is reported that by the agreement illlder which the mill was established, no other flour mill can be built in Fij i for a period of ten years an d flour can only be imported illlder government licence , wh ich is granted when the mill is illlab le to meet domestic demands. The company also received taxation and other incentives . For its part , the company reputedly guaranteed to produce flour an d sharps below the landed cost of equivalent imported products. 19 Sales are in 65 and 50 kg sacks , ex-mill; there are no bulk purchase discoilllts or company delivery services; credit is available.

In respect of distribut ion , government influence on FMF has been two-fold. The mill has been directed to sell to any customer who makes a minimum purchase of one tonne. FMF was also encouraged to assist merchants develop regional ware­ houses to hold approximately two weeks ' supply of flour an d sharps . Fourteen merchants are commissioned as 'warehouse agents' in the main centres of the Western Division and in Labasa, with FMF extending credit facilit ies to enable each to stock approximately 1000 sacks of flour and sharps. But warehous ing costs (including fumigation) , the disincent ive of capital tied up in stocks , together with improved connnun i­ cations between Suva and the Western Division led to a collapse of all but two of the me rchants within two or three months; these two me rchants now carry well below the in tended stock.

FMF has been respons ive to local market demands - the question of profiteering being sub judice . However, expansion of milling facilit ies in early 1977 is aimed at production for export . 20 Once export sales commence , it will be inter­ esting to see whether service of the local market retains high priority .

Another way of structuring a distribution system is through price incentives. One food producer that has attempted this is Cope Allman (South Pacific) Ltd, an Australian-owned 173

company located in Suva , which manufactures edible oils and fats, margarine and dry biscuits, in addit ion to soap and plastic products. Oil , fat an d margarine production is primarily for industry and large-scale users . Dry biscuit s, however , are aimed at local consumers andCo pe Allman manu­ factures in competit ion with one other local producer.21

Cope Allman has attemp ted to develop a three-tier system of distribut ion of its products, particularly for biscuits and soaps : 'dealers ' who make purchases of $30,000 to $40, 000 per month an d so obtain goods by quantity discotmts at prices that allow re-wholesaling or large retail returns; 'wholesalers ' who purchase in 100 to 200 carton lots at smaller discounts that do not encourage re-wholesaling; and 'retailers ' purchasing in lots of fifty cartons and less. Dealers in clude the three or four maj or wholesaler-retailers and at least one Indian jobber who places a single substantial order which he then bulk-breaks and distributes to a number of retailers - who frequently give the jobber a specific order.

Price and service support of products reflects the degree of competit ion in each market . For in stance, Cope Allman prefers to sell dry biscuits to only seven selected wholesalers in Suva , Savusavu and Labasa on a c&f Labasa/ Savusavu and into-store Viti Levu basis. A small depot has been open ed in Lautoka to solicit custom and co-ordinate ·supplies in the Western Division which is the locat ion of competing soap and biscuit manufacturers : 5 to 10 per cent of an order is made immediately and the balance is then sent from Suva . For products other than biscuits or soap where there is less competit ion , sales policies are not so generous. Sales on such goods are ex-factory or ex-Lautoka warehouse and, if they are offered at all , volume purchase discounts are considerab ly below those offered on soap and biscuits.

With the exception of milk and butter supplied by the Rewa Co-operative Dairy Company Ltd, distribut ion and product­ access policies of other local food processors are relat ively open . 22 For example, the three chief rice mills sell single 45 kg sacks of rice ex-factory and all have far-ranging sales teams . Two mills have established warehouses in centres away from their mills in attempts to increase trade . Rewa Rice, the third mill , relies on supplies to cane-farmers for a considerable share of its market , but is currently attempting expansion by selling rice in 'retail packs ' up to 10 kg in contrast to the 45 kg sacks offered by the other millers. 174

Non-retailing, non-processing wh olesalers . These are either specialized importers or general me rchants who import directly and purchase from local proce ssors and other suppliers. Both types of operation are of recent origin and have a minimal share in total wholesale trade.

The chief specialized importer-wholesaler is a Suva­ based Ch inese firm impo rt ing fresh vegetables and fr uit . Through exclusive distribut ive rights for products of the New Zealand Apple an d Pear Board (as we ll as possibly other overseas produce suppliers) and aided by extens ive contacts and experience , this importer handles much of the Central Division import trade of fresh vegetables an d fruit , in cluding potato, on ion and garlic imports. From a warehouse in central Suva , the importer supplies market vendo rs , retail shops, the maj or expatriate-owned supermarkets an d other wholesalers . Fruit and vegetable importing into th e Western Division is more open because of the greater accessibility of importers to Nadi airport , the point of ent ry of air-freighted produce.

A subsidiary of a New Zealand firm re. cently established a warehouse in Suva for the wholesale distribut ion of imported franchised grocery lines. It remains to be seen whether there is sufficient deman d for the exclusive products, which includ e a line of sweet biscuits and an in stant coffee, to support a spe cialized grocery distributor-wholesaler.

In a second type of non-retail wholesale operation , intermediaries both import an d purchase locally processed foods for re-wholesaling. As well as actually purchasing goods, some 'wholesalers' operate as 'brokers ', assembling orders for shipment to isolated customers . With good co­ ordination of supplies and ships, it is possible for these brokers to operate without extensive storage facilities . The few presently doing bus iness who are located in Suva and who serve island retailers have small bulk stores . Prices charged by these assemb ler-wholesalers are invariably higher than those available at other wholesale so urces. However, their survival is an indication that they provide a service not offered by other wh olesale outle,t s serving the same regions.

Trends in wholesaling

Grocery wholesaling in Fij i is part of a general whole­ saling syste m from which it does not differ sign ificantly in 175

organizat ion . Entry into wholesaling, as in to import ing, is easy . Wholesaling is primarily on a back-order system, which suggests the trade is under-capitalized. Because of ease of access to import ing, the spat ial concentrat ion of demand an d, for groceries , intervention of the government through price control , there are few in termediaries between importer/wholesaler or local pro cessor an d retailer. Gro cery wholesaling is generally 'open ' in that there is little exclusive brand representation . Al though their actual share of trade is unknown , wholesale trade is dominated by the two overseas-owned firms . This situat ion has additional significance in that the racial ownership of the firms an d their methods of bus iness organizat ion are different from mo st other participants in the grocery trade.

Recently, there have been some changes in grocery whole­ saling. One is the apparent increasing dominance by the overseas-owned firms , particularly Morris Hedstrom. In spite of in tentions to the contrary , price control measures seem to have enhanced the position of Morris Hedstrom as numerous independen t wholesaler/retailers have ceased dealing in price-controlled essential connnodities, wh ich has enabled Mo rris Hedstrom to cement its role as a gen eral wholesaler by using the essential connnodit ies as draw items for more lucrative lines. In addition , trade volume and capital resources give Morris Hedstrom access to favourable bulk­ trade discounts. While wholesale supply by Mo rris Hedstrom may be efficient, in creasing con centrat ion of trade raises the possibility of a monopoly over supply - or at least of overwhelming dominance .

An other wholesaling trend is the establishmen t of 'dist ributors' who import groceries and basic household requirements in bulk, particularly by container loads to gain freight cost benefit s and protect ion against pilferage , an d who then sell the products to retailers via their own distribut ion network. The range of products handled depends very much on ' special buys ' and does not in clude many basic foodstuffs. In the short-term such operators may cause some dislocation to larger established wholesalers by being able to sell a particular item more cheaply because of speculative purchasing and lower overheads. In the long run this compe­ tit ion could lead to better stock supplies and more competitive pricing.

Also evident is the establishment of regional warehouses by both distributors and manufacturers. Warehousing is 176 increasingly corrnnon in the Lautoka area for Suva firms , and vice versa, but as yet has not been established outside Viti Levu. It is possible that improved communications within Viti Levu will lessen the need for regional warehousing. It is as likely , however , that competit ion will be as intense in service as in pricing, so regional distribut ion centres will be maintained.

The dominance of Mo rris Hedstrom and the establishmen t of distributors and regional warehouses all point towards in creased specializat ion in wh olesaling. Trade is becoming concentrated in urban importer-wholesalers, even at a sub­ firm level.

Retailing

There are two basic forms of retail grocery outlet in Fiji. One is the mobile outlet or hawker, and the other the fixed outlet or store. Hawkers play a negligible role in foodstuff distribut ion today, although until recently they were important in both rural and urban areas . As either a purely retail outlet or the retail arm of a wholesale­ retail operation , stores are the consumers' basic source of import ed and locally-processed food.

Con comitant to the establishment of shops in 'urban' centres, and pre-dating this development elsewhere , foodstuffs and other goods were available from hawkers. The number of hawkers probably reached a peak in the inter-war period when Indians, in part icular , hawked both produce and processed foods , as well as non-food items , in Indian settlements, Fij ian villages and towns. Activit ies of hawkers were largely uncontrolled until 1924 wh en the �eneral Hawkers L icence was introduced. This limited products which could be sold by hawkers . For instance, hawkers could not deal in pigeons and wild birds but did not require a licence if they hawked beads , curiosities or food an d produce .2 3 Controls over hawking have developed steadily since 1924. Under the 1935 Towns Ordinance , all hawkers operat in g within gazetted towns had to be licensed. By-laws enacted by local town bodies during the 1940s main tained this requirement.

During the 1950s the number of hawkers dropped, partic­ ularly in the face of improvements in tran sport that expanded shop hinterlands . In the early 1960s, however, there was a rebirth of hawkin g associated with the development of tourism. 177

Concern over the in creasing numb er of hawke rs and complaints from retailers that hawkers competed with them by selling an almost un limited range of goods , led to new legislat ion . A significan t feature of the legislation was the definit ion of 'goods ' that hawkers could offer for sale . Re strictions on goods sold we re aime d primarily at non-food items . Hawkers were still ab le to sell pro duce so long as not in contravention of local market regulations - which generally prohib ited the sale of fresh vegetables and fruit within two or three miles of off icial markets except at licensed out lets, i.e. , stores . They were not prohibited from selling prepared food, but foodstuff sales have become largely con­ fined to confect ionery and snacks . 24 In small towns, hawkers are an portim ant source of food, part icularly on the rural out skirts of town and with foods like fish and game . As towns grow and hawker legislation is enacted and en forced, it is likely that this trade will diminish .

There are not many licen sed hawkers in towns with little tourist trade. Levuka, for instance, had on ly one licensed hawker in 1976. In other towns, especially in Suva and Lautoka, the demand for hawker licences can be so great that the numb er issued may be limited, as it was in 1976 in Suva to 700, 300 of wh ich were for foodstuff hawkers . Hawking today caters primar ily to the tourist trade and provides snacks at public events. Hawkers continue to be main ly Indians.

There is potential conflict between hawkers and local authorities in that the tendency of . handicraft an d curio sellers to operate from fixed pit ches is at odds with the definition of hawkers as mob ile vendors. In creasing demand for selling space and conflict between hawkers , police and tourists have led to reformulation of hawker legislation , especially in Suva where the number of hawkers is largest an d problems associated with them mo st acute. Th ere , pro­ vision is being made for a 'Bazaar Licen ce ' that will permit hawkers to operate from stationary sites. A similar form of licence is likely to be adopted in other towns visited by overseas tourists. The bazaar licence is designed to acconnnodate handicraft dealers and has no innnediate consequ­ ences for foodstuff retailing.

Retail stores are the chief source of imported an d locally-processed food products for the consumers of Fij i. In size an d variety of stock, stores range from sophisticated supermarkets in multi-funct ion department stores to small , 178

poorly-stocked concerns where goods are stacked in the corners of a living-room from which business is conducted. Between these extremes, a wide range of store-types can be distinguished on the basis of locat ion , ownership features, size and range of stock and clientele. Very few retail stores are owned by registered companies . Most are family businesses carrying groceries as part of a general stock. A number of general stores in towns have recently been con­ verted into self-service 'supermarkets'. The predominant store in rural areas is either a co-operative society outlet or a small general store owned and operated by an individual and his family.

Grocery retailing is characterized by easy access for participants, trading that is 'open ' in that there are no significant restrictive supply monopolies, both marked competition (e.g. in central Suva) and little competit ion (for example , in most rural areas) , and low returns relative to time invested. These characteristics contribute to con­ siderable in stability in food retailing. The annual gross turnover in retail stores ranges from $3-$4 million for the three chief supermarkets in Suva to less than $1000 at a cons iderable number of rural stores. At most outlets selling groceries, around 80 per cent of turnover comes from grocery sales.

Retailing in the Central Division

To gain an impression of the form and function of retail grocery stores in Fij i, it is useful to consider the stores of the Central Division . An immediate impress ion that comes from acquaintance with connnerce in the Division , as with Fij i at large , is that there are a great number of shops. In the Central Division there are over 1300 registered retail businesses , 782 of which deal with groceries, in addition to another 47 foodstuff wholesale-retail businesses (Table 3.5) . Re tail stores that do not sell food are located primarily in Suva city, although there are also a few in Nausori and Navua. 25 Only five shops in rural areas of the Central Division - all in the greater Suva area - do not sell any foodstuffs.

Three-quarters of the 829 shops in the Central Division that sell gro ceries are in rural areas . In Suva city there are 166 shops that sell food and in the towns of the Division there are another fifty-nine (thirty-one in Nausori, sixteen in Nav.ua , twelve in Korovou) . One-third (32.3 per cent ) of 179 the shops are owned by Fij ian individual s; 15.4 per cen t are co-operative society stores . 26 Indians own 35 .9 per cent of the shops and Chinese 12 .9 per cent. The remainder are operated by Europeans (0.6 per cent) and 'other ' races (1 per cent) .

Data on store ownership by race do not reveal locational aspects of store ownership and share of trade. Wh ile one­ half of all stores are Fij ian-owned, all but twelve (eight of which are co-operative societies) are out s ide urban areas. In contrast , Chinese stores are concentrated in urban areas (81 out of 107). Indian shops are connnon in both areas (40 per cent urban , 60 per cent rural) . European stores, for example, Morris Hedstrom , Woolworths and Burns Philp, probably handle about 35 per cent of the retail grocery trade of the Cen tral Division , Indian stores ab out 30 per cent, Chinese ab out 12 per cent, and Fij ian and co-operative society stores about 8 per cent each . 27

Bas ic data on retail food shops in the Central Division and their owners were obtained omfr a questionnaire survey conducted in May and June 1976. In all , 352 randomly-selected stores selling food were surveyed. The coverage in rural areas was about one-third; in Suva city it was about one- half an d was complete in Nausori, Korovou and Navua towns. 28

The stores. Most Cen tral Division stores (63 per cent) are operated either by an individual or as a partnership (12.3 per cent) , usually within the family. The connnon form of non-family ownership is the co-operative society (19.3 per cent) . Another form of Fij ian ownership is the 'club ' store wh ich is operated by a village group , such as a family , a youth or women 's club , or mataqali. A connnittee is usually selected to operate the store and the profits are put to a particular proj ect or shared among members of the operating group . A fifth type of store ownership organizat ion is the limited liability company registered with the Re gister of Companies and required to comply with the Companies Ordinance. In the Central Division there are only about ten registered companies involved in the food trade; about half are owned by expatriate European firms and about half by local Chinese and Indian merchants.

Differences in the ownership of stores, reflecting as they do upper and lower circuit participation , are pertinent 180

to many fo rms of store operation and organizat ion . In particular, store schedules, equipment and scale an d length of operat ion vary according to ownership .

As well as being predominantly 'family stores ' in terms of ownership , grocery stores are family concerns in that if the operator is not the owner (as in 52 .6 per cent of cases) he is likely to be a memb er of the owner 's family . Only sixty-five shops were encountered where the op erator was a paid employee; almost all these were co-operative society stores. No shops were surveyed where owner and operator were not of the same race. Almost three-quarters of shops are operated by the founder or by someone in the fotmder 's family. In the on e-half of stores where operators have changed, the change has occurred in the past four years and the new ownership continues to be family-based, albeit often from families other than that of the founder.

Family-based own ership and small scale of operat ion are reflected in retail store employment patterns. The work­ force of stores , particularly if registered companies and co-operative society stores are discounted, is largely from the owner 's family. At almost two-thirds of the stores operated and owned by individuals (181 out of 272), the owner is the full-time chief operator. In stores wh ere there is an other operator, it is most li kely to be a mem ber of the operator's family working without a fixed salary.

A large proportion (one-third) of store operators do not have regular assistance and an equal proportion have only tmpaid help. Eighty-one stores with some paid employees (in addition to the operator) were encountered; twenty-six of these also have un paid workers . All but four of the 314 recorded unpaid workers were relatives of the store-operator. In contrast , 106 of 160 paid employees were not related to the store-owner or operator. Stores are most likely to have unpaid , part-time assistance (33.9 per ce nt of stores) . Some have mo re than seven part-time helpers but the maj ority have one or two . Very few store-operators have full-time assistance and only 10.2 per cent have paid part-time help . Assistance , particularly paid help, is mo st connnon in urb an areas .

In retail stores there is little specializat ion of labour. Operator and assistants of ten perform similar tasks . In retail-wholesale stores , there is generally a division of labour between sales on the one hand and storing, packing 181 and loading duties on the other . As ide from the largest supermarkets and some retail-wholesale outlets where labour is more specialized, people of one race are rarely employed in establishments owned by another race.

The use of assistance shows distinct variation by race of store-owner (Table 3.6). Fij ians are the least likely to have assistance : two-thirds have none compared to ab out one-fifth of Chinese and Indian store-own ers . While all shops employ paid labour at about the same frequency - that is, in about one-in-ten shops - Chinese and Indian shops are about three times more likely to have help than Fij ian-owned stores.

Many shops are old and apparently securely established. Ten per cent of the surveyed shops have operated for more than thirty years, on e-quarter have for less than two and 40 per cent for under five years. At the same time , however, there are indications of instability in store operat ion . For example, of 140 retail food shops operating, and surveyed, in Suva in May 1976, twelve months later eight had closed , another eight were under new management and at least seven other food shops had been established. It is uncertain whether these changes are indicative of short-term instabil­ ities in food retailing or are an expression of a longer trend. Many rural stores have poorer survival rates than urban outlets, even after allowing for trading that can be markedly periodic; the irregular nature of trade reflects an endemic shortage of funds for maintaining stock and the variable financial fortunes of the clientele.

Many retail stores suggest a certain doggedness of survival. For in stance, store operators were asked the number of occasions on which the store had been closed for more than a day or two , aside from weekends and public holi­ days, since establishment. Only fifteen stores had closed for a considerable period and twelve of these had closed only once. The connnon reason for closure was the temporary absence of someone to operate the shop . Although many stores , par­ ticularly in rural areas , have a history of irregular closures that the survey did not reveal , others remain open for long periods when trade is moribund . 29

The expatriate-owned supermarkets and some other shops in Suva city operate on a definite schedule - Sam to Spm Monday to Thursday , Sam to 6pm on Friday and 8am to 12 . 30pm on Saturday - but few other stores do. 30 Over one-half of I-' ex:> N Table 3. 6

Employment of assistance by retail stores

Store ownership As sistance No. of stores Fij ian Chinese Indian Co-op. SOC .

None 137 57 12 31 36 Unpaid help only 134 17 36 76 3 Paid help only 55 9 8 15 27 Paid and tmpaid help 26 3 5 12 2

Total stores 352 86 61 134 68

Note : As sistance may be either full-time or part-time .

Source : Survey of retail food stores , Central Division , May-June 1976. 183 the surveyed shops open every day and an other one-third open on six days . In country areas shops may open at call and not to a fixed schedule, but otherwise there are three connnon patt erns of opening.

One opening pattern , found mainly in urban areas , is for shops to open from 6 or 6.30am until 7.30 or 8pm. Actual hours are flexible . Opening time is frequently contingent upon bread or newspaper deliveries , the signifi­ cance of the 'breakfast trade' in bread , milk and newspapers, and the general likelihood of custom. Not infrequently, small stores operating 'all day ' close to enab le the operator to visit the bank , do other errands or to rest . Another common arrangement is for shops to open an d close at the above times , but also to close for an hour or two for lunch . Shops in the suburbs of Suva and In dian-own ed stores else­ wh ere tend to follow this schedule. Finally , some shops open for specified one or two hour periods during the day, most commonly in the morning and again in the evening, as well as somet imes at midday . Co-operat ive society and Fij ian village stores most often operate on this schedule .

In the past stores tended to be at the road-facing end of a wooden business and residential building. Many newer stores are of a similar design but are const ructed of con­ crete. In town areas there is a trend towards single-function buildings , the residence of the store-owner being located elsewhere. Village stores are generally wooden non­ residential buildings . Variations in store design are apparent throughout the Central Division , particularly between stores in central business districts of towns and shops elsewhere, and to some extent also between Fij ian-owned and other stores. These differences are largely a product of the chronology of store construction .

There need not be a store building at all, of course. A few stores were encountered operating out of living-rooms , supplies being kept in bags and boxes in a corner. In all but five of the surveyed shops goods were displayed. Most display was through behind-the-counter shelves and/or display cases on the counter .

Recently a small number of shops have been constructed as , or converted to, self-service stores. These 'super­ markets' compris e less than 5 per cent of the survey sample and are almost exclusively in urban areas . They are a radical innovation in grocery retailing. 184

There is minimal development of storage facilit ies in retail stores . Few shops have separate storage facilit ies . Bulk items such as sacks of rice, flour , sharps, salt and sugar an d other goods not on display are stored on the shop floor in almost three-quarters of the shops surveyed, although 42 per cent of shops had a storeroom. Most store­ rooms are a small room in the shop building. On the whole, shops are small and crowded, even without being further cramped by stored goods . Seven per cent of the surveyed shops had a floor space of less than 9.3 sq.m. and one­ quarter less than 18.6 sq.m. ; only 30 per cent had more than 46.5 sq.m. of floor space and only four had more than 186 sq.m.

Essential store equipment is limited to shelves and display cabinet s. Only stores in towns are likely to have cash registers. Refrigerators or freezers are more connnon than cash registers. All town shops and some shops in the greater Suva-Nausori rural area have electricity and almost two-thirds have a refrigerator or freezer. Outside electrified areas , a few shops have kerosene-powered refrigeration tmits. On the whole, rural shops do not have refrigeration .

While the stores surveyed carried a broad range of stock, foodstuffs were always the basis of it. 31 Non-food items rarely accotmted for more than 'about 1/4' of stock by value , and were mainly goods like soap and other cleaning items , cosmetics , stationery , household utensils , kerosene and, less connnonly, a few clothes and some uncut material (Table 3.7 ). The most common class of goods , apart from food, was cleaning items , recorded in 92 .8 per cent of stores. Mixed goods such as kitchen utensils, clothing and other household requirements, were sold by over 80 per cent of stores and stationery products and drapery each by one-third . In over 80 per cent of the stores which sold non-food items , those goods comprised less than one-quarter of stock. Only twen ty­ two (6.3 per cent) stores did not sell non-food items . In only fifty-six (16.2 per cent) stores did food comprise one­ half or less of stock. As ide from soap , cigarettes and matches, the most widely carried non-food item was kerosene - used as a cooking fuel - which was recorded in 95 per cent of shops.

Measured by either the numb er of food types or food items on display or in storage, grocery stocks in retail stores were fotmd by the survey to be limited (Tab les 3.8, 3.9). 32 One-third of all stores had less than twenty different Table 3.7

Nature of stock in retail grocery stores

Share (by approx. value) of stock type in total stock Per cent stores Stock for stores selling selling stock <1/4 c.1/4 c.1/2 c.3/4 all

Groceries 100.0 0.9 0.9 14.4 77. 6 6.3 Cleaning items 92.8 83.5 16.5 Mixed goods 81 .2 95.3 3.9 0.8 Stationery 37.0 81. 9 17.4 0.7 Drapery 15 .1 81.1 15 .1 3.8

Note: Only shops selling food were surveyed. Valid responses for each stock category are within the range 345-348.

Source : Survey of retail food stores , Cen tral Division , May-June 1976. 186

Table 3. 8

Numb er of different food t�Ees stocked by retail stores

On display In storage a Food types Number of Per cent Number of b shops total shops shops

<10 26 7.4 43 10-19 93 26.5 13 20-29 82 23.4 14 30-39 71 20.2 40-49 51 14 .5 50-59 21 6.0 >59 7 2.0 Not available 1

Total 352 100.0 70

a A 'food type' is a generic group l.ike 'canned fish' , 'sauce' or ' sugar1 , irrespective of varieties or brands.

b Thirteen shops had goods stored but no further details were recorded:•

Source: Survey of retail food stores , Central Division , May-June 1976. 187

Table 3.9

DisElay and storage inventory of retail food stores

Number of In shoE In storage a food items No. ShOJ2S Per cent No . shOJ2S Per cent <100 10 3.1 ) 11 13.9 100-499 97 29.8 ) 500-999 66 20.2 11 13.9

1000-1499 44 13. 5 ) 12 15.2 1500-1999 34 10.4 ) 2000-2999 53 16.4 7 8.9 3000-3999 15 4.6 4000-4999 >5000 7 2.1 25 31. 6 Not available 13 16.5 No goods stored 178 No response 26 95

Total 352 100.0 352 100.0 a A 'food item ' is an individual selling unit . Where goods are stored in bulk, they are reduced to the most connnon selling units. For instance, a 112 lb bag of salt in storage is recorded as 112 units.

Source: Survey of retail food stores, Central Division , May-June 1976. 188 foods . They dealt in the bas ic items - canned fish an d meat, sugar , salt, tea, baking powder , flour , sharps and rice - and little more. Only twenty-eight of the surveyed stores stocked more than fifty food types. Over two-thirds of shops had no food in storage.

There was considerable variat ion in the size of stocks in storage. About one-third of shops with any goods stored had less than 5000 food items in storage and another one­ third had between 1000 and 3000 items . The remaining twenty­ five stores had over 5000 items in storage . The number of food types in storage was limited generally . Forty-three of the seventy shops for which data were available had less than ten different food types stored and none had more than thirty. Items most connnonly in storage we re sugar , salt, flour , sharps and rice. Survey data on stock suggest there is a correlation between mode of payment, that is access to credit , and the value of stock on display and in storage (Tab le 3.10) .

Aside from the upper Navua and Wainimala valleys, the Namosi area and the Rewa Delta, most stores in the Central Divis ion are on roads to local centres or Suva . Shops in Suva are supplied by Suva-based importers and wholesalers . These businesses are also the main suppliers for stores in other towns although a numb er of retailer-wholesalers, especially in Nausori, import directly . Rural stores are generally supplied from wholesaler-retailers in the nearest urban centre . Distinct supply areas can be delineated in the Central Division (Fig.16) . Sigatoka and Vaileka suppliers serve stores on the two road links near the boundary with the Western Division , some 100-llOkm from Suva. As well as being supplied by licensed retail-wholesale outlets in towns and at a couple of rural centres - such as co-operat ive wholesale association stores in the upper Navua valley and Serea and an Indian store at Nakelo - stores are also stocked from other 'retail ' stores . Twenty-two of the surveyed shops in addit ion to six licensed retail outlets. operate as unlicensed re-wholesalers for other stores.

One reason for the grocery supply system of the Central Division being focused on Suva-Nausori is the relative com­ pactness of the region wh ere stores are located . The mean distance of surveyed stores fr om Suva is 28.3km: two-thirds (65.9 per cent) are within 32km of the city and over one­ quarter (2 8.4 per cent) within 8.lkm. While 90 per cent of Table 3.10

Mode of Ea�ment for suEplies and food stock characteristics of retail stores

Value ($) Value ($ ) Per cent Per cent Mode of Number stock on stock in stoc stock payment stores � c display storage local b as1c.

Cash 190 393. 33 181.16 27.3 72 .o a (464.04) (702.38) (15.6) (16.0) Cash and credit 108 921. 88 506.29 25.6 65.6 (1063.38) (1614 .06) (16.2) (19.4) Credit 14 1125 .27 539.04 25.6 61. 9 (1859 . 57) (1703.10) (15.0) (23.3) Unknown 11 1381. 53 252 .20 22 . 0 56.1 (913. 53) (602.25) (11. 2) (22.6) a Figures in brackets are standard deviations of mean values. b 'Local ' foods are taken as bottled milk, eggs , butter, frozen chicken , ghee in 26oz. and smaller containers, canned tuna , local fish, rice , dry biscuits, bread, sugar , an d local fruit and vegetables. c 'Basic ' foods are taken as canned beef and mutton , dripping, milk of all types, ghee , canned mackerel,. rice, flour , sharps, dry biscuits, potatoes , onions, garlic, sugar , tea, oil and salt. Source : Food stock inventories, survey of retail food stores, Central Division , May-June 1976. 190

stores can be reached by road, the access to some is mo re difficult . Stores in the Rewa Delta are generally accessible only by water, often by a one hour or so power boat trip . Some shops in the Wainimala valley are well beyond a three­ hour boat journey. Others in Namosi are accessible on foot only.

Distance from Suva and means of access are related to the period of operation of stores . The further a store is from Suva, the shorter the period it is likely to have been open . Over 40 per cent of stores more than 64km from Suva have been open less than three years : 44.6 per cent of stores within 16km of Suva have been open for more than ten years , compared to 20.7 per cent of stores located more than 64km away .

Stocks of all shops, except for a few in Suva and other towns, are limited; the average value of displayed gro cery stock is under $700 with a further $300 worth stored . There are some interesting correlations between stock size and distance from Suva . One is that the number of food types and inventory size is correlated indirectly with distance from Suva. For instance, the proportion of stores stocking less than twenty types of food increases from 12 .1 per cent for stores within 8km of Suva to over 79.4 per cent for stores over 64km away . Likewise, only 4.4 per cent of stores within 8km have fewer than 500 items in their inventory, but 73.5 per cent of those over 64km away have inventories of fewer than 500 items . Mean stock values range from $1504 to under $300 for stores in the 1.5-14.5 and over 64km from Suva distance groups. One half of the stores within 8km of Suva have no foods stored: 90 per cent of those more than 64km away have none stored.

Whilst the proportion of stock that is gro ceries does not vary greatly with distance from Suva, stock composition does: in essence , the nearer a shop to Suva, the smaller the proportion of grocery stock that is locally produced and of 'basic' food items (Tab les 3.11-3.15). 33

Distance from Suva is on e measure of locat ion for stores , but a more prosaic indication is accessibility, especially as many shops in the Division are not served directly from Suva. 34 The maj ority of shops are on reasonable roads within 56km of the city, and others are on the two chief llOkm long roads leading to the Western Division , but a not inconsider­ able number are in locat ions that require a change of 191

Tab le 3.11

Food stocks and store distance from Suva

a Number of food types

Distance Number of Per cen t stores in food type size group from Suva stores <10 10-19 20-2 9 30-39 40-49 50-59 >59 (km)

<10 99 3.0 9.1 20. 2 26.3 21.2 13. l 7.1

10-16 33 6.1 12 .1 24.2 39.4 18.2

17-32 99 7.1 29.3 32 . 3 18.2 11. 1 2.0

33-48 60 5.0 35 .0 23. 3 11. 7 15 .0 5.0

49-64 26 7.7 46.2 11. 5 11. 5 15.4 7.7

>64 43 26.5 52 .9 14 .7 2.9 2.9

Size of food inven tory

Distance Numb er of Per cent stores in food itenPsi ze group from Suva stores 100- 500- 1000- 2000- <100 >2999 (km) 499 999 1999 2999

<10 91 4 .4 23.1 33.0 22 .0 17 .6 10-16 28 25.0 21. 4 32 .2 21. 4

17-32 94 5.3 35 .1 26.6 17 .0 13. 8 2.2

33-48 55 40. 0 12 . 7 29.1 14 .5 3.6

49-64 24 4.2 41. 7 16.7 12 .5 16. 7 8.4

>64 34 11 . 7 61. 8 8.8 11. 7 5.9

a A ' food type ' is a generic group like canned fish , sauce or sugar , irrespective of varieties or brands . b A ' food item' is the individual common selling unit , such as a can of meat or a pound of rice .

Source : Survey of retail food stores , Central Divis ion , May-June 1976. 192

Tab le 3. 12

Store distance from Suva and food stock characteristics

Distance Value ($) Value ($) Per cent Per cent Number from Suva stock on stock in stoc stock stores � c (km) display storage local basic

1-15 129 1060 .41 444 .33 25.3 60.5 a (1305 .84) (1247 .23) (13. 8) (19.9)

16-31 92 479.29 97 .37 24.1 71.4 (684.74) (400.69) (13.4) (16.5)

32-47 62 483. 01 630.18 24.4 70.2 (570 . 18) (2050.93) (16.0) (15.1 )

48-63 28 541.34 160.37 33.0 73. 3 (869 .14) (496.94) (20.8) (18.1)

64-79 28 208.13 11 . 82 35 .4 77.5 (450.55) (39.60) (18. 7) (14.9)

80-112 8 128.27 145 .33 29 .1 90.8 (163 .18) (380.54) (18.1 ) (11.4)

Unknown 3 928.18 849.37 34 .3 78.7 (1189 .16) (1471. 15) (16.2) (13.3)

All shops 350 673.2 323.4 26.4 68. 3

a Figures in ULackets are standard deviations of mean values. b ' Lo cal ' foods ai:e taken as bottled milk, eggs , butter, frozen chicken , ghee in 26oz. and smaller containers , canned tuna, local fish , rice, dry biscuit s, bread, sugar , and local fruit and vegetables. c ' Basic ' foods are taken as canned beef an d mutton , dripping, milk of all types , ghee , canned mackerel, rice flour, sharps, dry biscuits, potat oes, onions, garlic, sugar , tea, oil and salt .

Source : Food stock inventories , survey of retail food stores, Central Division , May-June 1976. 193

Table 3.13

Food stocks and retail store accessibility

Number of food t�']�esa

Per cent stores in food type size groups Means of Number of access stores 10- 20- 30- 40- 50- >59 <10 19 29 39 49 59 types

Road 312 5.1 23. 7 23. 7 22 .1 16. 3 6.7 2.2 c Boat 25 20.0 52 .0 24.0 4.0 d Boat 10 50.0 50.0 Road , foot 2 50.0 50.0

Size of food inven tory

Per cent stores in food itenP size groups Means of Number of access stores 100- 500- 1000- 2000- >2 999 <100 499 999 1999 2999 items Road 287 1.4 27.2 20.2 25.8 18. 1 7.3 --- - e Boat 25 16 .0 44.0 20. 0 16.0 4.0 d Boat 10 20. 0 80.0

Road, foot 2 100.0

a A ' food type ' is a generic group like canned fish, sauce or sugar, irrespective of varieties or brands. b A ' food item ' is an individual coIImlon selling un it , such as a can of meat or a pound of rice. c Boat trip less than 45 minutes . d Boat trip over 45 minutes (and up to three hours).

Source : Survey of retail food stores, Central Division , May-June 1976. I-' "° Table 3.14 �

Trans2ort access and food stock characteristics

Value ($) Value ($) Per cent Per cent Transport Number stock on stock in stoc stock access stores � c display storage local b as1c .

Road 311 732.79 331.63 26.1 66.9 a (1032 .80) (1226. 42) (15.3) (18.8) Road and boat 8 393.35 455.42 26.3 70.5 (<45 mins .) (508.89) (871.50) (16.4) (11 . 4) Road and boat 27 118 .15 56.16 27.8 81.4 ( >44 mins .) (121. 89) (209 .54) (17.5) (15.0) Road and cons id. 2 123.14 36.0 80.5 foot (12 . 47) (9.9) (9.2) a Figures in brackets are standard deviations of mean values. b 'Local ' foods are taken as bottled milk, eggs , butter, frozen chicken , ghee in 26oz. and smaller containers, canned tuna, local fish , rice , dry biscuits, bread , sugar, and local fruit and vegetables . c 'Basic' foods are taken as canned beef an d mu tton , dripping, milk of all types , ghee , canned mackerel , rice , flour , sharps, dry biscuits, potatoes , onions, garlic, sugar, tea , oil and salt . Source : Food stock inventories, survey of retail food stores, Central Division , May-June 1976. Table 3.15

Storeowner 's possession of business transport and food stock characteristics

Value ($) Value Per cent Per cent Numb er ( $) Type of stock on stock in stoc stock cases � c t.ransport display storage local basic

Vehicle 95 1352 .09 653.54 24.2 64.6 a (1420.23) (1637.22) (14.0) (18. 7) Boat 20 149 .61 142 .63 30 .8 86.3 (263 .16) (384. 07) (17. 3) (11. 9) No transport 221 399.69 224.40 27.1 69.2 (448.13) (102 7. 06) (16.0) (17 . 3) Unknown 14 1132 . 72 24.0 53.4 (1753.30) (16.2) (26. 7) a Figures in brackets are standard deviations of mean values. b 'Local ' foods are taken as bottled milk, eggs , butter, frozen chicken , ghee in 26oz. and smaller containers, canned tuna, local fish, rice, dry biscuits, bread, sugar , and local fruit and vegetables. c 'Basic' foods are taken as canned beef and mutton , dripping, milk of all types , ghee, canned mackerel , rice, flour, sharps , dry biscuits, potatoes, onions , garlic, sugar, tea, oil and salt. Source : Food stock inventories , survey of retail food stores , Central Division , May-June 1976. 196 transport mo de . For instance , about thirty of the stores surveyed are in the Rewa Delta area and can be reached only by boat (by trips of up to an hour) from landings on roads that lead to Suva and Nausori. As noted above , stores in the upper Navua and Wainimala valleys and Namosi areas are even more inaccessible.

Stocks are in fluenced by the accessibility of stores. Of the thirty-five stores reached by boat, 80 per cent have fewer than twenty food types and only one has more than twenty-nine . The corresponding rates for road-accessible stores are 28.8 and 47. 3 per cent . Moreover, none of the stores reached by a boat trip taking longer than 45 minutes has more than nineteen food types. A similar pattern is apparent in the number of items stocked : 71.4 per cent of stores accessible by boat (and all of those reached by a boat trip over 45 minutes) have fewer than 500 items , whilst only 28.6 per cent of road-accessible stores have stocks of this size. The amount of goods in storage also reflects access: only two of thirty-seven shops inaccessible by road have goods stored.

Four basic forms of store ownership were encountered in the Central Division : businesses owned by individuals or in formal partnerships but not registered as limited liability companies (hereafter called 'private stores ', 265 cases); co-operative consumer society stores (68) ; village 'club ' or rnataqali stores (14); and registered companies (4) . These forms of own ership imply different racial, economic and legal requirements so it is to be expected that stores of different ownership will vary in particular ways . The retail store survey revealed (by chi-square analysis) that stock-related variables like stock composition or size are independent of ownership form (Table 3.16) . Other variables like display form, operator, founder, operating schedule and variables reflect ing the scale of operation are not independent of the type of ownership .

A marked difference by ownership form is in the use of help in shops. Private stores rely most on unpaid labour, 42 .2 per cent having part-time labour and 17.5 per cent unpaid full-time labour . Village stores have the next highest use of 1.lllpaid labour (14.3 per cent part-time , 21.4 per cent full-time) . Co-operative society stores have least unpaid assistance - their storemen and other officials are usually salaried. Registered companies have highest employment frequency of paid labour, all having full-time and part-time Table 3.16

Form of store ownershiE and food stock characteristics

Value ($) Value ($) Per cent Per cent Form of Number stock on stock in stoc stock ownership stores � c display storage local bas ic

Licensed individual , 265 763. 33 384.75 25.5 66.1 a family (1005 .16) (1316.62) (15.5) (18. 9) Co-operative 68 250.38 49.60 28.7 77. 7 (389.55) (216.30) (16.4 ) (13.0) Village group 14 160.13 18.57 33.8 72 . 3 (179. 60) (57.14) (13.1) (15. 3) Registered company 3 4963.22 2621. 65 22.0 28.0 (96. 80) (2360. 94) (4.4) (20. O) a Figures in brackets are standard deviations of mean values. b 'Local ' foods are taken as bottled milk, eggs, butter, frozen chicken , ghee in 26oz. and smaller containers, canned tuna, local fish , rice, dry bis cuits, bread , sugar, and local fruit and vegetables . c 'Basic' foods are taken as canned beef and mutton , dripping, milk of all types, ghee, canned mackerel, rice, flour , sharps, dry biscuits, potatoes, onions, garlic, sugar, tea, oil and salt.

Source : Food stock inventories , survey of retail food stores, Central Division , May-June 1976. 198 paid help. Village stores (7.1 per cent full-time and part-time ) an d private stores (20.2 per cent) have lowest rates of paid help.

Many differences in store operat ion that may at first appear attributable to form of ownership are more probably the product of a variety of interrelated factors. These include, in addition to ownership form, variations in oper­ ational modes by race of owner , involvement in a particular commercial 'circuit ' and the racial and socio-economic character of the clientele. Consequences of these factors are examined- below. It is sufficient here to make one bas ic point ab out retail operat ions in Fiji: there are significant differences in the form and nature of operation between outlets operated by different races .

Table 3.1 7 presents data on the locat ion of Central Division stores and the period they have been functioning, by race : 35 Fij ian-owned stores are both the most distant from Suva, the chief source of supply and the most recent ly established. Fij ian stores have been operat ing for less than half as long as all shops and average only one-sixth the life span of Guj erati shops. Co-operative society stores have been operating lon ger than individually-owned Fij ian stores (8.3 years compared to 5.1) but still for shorter periods than any other shops. Fij ian shops are located further from Suva than shops of other races. The average distance from Suva for all shops is 29km. For Fij ian shops the average is 39km, for Chinese llkm an d for Indian from 19 to 32km.

Similarly, in terms of inventory size, origin (local/ non-local) and composition (basic/non-basic) , shops of the chief racial groups fall into distinct categories (Tables 3.18-3.21) . Fij ian shops have the smallest stocks in storage or on display, and the largest share of stock that is local and basic. Chinese and In dian stores have grocery stocks of approximately the same size (almo st ten times the value of stock in individually-owned Fij ian shops) and local and basic foods are of equal importance in shops of both races. In addition to having larger and more diverse stocks, Chinese an d Indian shops are less dependent on groceries than are Fij ian stores.

Racial variables in retail ing and in the processed foods dist ribut ion system in general are examined further below . It is enough here to conclude by noting that while 199

Table 3.17

Distance of retail shops from Suva and years of operation : comparison by race of owner

Km from Suva Years operating Storeowner a Mean (N) Mean (N) Fij ian 39 .0 (141) 5.1 (58)

Other Pacific Isl. 6.4 (3) 5.0 (2)

Indian

Moslem 31. 6 (12) 18.3 (11)

Hindu 24.6 (96) 17 .2 (17)

6uj erati 19.5 (15) 29.7 (11)

Punj ab i 29.6 (10) n.a. n.a.

Indian, unspec. 25.6 (8) 3.7 (6)

Chinese 10.6 (51) 14.8 (23)

Part-European n.a. n.a. 8.0 (3)

All shops 28. 2 (352) 12 .4 (236 ) a Mean calculated from number of shops for which there are valid responses (N) .

Source : Survey of retail food stores , Central Division , May-Jnne 1976. N 0 Table 3.18 0

Number of items disElaled and stored.- b�·-race of storeowner

Number of items on disElal Total Owner 100- 500- 1000- 1500- 2000- 3000- >8000 stores <100 500 1000 1500 2000 3000 4000 No info . a Fij ian 9 77 32 14 4 6 2 144 Indian 17 21 23 22 32 6 4 17 142 Chinese 11 6 8 13 8 5 51

Numb er of items stored Total Owner 500- 1000- 2000- >5000 Goods stored No goods stores <500 1000 2000 3000 no details stored No in fo. Fij iana 5 4 2 1 118 14 144 Indian 5 4 6 4 15 8 43 57 142 Chinese 3 4 3 9 2 8 21 51 a Co-operative society and village stores , and three stores operated by non-Fij ian Pacific Islanders , are treated as 'Fij ian' .

Source : Survey of retail food stores , Central Division , May-June 1976. Table 3.19

Race of owner an.d food stock characteristics of individually-owned stores

Value ($) Value ( $) Per cent Per cent Number Owner stock on stock in stock stock stores display storage locale basic?

Fij ian/Pacific Is.a 62 145.15 27.47 30.4 78.6 (172 . 89)b (138 .76) (18 . 7) (16.3 )

In dian 142 1000 .59 449.63 23.5 63.8 (1218 . 47) (1445.83) (14.0) (18 . 7)

Chinese 51 965.61 805 .98 25.2 58. 6 (794.26) (1809 .56) (14.5) (17.4)

Part-European 4 464.54 30.78 29.8 57.8 (462 .41) (61. 56) (10. 7) (8.0) Unknown 11 1381. 53 252 .20 22.0 56.1 (913. 54) (602 .25) (11. 2) (22 . 6)

a Stores operated by groups (e.g. villages or mataqali) or co-operative societies are included under 'Fij ian/Pacific Islander '. b Figures in brackets are standard deviations of mean values . c 'Locar foods are taken as bottled milk , eggs , butter , frozen chicken , ghee in 26oz. and smaller containers , canned tuna, local fish , rice , dry biscuits , bread , sugar and local fruit and vegetables . d 'Basic' foods are taken as canned beef an d mut ton , dripping, milk of all types , ghee , canned mackerel, rice , flour , sharps, dry biscuits � potatoes , on ions, garlic , sugar , tea, oil and salt . Source : Fo od stock inventories , survey of retail food stores , Central Division , N 0 May-June 1976. I-' Table 3. 20 N 0 N

Value ($) Value ($) Per cent Per cent Number Ownership stock on stock in stock stock stores c d display storage local basic

Fij ian and other 143 196. 33 37.12 29. 9 77.5 a Pacific Islander (300.11) b (175 .48) (17.2) (14 . 7) Indian 142 1000 .59 449.63 23.5 63.8 (1218. 47) (1445.83) (14.0) (18 . 7)

Chinese 51 965.61 805.98 25.2 58.6 (794.26) (1809 .56) (14.5) (17.4) Part-European 4 464 .54 30.78 29.8 57.8 (462 .41) (61.56) (10. 7) (8.0) Unknown 10 1436.68 333.16 22 .1 54 .5 (1867.88) (1037 .27) (10.1 ) (22 . 6) a Stores operated by groups (e. g. villager or ma.taqa li) or co-operative societies are in cluded un der 'Fij ian and other Pacific Islander ' . b Figures in brackets are standard deviations of mean values . c 'Local ' foods are taken as bottled milk, eggs , butter, frozen chicken , ghee in 26oz . and smaller containers , canned tuna, local fish , rice , dry biscuits , bread , sugar, and local fruit and vegetables . d 'Basic ' foods are taken as canned beef an d mut ton , dripping , milk of all types , ghee , canned mackerel, rice, flour , sharps, dry biscuits , potatoes , on ions, garlic, sugar, tea, oil and salt. Source : Food stock inventories , survey of retail food stores , Central Division , May­ June 1976. Table 3. 21

Food stocks of Indian-owned retail stores

Value ($) Value ($) Per cent Per cent Number stock on stock stoc stock stores in � c display storage local basic

All Indian stores 142 1000.59 449.63 23.5 63.8 a (1218 . 47) (1445.83) (14.0) (18. 7) Mo slem 12 1069 .57 1357 .98 30.8 65 .8 ( 821. 50) (2713. 77) (10. 7) (18. 4) Hindu 96 838 .21 366 .25 23.4 64 .4 (915 .65) (1356 .52) (14.4) (17.4)

Guj erat i 15 1837.08 250.07 14 .5 60.6 (2392 .82 (812 . 05) (8. 0) (25.2) Punj ab i 10 1684.94 651 . 26 28.4 61 .8 (1535 .32) (1063 . 23) (14.9) (25. 2) Unknown Indian 9 483.63 189 .46 24.5 63 .6 cultural group (381 .13) (494.57) (16.4) (15.5)

------·-

a Figures in brackets are standard deviations of mean values . b 'Local ' foods are taken as bottled milk, eggs , butter , frozen chicken , ghee in 26oz. and smaller containers , canned tuna , local fish , rice , dry biscuits, bread , sugar, an d local fruit and vegetables . c 'Basic' foods are taken as canned beef and mutton , dripping , milk of all types , ghee , canned mackerel , rice , flour , sharps , dry biscuits, potatoes , onions, garlic, sugar , tea, oil an d salt. N Source: Food stock inventories , survey of retail food stores, Central Division , 0 May-June 1976. w 204

retail grocery stores in the Central Division are, like those elsewh ere in Fij i, primarily small , family-owned an d operated outlets with foodstuffs as a dominant part of a general stock, there are marked differen ces in the location , scale and mode of operation of- outlets that are related to the owner's race.

The storeowners. Storeowners - at least those covered in the Central Division retail food store survey - share a numb er of common characteristics. 36 As a group , storeowners are male, married and over thirty years of age. The mean number of dependants is 5. 6. About 4 per cent are illiterate and an equal proportion have not been to school but claim literacy. One-third have only. some primary school education . About the same number (37.1 per cent) have completed primary school but very few (7.8 per cent) have completed secondary school. Levels of formal literacy, of course, have little relat ion to store operation . Few retailers keep written records of transactions and many are fllllctionally illiterate.

Almost all storeowners have had an occupation prior to operating the present store, the most common being farming or working in another store or small business (Tab le 3.22) . Slightly more than one-half of storeowners have another occupation apart from their shop. The most common forms of additional employment ·are farming (34.2 per cent of store­ owners with other occupat ions) and other store or business interests (26.5) . Salaried or wage employment as a second occupation is not common .

Personal characteristics of storeowners vary between racial groups. 37 For instance, forty-eight of fifty-two Chinese shopowners were born overseas compared to eighteen of 142 Indians (and no Fij ians) . Chinese storeowners are considerab ly older than others: 32 .7 per cent are over sixty years old compared to 11.3 per cent of Indians and 17.5 per cent of Fij ians . Indian storeowners tend to be the youngest. One in six Indian storeowners is under thirty years of age (and one in two Guj eratis) : only one in eight Fij ians and one in twenty-six Chinese storeowners are of comparable age . As ide from being younger , Indian storeowners stand apart through being better educated (particularly Guj erat is and Moslems ) and are more likely to be male.

Maj or differences between storeowners of different races are apparent in business experience. Only one of 205

Table 3. 22

Occupation of storeowners (prior to and concurrent with operat ing present store)

Prior to operating At present Occupation present store

Number per cent Numb er per cent

Farmer, villager 59 27.3 40 34.2

Un skilled trade 21 9.7 2 1. 7

Skilled trade 26 12 .0 6 5.1

Semi-skilled trade 18 8.3 4 3.4

Professional 19 8.8 9 7.7

Semi-professional 21 9.7 8 6.8

Store , small businessa 52 24.1 31 26.5

Other 17 14.5 Sub-total 216 99.9 117 99.9

Not available 9 2

No prior/present occupation 35 141

Total 260 260

a Includes market vendors (2) and real estate agents (5) in present occupation .

Source : Survey of retail food stores, Central Division , May-June 1976 . 206 fifty-six Fij ian storeowners owned a store or business prior to the one operated at the time of interview. In contrast, 27.7 per cent of Indian and 29.5 per cent Chinese storeowners have had another shop or business. Fij ian storeowners are from predominantly agricultural backgrounds (41.1 per cent) : the rate for Indians is 24.1 per cent and for Ch inese , 15.6 per cent . Among In dian storeowners , no Moslems or Guj eratis have been farmers but 27.2 per cent of Hindus have been . A similar pattern of agriculture and store/business ownership is apparent among storeowners with another source of income . For instance , of thirty-eight Fij ians who have another in come source, 65.6 per cent are in agriculture and only 2.6 per cent have another store or business: the rates for Indians are 23 and 36.1 per cent respectively. Chinese storeowners with other income sources are not in agriculture but mainly have another store or business.

The sections below on 'A typology of retail food stores ' and 'Race and trade ' cover further aspects of storeowners. It is eno ugh here to note that the two connnon features of all own ers is their limited education , which rarely includes commercial training, and their diverse former and current employment backgrounds. There are marked differences in these variables by the race of the owners; these are discussed below.

Grocery retailing elsewh ere in Fij i

To ch eck the validity of data from the Central Division retail store survey and to obtain informat ion about retail food out l ets elsewhere in Fij i, small surveys were conducted in Ba and Labasa towns and surrounding di strict s in November 1976. In each area, all licensed stores in town selling food and an approximately equ al numb er in adj oining rural district s we re covered . Sixty-eight stores were surveyed in Ba and Labasa.

The basic finding of these surveys is that the structure of gro cery retailing in Ba, Labasa and the Central Divis ion has much in common . Ownership forms , for instance , are similar in all areas: in Ba and Labasa, over 85 per cent of stores are individually owned and an equal proportion of shops in each is owned by Indians. In both Ba and Labasa towns and regions , six Chinese shopkeepers were encountered, as well as four and three Fij ians respectively. Patterns of ownership by race and location are also similar . In neither Ba no_r Labasa towns is there a Fij ian-owned shop , although 207 each has a co-operative society store. Chinese and Indian stores are about equally divided between town and country , although in both areas Guj erati shops (twelve in Ba, six in Labasa) are located only in towns .

One difference between retailing in Ba and Labasa is the time shops have functioned : shops in Labasa are longer established than those in Ba (20.2 compared to 13. 3 years) and have also operated longer under the present owner (10.4 years to 8.7 in Ba) . The period owners have operated any shop is similar in the two areas (14.2 years in Labasa; 13. 8 in Ba) . In both areas , town shops have operated longer than rural shops, although the difference is not as great in Ba as in Labasa where shops in town are on the average twice as old as rural shops.

Features of ownership were also compared. The survey revealed that shops in Ba are more likely to have been established by someone other than the present owner (in two­ thirds of cases) than in Labasa (one-half), and that whilst owner-operators predominate in both areas , they are less common in Ba (52 out of 68 cases) than in Labasa (61 out of 68) .

In terms of racial variables, key features of shop operat ing experience in Ba and Labasa are similar, and are also similar to the general trends revealed in the Central Division . Fij ian storeowners have operated shops for 3.3 years in Ba and 2.1 in Labasa; Indians for 15.7 and 13.l; and Ch inese for 13. 3 and 32 .8 years. Table 3.17 presents similar data for Central Division stores . Amon g Indians, Punj ab is have the longest shopkeeping experience , followed by Guj eratis who have double the years experience of other Indian storeowners; the situat ion of Central Division Indian storekeepers is not greatly different.

The Central Division survey revealed that Chinese and Indian shop$ selling foodstuffs rely less on food sales than Fij ian shops and that shops located in rural areas have greater dependence on food than other shops . Th ese findings were upheld in Ba and Labasa. For instance , in both areas groceries comprise at least three-quarters the stock of all Fij ian stores but less than this in at least one-third of Indian and one-half of Chinese shops. Food is the basic stock of all general retail stores. 208

Detailed investigations of store trade like those done in the Suva area were not conducted in Ba and Labasa. How­ ever, data on store stocks and general observation in those areas and elsewhere suggest that the relative share of retail trade by race calculat ed for the Central Division (cf. note 27) is roughly representative of the situation elsewhere in Fij i, especially if it is adj usted to reflect the distri­ but ion of store type (see Table 3.4 ).

A typology of retail grocery shops

The above analysis of gro cery retailing derived from surveys of outlets gives an impression of the broad features of retailing in the Central Division and, more generally, elsewhere in Fij i. It suggests that there are a number of distinct forms of retail business, the key variables for these being shop fl.lllction, location , form of bus iness organ­ ization and race of the store owner. The variables are inter-related but even so basic store types are identi­ fiable. Examination of these is useful for the conceptual­ ization of the retail system that resul ts and also because of the opportunity it offers to examine in detail the oper­ ation of typical outlets of each type.

In the Central Division there are seven basic types of retail grocery store:

(a) expatriate-owned supermarkets; (b) specialized food retailers; (c) suburban neighbourhood shops;

(d) suburban com1mm ity shops; (e) rural general stores; (f) village stores; (g) co-operative society stores .

The typology is based on retail outlets in the Central Division , especially in Suva and the adj acent area. The first two types of shop, for instance, are largely confined to Suva. However, if allowance is made for the greater sophistication and specializat ion of grocery retailing in Suva than elsewh ere, the typology can be extended to other areas of Fij i. The seven types of outlet are examined below. 209

Expatriate-owned supermarkets are operated by the two chief expatriate European companies in Fiji, Burns Philp (South Sea) Co . Ltd, and W.R. Carpenter (South Pacific) Ltd. Burns Philp operates merchandise out let s at seven sites in Fij i, each with self-service grocery retail facilit ies as part of a general store. The supermarket concept is mo st developed in the Suva Burns Philp store where a recently constructed supermarket is a sect ion of a centrally-located department store. W.R. Carpenter retails through eleven Fij i outlets. Two are in central Suva operated by the sub­ sidiary Woolworths Ltd on a cash basis . The others are operated by Morris Hedstrom Ltd as wh olesale-retail outlets that deal in cash and credit, with foodstuffs an d associated household products being sold in general departmen t stores. The two Woolworths stores and the Burns Philp 'Foodhall ' in Suva are the most developed forms in Fij i of the broadly­ stocked self-service store connnon in mo·st Western countries. It is not coincidental that this recently-introduced form of retailing has attained greatest development in Suva, where grocery retailing is mo st competitive .

The expatriate-owned supermarkets stand apart from other retail food outlets not only in terms of ownership and business methods, but also because of their central location two are adj acent to the Suva �arket and bus stat ion and the third is within a couple of minutes ' walk - and the scale of operation ; each has an estimated gross annual turnover, in groceries , of more than $3 million .

As well as being similar in design to supermarkets in Australia, the home of Burns Philp and Carpenters, the outlets are operated on the upper circuit business principles common to Western economies. Employees are on regular wages or salaries and are unionized. Sales are cash only. Strict operat ing schedules are followed. Sales techniques include 'draw items', special-sales, compet itive gimmicks , intra­ store public address systems and media advertising. Stock and sales are continuously mon itored and stock ordered in anticipat ion of customer demand. No other shops in Suva or elsewhere have a stock range that approaches that of the supermarkets. The range of stock available in the super­ markets - including fresh vegetables and fruit , dairy products, bas ic andlu xury grocery lines and common household items - makes possible one-stop shopping.

The expatriat e-owned supermarkets are more than maj or retail outlet s. They have evolved from general purpose, 210 largely pas sive wholesale-retail outlets to be at the front of a competitive grocery trade. In competition between themselves and with locally-owned stores for an increasingly mob ile and richer clientele, the sup ermarkets have become pace-setters in retail innovat ion . Some innovations, such as prepacked products, advertising, sales and draw items , have been adopted by other retailers . Moreover, price com­ pet ition between expatriat e firms has resulted in the consumer becoming more price-conscious and the more competitive retailers offering cost and service incentives. However, apart from Suva and the larger towns , the expat riate-owned grocery outlets continue to provide the traditional general wholesale-retail service . But some of the supermarkets ' innovations are reaching even these towns which suggests the far-reaching consequences of the trade practi ces of the expatriate-owned supermarket s.

In Suva, and to a less extent in all towns, there is some hierarchical development of retail outlets ranging from local , isolated general stores , through neighbourhood and commtmity centres to central , wide-drawing outlets. An impression of differences in patronage of these outlets is available from a comparison between custom at the expatriate­ owned supermarkets in central Suva and three shopping centres in Suva. One is a neighbourhood centre in a low income area (Milverton Road, Raiwaqa) ; another (Flagstaff) has a wide range of functions and a high-access location and draws customers from a considerable area; the third, Toorak, is intermediate in respect of funct ion and access (Tab le 3.23) . 38

The trade of the expatriate-owned supermarkets differs in many ways from that of the shopping centres. The super­ markets are Suva 's only retail food shops that draw from a regional hinterland. They are used for casual supplies, snacks and even meals by people living or working in nearby areas, and they are a main source of groceries for families living throughout the greater Suva area. Expatriate Europeans and wealthy local people do a large part of their food shopping at the supermarkets. At the same time , the stores have a significant poorer clientele who frequently spend small amounts while in the central city area ·for other reasons, particularly for work or to visit the market . Around one-half of supermarket customers spend less than $5 on a visit. The mean amount spent is $8. 04 - $7. 31 on food­ stuffs. Sales at stores serving a neighbourhood clientele are frequently less than $1 per person and are often about 35�. 211

Patronage at shopping centres and expatriate-owned supermarkets in Suva

Super- �ping centres Patronage features markets Milverton Rd�staff Toorak

Per cen t of customers male 40.4 45 .8 47.3 52 .2

Position in household head 44 .9 10.3 29.4 27. 1 wife 44 .3 10 .9 28.2 30.6 child 6.0 68.5 20.9 30.2 Occupation of household head professional 37.2 12 .5 38.0 19.5 semi-professional 21 .1 20.0 16 .6 8.3 skilled, semi-skilled 25.5 29.0 22.0 39 .0 un skilled 10.2 23.5 7.5 21. 7 unemployed 5.0 n.a. n.a. n.a. Race Fij ian , other Pac . Isl. 37.9 61.2 46 .0 38.8 In dian 39 .5 29.4 31. 6 51. 5 Ch inese 4.0 1.0 4.4 7.0 European 12 .0 9.7 1.0

Visit for home consumption 94 .3 97 .0 93 .8 89 .4 Distance of residence from shops <.4km 82 .4 43 .5 76 .3 2.1 .4-. 8km 15 .6 23.4 6.6 . 8-1. 6krn 15 .6 1.5 13 .7 4.5 >l . 6km 82 . 3 18 .3 12 .4 Means of transport foot 9.4 95.6 51. 9 78.9 bus 58. 8 1.0 7.4 5.7 Last time h'hold memb er bought he re today/yesterday 25. 7 79 .7 4 7. 9 65 .8 within week 77.2 82 . 7 54.0 69.0

Payment by cash 100.0 98.0 91.1 92 .5 Main source groceries this centre n.a. 33. 7 35 .6 61.0 central Suva 77 .1 62 . 8 42 .7 24 .8 Main source fresh fruit and vegetables market 65 .0 85.0 78.2 85 .2 (+ 16.4 part mkt ) Money spent on food <$2 26.9 17.4 4.8 6.6 $2-$5 26.4 33 .3 16.7 17.0 $5-$10 23.5 23.9 16.4 27.4 >$10 23.4 25.4 62 .1 47.0

Products bought by more cereals , fresh vegs , bakery products, fresh milk , than 20% customers bakery prods , dairy fresh me at . prods , sugar/salt , oil/ fat , poultry/eggs , fresh/ froz. meat , canned fish , beverages, fresh milk. · -�� Note : Figures are percentages of valid responses in each category - out of 824 interviewed at the supermarkets , 203 at Milverton Road , 340 at Flagstaff and 301 at Toorak. Source : Patronage surveys at three ex patriate-owned supermarkets in Suva and three shopping centres, Jlllle 1976. 212

Another difference between patronage of expatriate­ owned supermarkets and local shopping centres is the distance customers live from the place of purchase. Ninety per cent of customers at suburban shopping centres live within one mile of the shop : the comparable rate for supermarket customers is 17.7 per cent. Differences in residential location are reflected in transport means: almost three­ quarters of shopping centre customers walk to shop compared to 9 per cent of people purchasing at supermarkets.

Other dissimilarities in clientele are race (European purchasing concentrated in the central supermarkets) , the shopping 'team' (husband and wife accollllt for ab out 90 per cent of supermarket patrons compared to about 40 per cent at shopping centres), frequency of visit (a household memb er had been to a supermarket within three days of interview in 25.7 per cent of cases, compared to 62 per cent for clientele of shopping centres ), credit access (not available at super­ markets but used by 6 per cent of centre customers) , socio­ economic status of the household head (57.3 per cent of household heads of supermarket shoppers are in professions or skilled trades, compared to 39 .8 per cent at shopping centres), dependence on central city shops for groceries (the 'main source ' of groceries for 77.1 per cent of super­ market customers , compared to 41.3 per cent at shopping centres) and dependence on the market for fresh fruit and vegetables (82.4 per cent of shopp ing centre patrons cited the market as the 'main source' of fruit and vegetables compared to 65.0 per cent of supermarket patrons) .

Supermarkets and suburban shopping centres have differ­ ent fllllctions in the supply of food. Not all shopping centre cust omers patronize the supermarkets and when they do it is to purchas e different items than they would at the centres. More than 30 per cent of supermarket customers buy some 'cereals ', 'fresh vegetables', 'bakery products', 'dairy products other than milk' , 'beverages ' and 'fresh and frozen meat '. At shopping centres only 'bakery products' were bought by more than 30 per cent of customers. 'Milk' and 'meat ' we re the next mo st popul ar items , each purchased by 20 per cent of patrons. While at supermarkets eleven different types of food were purchased by more than one-fifth of customers, only three items had such support at shopping centres. A reason for the difference is that most shopping centres are sources of basic groceries and are used only to top up immediate needs of other items . Purchases are small 213 and frequent . Some shops in centres provide more specialized services , e.g. liquor sales, but concomitantly most meet immediate local demands .

Specialized food retailers are chiefly butchers , although there is increasing specializat ion in shops retailing fresh fruit and vegetables. But cher shops differ from other retail food outlets in the degree of specializat ion and the develop­ ment of chain outlets.3 9

There are licensed but chers in all centres save Vaileka , Korovou, Navua and Levuka. Ten of the twenty-five but chers are in Suva, including two halal but chers. Ap art from Lautoka (three) and Suva, no centre has more than two but chers. Butchers are the one local food retailing activity in Fij i with some development of retail chains. Twelve of the twenty-five licensed butcher shops are operated by four chains , two in Suva and the two in the Western Division , one of wh ich operates in four different towns.

No retailer deals exclusively in fresh fruit and vege­ tables, but at least one Chinese merchant in Suva has a business based on wholesaling an d retailing imported and local 'temp erate' fruit and vegetables. The expatriate-owned supermarkets also have fresh produce sect ions.

It is apparently becoming more common for general grocery shops to sell some fresh , mainly local , produce. Bananas, oranges , coconuts, dalo , pumpkin , ginger , chillies and eggplant are often sold. Perishable leafy vegetables are also sold occasionally. The produce comes from the shopowner's garden or is bought from pro ducers or middlemen at the shop or market . Often vegetab les are left at shops by a friend of the storeowner who supervises sales as a favour .

The expansion of grocery shops in to sales of local fresh produce is most marked in Suva , where about twenty of the 166 shops retailing food sell produce. Outlets in central Suva deal primarily with local or imported temperate produce and , as sales are directed to the expatriate and other wealthy trade, are not in marked compet ition with the market . Elsewhere , supplies are limited and erratic and trade is aimed at the urgent needs of neighbourhood customers for whom the market remains the prime source of supply . 214

Suburban neighbourhood and community shops range from moribund general shops to aggressively-managed local super­ markets and stores , many of wh ich have a liquor off-licence. Aside from the scale of operation , larger neighbourhood and smaller community shops are not readily distinguishable. They share common functions (such as supplying basic groceries an d household goods and acting as a credit source), ownership forms and business methods , and even their custom may be from either the local ?eighbourhood or the broader community. Some aspects of the trade activity of twelve neighbourhood and community shops are described in Table 3.24.

Suburban neighbourhood shops do almost all their trade serving the basic grocery and non-food requirements of the immediate neighbourhood . The shops may be isolated , in a small neighbourhood shopping centre or in a widely-drawing regional centre . Even when in a community shopping centre or on a high-access route, neighbourhood shops have limited stocks and frequently an unprepossessing appearance. Store owners are passive , patronage limited and trading periodic. The shops are busy from the time of bread and newspaper deliveries (around 6.30am) until soon after 8am when most schools and employmen t commence. Late morning sees a small trade revival as households purchase lunch needs but the shops are not really busy again until children get out of school and preparations begin for the evening meal.

Almost all suburban neighbourhood shops are owned and operated by Indian or Chinese families . Assistance from outside the family is rare. The shops can be operated by most times by one person and children help out in busy periods, mainly by keeping records as the owners are often functionally illiterate. Stock comprises a wide range of small non-food items as well as groceries and household cleaning goods. Groceries usually account for at least three-quarters of trade. The shops function as a supply equalizer for low income neighbourhood families by being a credit source to tide over between pay days .

Another important function of suburban neighbourhood shops is to act as a break-of-bulk point . As well as breaking wholesale units, such as lOOlb sacks of rice and 224lb sacks of sugar, into common consumer sizes, they also sell items in units smaller than the manufacturer's standard retail pack. Table 3. 24 ------Trade cha racteristics of twelve grocery shops in or near Suva

Est. Customers Sales Daily Sales Sales in Av. no. Av . sale weekly from for Shop turn­ in 'basic' customers per Commen ts turn-,. local credit over food foods daily customer over area (approx. ) $ $ % % $ % % 1 10-35 120 67 90 50 0.8 7 100 5 Owned by 'Pranj ivan ' (see text); on outskirts of Nausori oppo- site store 3. 2 20-75 220 84 81 10 3.39 100 20 Owned by 'Krishna ' (see text); 24km from Suva , opposite store � 3 20-80 280 22 39 55 0.79 100 0 Indian-owned; on outskirts of Nausori opposite store l; liquor licence . 4 6-100 290 47 69 50 0.9 0 85 0 Indian-owned; on King 's Road at Nasinu; adj acent to store 10. 5 30-75 370 64 72 70 0. 82 100 4 Indian-owned; in isolated rural­ urban settlement llkm from Suva. 6 50-160 400 63 67 75 0.8 3 100 1 Indian-owned; in small shopping centre in prosperous Suva suburb. 7 45-110 430 47 63 145 0.45 90 10 Owned by 'Lai' (see text); in a prosperous Suva suburb , isolated . 8 80-100 520 67 51 110 0.7 9 85 5 Chinese-owned in an active Suva shopping centre . 9 30-380 650 72 58 25 3. 92 85 n.a. Owned by 'Vinod Prasad' (see text); 24km from Suva, opposite store 2. 10 30-225 770 28 48 95 1.25 85 5 Indian-owned; adj acent to store 4; liquor licence. 11 100-350 1400 29 49 200 2.06 82 1 Indian-owned; in a prosperous Suva suburb ; isolated; liquor licence. 12 60-350 1500 30 52 140 3. 14 65 1 Owned by 'George Sultana' (see text); in prosperous Suva shop­ ping centre ; liquor licence. Source : Survey of transactions in shops over six consecutive day s in December 1976. (See note 40 for me thodology .) 216

One pound packs of butter, for in stance , can be purchased in 10� slices. Single cigarettes are available at 3� each . Kerosene is sold by measure into the customers ' bottles , the only break-of-bulk fo r kerosene which is delivered to shops in 44-gallon drums or larger amounts. Supplies are purchased in cash or on credit . Turnover, and the extent to which expansion is possible , is largely a product of store location , particularly access to a relatively wealthy clientele. In Suva, a poorly located neighbourhood shop may have an annual gross turnover under $6000 ; turnovers up to $10,000 are not uncommon .

One neighbourhood shop in Nausori was studied in some depth. 40 The shop is located on a main road leading from the town but serves mainly local semi-professional and skilled Indian and Fij ian residents rather than through-traffic . Few customers live more than a hundred or so metres from the shop. Opposite the shop is another with a licence to sell liquor; this store has a considerable non-local trade.

No more than fifty customers come to the shop in any one day. It opens soon after 6am and closes an hour or so after nightfall. Two-thirds of sales , by value , are of food, and 90 per cent of these sales are of basic foods such as sugar, salt, flour , sharps, rice , onions, potatoes, garlic, .dhal, tea, milk, bread, butter and canned fish and meat . The main non-food items sold are cigarettes, soap , ice cream and sweets. For a neighbourhood outlet the store is unusual in that newspapers are not sold and sales of kerosene are negligible. The average sale is 37�, with each customer buying an average of 1.4 different items . About 5 per cent of sales are on credit . Daily gr oss turnover ranges from $10 to $35 for a weekly rate of about $120 or little more than $6000 a year. As the shop is open for around eighty hours a week and assumin g that net profit , not costing labour, is probably no more than 10 per cent of gross sales - it is considerably less on price-controlled items that are the basis of trade - the return on labour is about 15� per hour.

The shop fotmder , a Guj erat i man named Pranj ivan , came to Fij i shortly after Wor1d War II. After working for a short time for another man as a tailor, which was his previous occupation in In dia, Pranj ivan established his own tailoring bus iri.ess in Suva. In 1952 he moved to Nausori to be near other merchants from his home area in India . After a couple of years he obtained a 75-year 'native lease ' on some land 217 and decided to build a general store . The 6m x 18m single storey concrete shop, with family residence, was built in 1955 with $1000 of Pranj ivan 's money and a commercial bank loan of $5000 . Bus iness was quite good as the shop was the first in the immediate area, but the establishment of better shopping facilit ies in Nausori and the opening of other grocery shops near his store soon caused a decline in business.

Af ter operating the shop for almost five years, Pranj ivan became ill . As no one in the family was available to continue running the shop, it was rented to another person . In 1971, the year before Pranj ivan died, the family 's second eldest son, the eldest son having left the family some time before, connnenced work in a Suva bank and decided he would also take over the shop. While he works in Suva , a younger sister - the only one not yet married - looks after the shop , being helped occasionally by her mother or two younger brothers who live at home . The son and daughter, both of whom com­ pleted secondary school, look after all shop accounts , but hire an accountant to complete annual tax returns.

Supplies are purchased from salesmen who collect orders and later deliver to the shop. The shop has thirty-day credit accounts with its main suppliers. Chief competition comes from the shop across the road which has a similar grocery stock but the added draw of liquor sales : the weekly turnover of the competit ion approaches three times that of Pranj ivan 's shop. Pranj ivan 's son admits that not much money is made from the shop and that there is little chance of expanding business, especially since he is busy with a full­ time job. On the other hand, the shop does help the family when they are short of money.

As noted above , suburban connnunity shops are in many ways similar to neighbourhood shops, in terms of location , ownership and organization, the difference being they are operated on a larger scale. They include the mo st prosperous grocery shops operated by local people, and some even compete with expatriate-owned supermarkets. A range of shops com­ prise this category , including neighbourhood supply shops that attract a broader clientele because of their location, supermarkets and shops with additional specialized functions (such as selling liquor) .

Most suburban community shops are family bus inesses owned, operated and staffed by family memb ers . Only a handful are registered companies. Ownership is almost 218

exclus ively Indian or Chinese. The shops have developed generally from stores with local , limited trade . The larger shops, wholesale-retail, supermarket and liquor operations employ non-family labour but none is tmionized . While multiple grocery store ownership by local people is unconnnon , suburban connntmity shops are the mo st likely retail outlets to form a 'chain'. The usual route of connnercial expansion is to move into import ing and wh olesaling via a wholesale­ retail outlet rathe r than to establish a duplicate retail outlet . There are around twenty suburban connnunity shops in Suva, excluding wholesale-retail stores.

An isolated suburban connnunity shop in a relat ively wealthy residential area can have an annual gross turnover of $20,000 to $30,000. Shops on major thoroughfares have higher turnovers, whilst shops with a liquor licence an d/or supermarket facilities in a high-access location can have an annual turnover in excess of $125,000. The few larger supermarkets that import some stock and make home deliveries have gross annual turnovers of arotmd $300 ,000, about 80 per cent of which comes from grocery sales . Because of the range in form of operat ion of connnunity stores, it is useful to review two case studies of different levels of activity.

The first example is a store operated by a thirty-five year old Chinese lady on one of the main roads leading from Suva to the northeast . The father of the owner, Lai , operated a small pig and general market farm, as well as a numb er of leased shops and a cafe, in the Suva area, before he became ill and stopped working. For some years the mother kept the family by selling produce in Suva market, but in 1961 , with the father's en couragement, it was arranged that the family would buy the shop and the att ached house from an tmcle. Half the price ($8000) was raised within the family. The remainder came from a connnercial bank loan. One year after leaving high school, wh ere she failed Form 4 exams , and after searching for employment because she was not keen to work in the shop, Lai began operat ing the shop with her elder sister ; their father gave advice and encouragement but little other help because of his illness. After nine years , the elder sister got married and moved to another town . Since that time Lai has run the store by herself apart from periodic help from a brother with the accounts and tax returns. Like practically all other local retailers , wholesalers and importers , Lai has no formal commercial training. 219

Lai works alone in the shop every day of the week from 6 or 6.30am until 7.30 or 8pm, with an hour or two off in the early afternoon . Stmdays are particularly busy as then the two shops that offer most competit ion , each a couple of htmdred metres away on the same road , are open only for a limited period. All stock is purchased at the shop from wh olesalers; Lai rarely goes away from the shop . Stock comprises groceries, household cleaning items , cigarettes, kerosene, sweets and a few vegetables from her or her neighbours ' gardens. An, upper shelf is stacked with cooking pots: they were bought a long time ago by her mother and have proved un saleable. Lai herself now does all the stocking on the basis of lines wh ich sell quickly and is cautious about introducing new ones.

Over the week, the shop is visited by an average of 145 customers daily, each of whom purchases slightly fewer than two items . The average out lay per visit is 45�. Daily turn­ over ranges from $45 to $100 and weekly turnover is a little more than $400 (annual rate, $22, 400) . Almost one-half of sales , by value , are food. If a 10 per cent net profit is assumed, and Lai's wo rking week is taken to be eighty hours, she has a gross income , prior to expenses an d taxation , of a little more than 50� an hour .

Forty per cent of food sales are in milk, bread and dry biscuits and a further 23 per cent are in other 'basic' items such as sugar, salt , flour, sharps, rice, garlic, onions, potatoes , dhal and tea. Over 80 per cent of non-food sales are in newspapers, cigarettes, kerosene, sweets and ice cream. The shop is unusual in that there is a relatively large non­ local trade (about 10 per cent of sales) and a high proportion of credit sales. Lai finds it difficult to limit credit . She has called the police over dishonoured cheques but has not made other legal moves against debtors. She is well aware that too much pressure on debtors can drive them away , leaving no hope of recovery and one less customer. In terms of breaking bulk and supplying immediate needs through a multitude of small sales, Lai's shop is essentially a neigh­ bourhood store, although in a reasonably wealthy neighbourhood. Access to through-traffic raises trade above the leve.l of mo st neighbourhood stores to the lower financial margins of suburban connntmity outlets.

A more pro sperous community shop is the supermarket run by George Sultana , a fifty year old part-European high school graduate. The medium-sized supermarket is a converted shelf- 220

and-counter store on a high access junction a few kilometres from central Suva . The owner ob tained a share in the shop in 1965 and a few years ago received a loan from a commercial bank to buy the share of the partner . At that time , the bus�iness was incorporated as a limited liability company . Sultana worked in a variety of clerical and managerial jobs before getting involved in retailing .

The shop is part icularly busy as it sells liquor , makes some home deliveries and does a little wholesaling to other stores . A number of basic gro cery and liquor lines are imported directly. Sultana is full-time man ager and operator, overseeing ordering , costing, stocking, indenting, customs clearance and general operat ions . He has two full-time shop assis tants and a numb er of paid part-time helpers . His wife holds a full-t ime job elsewhere and handles the correspondence; in the early years with the shop the wi fe's income helped meet household expenses . An accountant is hired to prepare tax returns .

The shop is open from 6.30am to 6pm Monday to Friday, apart from a 1.15 to 2.15pm ltmch break. Paid employees start work at 8.30am. On Saturday the shop closes at 12 . 30pm and does not open until Monday . Some residents of the immediate neighbourhood use the shop for basic food supplies . A mainstay of trade is liquor and grocery sales to employees of government and nearby institut ions who shop by car . In addition to good grocery stocks and the liquor licence, an important att ract ion for these shoppers is the fact that the adj acent shop is one of the few butchers out side the central city. Sultana is considering retail ing fresh fruit and vege­ tables to at tract greater patronage . General homeware is carried but sales are in significant compared to those of groceries and liquor. There are seven other general stores in the cent re . To attract the non-breakfast trade customer, Sultana has a more liberal ch eque-cashing policy than most food stores but risks are high . There were bad debts of more than $4000 in the past year , a large share of wh ich came from dishonoured cheques .

Data on the trade of Sultana's shop are not as complete as for other stores studied. It appears that about 140 cus tomers use the store daily and spend slightly more than $3 each , giving a weekly gross of about $2500. Less than one-third of sales are food. One-half of food sales are in basic items , including sl ightly mo re than one-quarter in 221 milk, br�ad and dry biscuits. Sixty-five per cent or so of customers live in the immediate neighbourhood. Few sales are on credit .

Rural general stores are ope rated by Indian and, less commonly, Ch inese merchants at dispersed and often isolated locat ions . In terms of trade volume and residential location of customers , the stores fall in the middle range of neigh­ bourhood and community shops. All handle some neighbourhood trade but some have broad trade hinterlands. In form, function , ownership and mode of operation , the rural general stores hardly differ from suburban neighbourhood and community shops. The stores are invariably attached to the residence of the storeowner, are operated by the owner and his family and depend largely on the grocery trade although they frequently carry broader non-food stocks than urban counterparts.

After the initiative of the storeowner, the scale of operation depends largely on the ty?e of clientele served. Stores with a clientele from beyond the immediate residential community can do well. Stores on main roads, serving a prosperous community or at an incipient nucleus of settlement may have a considerable trade. Shops at breaks-of-transport are often re-wholesalers for more distant stores. Annual gross turnover of rural general stores in the G.entral Division can reach $60, 000, but the mean probably is around $20,000 .

Two rural general stores about 24km from Suva were studied. The stores are a little unusual in that they are situat ed in a . relat ively wealthy area where the populat ion is more dependent on wage labour than agriculture for income , since rice and other farm product ion is primarily for home consumpt ion . They are opposite one another across a narrow road. Both are owned an d operated by Indians.

The smaller of the shops has been functioning for ten years . The owner, Krishna, is ab out forty-five years old and married with five children . After completing primary school and before getting married, he worked as a sales assistant in a Nausori shop . For almo st ten years after marriage he continued this work before deciding to attach a shop to his house on the small piece of freehold land inherited from his father. Capital for construct ion ($800) came equally from personal savings and a loan from relat ives, who also 222 helped build the shop. When the store was completed in 1967, Krishna resif1ed from his job and has operated the store ever since . 4 In addition to operating the shop with the assistance of his wife and daughter , Krishna also grows enough rice on his land for their domestic consumption . The only outside help is an accolllltant paid to compile annual tax returns. Four of Krishna's children are still at school; the fifth, the eldest , has a part-time sales position in Nausori. The only other source of in come in the family is a taxi Krishna owns and operates through a hired driver in Nausori.

While Krishna had not operated a shop prior to building his own , he picked up pract ical knowledge of business during his fifteen years as a sales assistant. Once the shop was completed his biggest problem was to attract customers from the longer-established shop across the road, a task made especially difficult by the limited stock which was all he could afford to carry . 42 As ide from the continuing problem of competition, which he has tried to counter by applying for a liquor licence, a request which has thrice been refused , Krishna's problems are those connnon to small shopkeepers . By allowing regular customers to purchase on credit , he is frequently without sufficient cash to purchase new stocks . As many of his customers are dependent in part on agriculture , their income is spasmodic. Krishna finds it difficult to refuse credit requests from such people. Usually he can obtain thirty days ' credit from wholesale suppliers in Nausori but occasionally he has to plead for an extension of credit .

While Krishna finds storekeeping a secure occupat ion which enhances his self-respect , he sometimes feels that he would be better off in wage or salary employment, as prac­ tically all store income is used for family and educat ional expenses. To make matters worse, he feels that few of his customers appreciate that only very small markups are per­ missible tmder price control , or are possible in the competi­ tive grocery and general household goods trade . While there are not 'price wars ' between his an d the opposite store, the situation is competitive in that prices can be compared and customers have a readily accessible alternat ive source. For fear of losing customers , Krishna cannot afford to close the shop for any length of time between the 6.30- 7am opening and the 7.30-8pm closure. 223

Krishna's trade is ma inly in basic grocery lines, although he has a freezer which enables him to sell frozen goods. All sales are retail and all customers are residen t in the immediate neighbourhood. An average of ten customers a day come to the shop , the number ranging throughout the week from five to sixteen. Daily turnover is from $20 to $75, weekly turnover being about $220 or around $11, 400 a year. One-fifth of sales are on credit , each customer buying an average of 4.6 items per visit. Goods sold are primarily groceries (83.7 per cent) ; 81 per cent of food sales are in basic foodstuffs.

Vinod Prasad , Krishna's opposit ion , has an enviable connnercial pedigree. In the early 1940s, Vinod and a brother obtained a hawker licence and during the mornings and evenings sold rice , soap and other items around their settlement and in a nearby Fij ian village . During the day they farmed rice and sugar, whilst their father worked as a carpenter. They converted the verandah of their family hous e into a shop in 1943, thus establishing the first shop in the area. Their commercial activities continued to expand. In 1950 they purchased a shop from a bankrupt Chinese merchant , then dis­ mantled and recon structed it at the ir home . The brothers operated the shop together until 1961 by which time they had saved enough to purchase a shop in a Suva suburb .

While his brother moved to the Suva shop , Vinod remained at the original shop which in 1973 was replaced by a modern­ design con crete self-service store, the only one in the area and one of the few outside any town. The self-service format was adopted to attract and please customers. All capital to construct and expand shop facilities, as it was to build the original shop and for development expenses, was generated by trading. The shop is operated by Vinod and a son; another two sons are working with their uncle in the second shop.

Vinod's shop is more than a basic retail grocery outlet : groceries are delivered by van to customers ' homes, some wholesaling is undertaken and some food lines are imported direct . Sometimes boys are employed to help Vinod and his sons pack and deliver orders. In addition to the shop , Vinod has 24 hectares of land, one-third of which is used to grow rice for home consumption and the remainder is grazed by cattle. Vinod also leases a stall daily at Nausori market . Each Saturday his son and some friends occupy the stall and sell onions, potatoes, garlic, spices, dhal and rice. In conj unct ion with his brother , Vinod imports most groceries 224

that are not locally manufactured.

Over the monitored week, Vinod's shop had a turnover of about $650 or an annual rate of nearly $35 ,000. Including customers to whom deliveries were made, an average of twenty­ five people purchase at the shop daily , about 85 per cent of whom live in the immediate neighbourhood. Daily turnover ranges from $25 to about $400, almost one-quarter of sales being on credit . Compared to the adj acent store, a smaller proportion of sales is in foodstuffs (72 per cent) and basic foods are less significant (58 per cent of food sales compared to 81 per cent) . Largely through service innovations like home deliveries and self-service facilities , as the first store in the area Vinod's store has been ab le to expand trade an d withs tand the competition of one adj acent store and three other outlets in the area.

The studies of Krishna's and Vinod 's stores suggest some common features of retail outlet establishment and competition . Entry into retailing has few restrictions, which can lead to an over-supply of outlets. Patronage of stores is to some extent unchanging, but aggressive entre­ preneurship can expand trade. Proven expansion techniques include generous credit policies and provision of broad stock rather than price competition . Moreover, as the over­ heads of family stores are minimal and alternative employment uncommon , it is unusual for a shop to close in the face of competition. Given tha� particularly in rural areas, possi­ bilities for market expansion are limited, it is often more profitable to establish a second outlet elsewhere than to attempt to capture all local trade.

Village stores are operated by Fij ians in Fij ian villages by an individual or occasionally by a kin group or even on a basis other than kinship , such as a youth or women 's group . Stocks are usually minimal and comprise little more than basic foods, cigarettes , soaps and kerosene. The stores are non-residential buildings of roughly-hewn or used wood and iron . Facilities are minimal : few stores have more than a cotulter and shelves.

The village storeowner is often helped by relatives and friends in the shop but paid help is not used. Storekeeping is not a full-time business and the owners plant food gardens for family consump tion and perhaps sale, an�- may have other irregular local sources of income . Village stores serve the village in which they are located; there may be three or 225

four stores in a village , each with a clien tele fairly well defined on kinship lines. Even where a village is on a road and the store could attract through traffic trade , this is rarely sought and is anyway minimal, in part because of store schedules. The stores are open from soon after dawn until 8am or so, an d again from late in the afternoon until after dark when the evening meal has been eaten . During the day the operator or his deputy may open the store upon request but daytime trading is negligible.

Many village stores have erratic trading activities. One reason is that demand for groceries and other purchased goods is very elastic. Villagers grow all required staple foods and often need to buy only salt, sugar , matches and sometimes kerosene. Flour , tea, canned mackerel and corned beef , dry biscuits, rice , baking powder and cigarettes are also purchased but they are luxuries which are foregone if cash or credit is unavailable. Moreover , because of costs of goods , transport costs and sometimes bus iness inefficiency, prices are ofte n considerably higher than in towns and even in other nearby non-village shops. Trips are rarely made to towns solely for shopping, but food and other supplies are purchased on such trips.

A factor further contributing to erratic trading is that the store operators are sometime s as adverse to receiving credit as suppliers are to granting it . Supplies are usually purchased from wholesalers, or from retail grocery outlet s in the larger centres , only when there is adequate cash to cover purchases . Trading is frequently in terrupted by periods of inactivity while extra store sources are tapped to raise capital to purchase addit ional stocks . Stock purchased at any time is often small enough (perhaps a couple of cartons of goods) to be transported to the village on a public bus or in a taxi. Village stores rarely have specialized storage space. Records are kept infrequently. Owners do not prepare tax returns .

Turnover of village stores varies in relation to a number of factors, including competitive outlets, the general wealth of the village, and the level and consistency of stocking. In mo re remote areas where there is limited money in circu­ lation and where supplies are difficult to obtain , many village stores have weekly gross turnovers of less than $20, or around $1000 annually . Other stores , particularly those serving wage-earning populat ions, as around the city of Suva, have estimated turnovers exceeding $10,000. The mean annual 226

turnover of village stores is probably under $5000.

Village storeowners are not in an enviable commercial position. Aside from being poorly educated, they rarely

receive commercial training and operate in a distinctly _ non-commercial (or at least, a non-competitively commercial) environment . Cus tomer loyalty comes from ties to the store­ own er but if stock is unavailable another store is readily patronized . There is no price competition between stores in a village . Notions of cost accounting and pricing are hazy. In des cribing how selling prices are determined, storekeepers speak in terms of adding a cent or two to the purchas e price , in accordance with what they believe the market will bear. No formal exercise is undertaken to compensate for tran sport and other costs. In rural village stores, store proceeds are kept in the owner's house, so that what is not dissipated is used to purchase supplies as store stocks near exhaustion. It is uncommon to encounter village storekeepers who have financed capital developments or who make substantial purchases with store profits. Nonetheless , there is no shortage of vil lagers operating retail stores . In the Central Division , there are 266 individual Fij ian storeowners compared to 269 Indian and 96 Chinese storeowners . It is necessary to examine the mo tives leading to the establishment of village stores before it is possible to understand why there are so many.

Motivation to operate a store is as much service-oriented as en trepreneurial. In one form, it is service to the community by providing a convenient source of purchased goods . As one storeowner put it ,

I established the shop to cater for the haphazard

way of life of the taukeis - they sit down to eat and only then realize there is no salt , or they pour out the tea and then see there is no sugar . Of course the children are always sent at once to buy these urgen t goods .

Another aspect of service is that if they do not make money , at least the owner's family, or the broader group operating the shop , gets access to credit by supporting the shop . As the storeowner quoted above said , one reason for starting the shop was

to help the family out during times of difficulty; for example , at any Fij ian gathe ring, when they 227

are ab le to take goods on credit . Or , if they [family members ] are required to donate money to some causes , they normally borrow from the shop and they pay later. At the moment the shop is largely not profit-oriented, but stands as a kind of bank or security for the shareholders.

Factors affecting village store operation are apparent in the following interviews with two storeown ers from Lau who have stores in villages a few kilometres from Suva . Their situation is tmusual relative to the experience of people with stores in their own villages , and because the clientele is wealthier than in mo st rural villages , but nonetheless they do have much in common with other village storeowners .

Viliame Lomaloma and his wife from Lakeba, Lau, operat e a shop in Walua village on land they have been loaned through traditional channels by the villagers . Th e shop is successful in terms of trade volume and apparent profits. Stock is that which is basic to Fij ian shops - canned mackerel an d corned beef, evaporated and powdered milk, tea, sugar , rice , flour , dry biscuits, edible oil, dripping and bread, in addition to connnon non-food items . The shop was built eleven years ago when Viliame came to Suva for a visit from Lau. He intended to leave the store with his brothers in Suva, planning that income would be sent regularly to him in Lakeba. However , it did not work that way :

I was not educated in any popular school. At home I am the leader of a mataqali. Everything to do with their lives was my responsibility. I did everything I could to improve my mataqali., but after that I left home to look for a business [opporttmity], wh ich brings me to wh ere I am now

•.. I just came to Suva, built the shop , loaded it with groceries and left it with my brothers while I returned home . At the end of every year for seven years I came here and changed the storekeeper. This happened for seven whole years. After those seven years there was nothing worthwhile or of any value that came out of the business .

[His wife had received a similar amotmt of formal education (up to class 6 in primary school) , but even in Lau before she came to Suva in 1973 she had l I shown commercial initiat ive . As she said ...] Al 228

did was to look for things that I could do for sale, like fishing, weaving mats and doing other things that could bring us some cash that could

be used in the household ..• I learned how to sew clothes and this meant we could stop buying clothes for the children ... I bought a pair of scissors and cut their hair at home - all in the quest of not spending money tmnecessarily.

When I entered the shop, our second eldest son was the shopkeeper . As he could not finish his studies , he stayed home and looked after the shop . All that was in the cash box was $2. 00 and there was nothing left in the shop except six small packets of Omo . We had left $30 with the eldest son to put a deposit on the refrigerator, so I decided to take back that money and buy some items that I knew would be sold quickly, like sugar, salt and kerosene. After one week of trading I had back that $30 and a profit of $50. I used that profit the next week and things began to grow. After three months I was satisfied with what I had achieved. By Christmas all was rtmning well . Educat ion fees had been met and $150 had been deposited in the bank. Then a child got sick, so I asked my husband to come to Suva and help run the business for I could not do it alone. He arrived and soon we made up a plan for our business. All I can say is that although I am not well educated, just a villager, I have been able to build a business right from rock bottom for only $30.

[Husband and wife have now operated the store for four years . Income from the store has been adequate to cover the education expenses of their three adopted children still at home and four grand­ children. Viliame continues ...] My family is quite a big one , there are nine of us altogether today and the cost of support ing this family - you can work it out - is staggering. All money for the family comes from this shop alone; what I buy an d sell an d what we eat are all taken from this shop. After just one and a half years I built myself a good and beautiful house that cost me $1500 - that was possible only through this small room [the shop]. The furniture and everything you see arotmd you are all directly the product of this 229

small business. We have multiplied the original $30 by so many hundred times . I'm a bit hard on people requesting credit . Only those I can trust I may allow credit . I don't even like to give credit to my brothers , the real ones , because if things go bad it is only me who suffers .

[Goods are sought from wholesalers in Suva but even if the wholesaler does deliver, it is only to the end of the road, about 0.8km from the store. Viliame himself has to carry the goods, wh ich include sacks of rice , sugar and flour weighing 48kg and more , to the store , sometimes through kn ee-deep mud . Viliame buys and prices art icles so that they move quickly. ] I buy items , consult my arithmetic table and try to put a few cents interest on . It might only be 1 or 2 cents but I have this small increase because I want to sell quickly and bring in some more stock. I'll only stock with commodities that I feel will be quickly sold.

One mu st suffer, sweat , be gossiped about . This must come first before we reap the fruits: they are the roots of success - they may be bitter but the sweetness of the fruit is worth mo re than any pain . The main reason for starting the shop is that now is the time of business and progress. To survive in this society, one needs money . The main foundation of success is gut s and determin­ ation . If we start a business with this we will succeed. If we don't have this , we cannot hope to have a business. This work needs courage , a manly stand, good judgment, ab ility to take risks . Now is the time for progress. Everybody else has moved, but wh at about the taukeis? We also can be like them [the other races in Fiji] . We have the same ab il ity. Why shouldn 't we also be able to achieve their goals? We can if we try harder. Let 's stand up and walk hand-in-hand together.

Jone Yaqona (also from Lakeba) is a 51-year-old widower who stays with his two sons at Laucala village. Jone an d his sons do not have land for food garden ing and consequently are dependent on the sons ' wages and income for food supplies . Jone describes how the store came to be established and how he operates. it : 230

I used to stay in the village [in Lau] in a rented house with some relat ives. Both my children had left home to come to Suva and as there was no one to look after me in my sickness [he is partly crippled] , I decided to come to Suva . Here the rent was too great so I asked my sons to try for a privat e house . That 's how we built this house. Last year we started to plan this shop. One after­ noon I asked my children if they were willing to leave some money aside for a shop . As the result of this conversat ion , by 31 March 1976 we had built this shop . The plan came ab out when I told my sons that I'm weak an d can 't be engaged in any paid jobs. So I'm depending on them here in Suva where everything mu st be paid for. Before having this shop, I noticed that most of my sons' wages were spent and that only a very small port ion was being banked . Maybe if we had a shop , I thought , the money will make things a bit better. So when they decided to set one up we built this one. It cost $187 to build and supply .

I only stock the items for daily consumption . that are most needed by those people living around here. I buy these items from other shops. My customers are only those people staying here around me . I sell right from morning till night but if I'm tired I might call the children to come and help . To get the stock I go around from shop to shop [in Suva] and only buy from wh ere the needed items are cheap as this store has just started and our capital strength is very weak . After I buy the commodit ies I search for a taxi to transport my articles here. Taxi fares vary . Some charge $1.20 and others $1. 10 ... Transportat ion is difficult. The main road is a bit far off and if boys carry the commodities from the road they 'll demand to be paid. If we just ask them in traditional style, they ref use .

Ab out the nuisance of credit . Yes , some of them [customers] are good but others are very bad. When I tell them to pay for their credit they refuse but when in trouble they come back. But some of them are true - they pay for their credit at the right time . Only after being three or four weeks overdue , will others come and pay . I tell them, 'Do you 231

think I'm going to steal items from the companies ? If you don't pay , I won't be able to get mo re stock, so I'll stop giving you credit .'

The entire daily running of this shop is done by none other than this old man [himself] . No educated person is running it , the small thing of a foolish man. I never attended school. I only do what my min d thinks fit . For example, if mackerel were too expensive , I would not buy it . I would turn to other cheaper articles. So i buy stock in small amounts . I could buy mo re but because I do not know any arithmetic I might lose. I really want to expand this shop, but I'm uneducated and I might run at a loss if I expand. Another reason is that the mo st essential thing for business, its 'mother ' [money] is lacking. Since I first started the shop , I haven 't known any problem. Maybe it 's because the business is so small ; maybe after another year or two I might notice some problems . But if money continues to come in properly there will be no problem. Up till now there 's no prob lem, but with a lack of money I may end up in real big trouble.

Co-op erative society stores , the retail outlets of consumer and consumer-market ing co-operative societies , are an important feature of retailing in rural areas . In some ways the stores are similar to stores operated by in dividual Fij ians. Stock tends to be mainly groceries an d basic household requirements . The stores tend to be passive 'service depots' for the connnun ity rather than a retail outlet actively tapping non-local trade. Mo st shops are located out of urban areas and serve exclusively Fij ian customers. Co-operative society stores differ from village stores in that they have access to the services of the Departmen t of Co-operatives, in cluding training, general advice on connnercial practice , semi-regular auditing of accounts and access to co-operative who lesale facilities .

In 1974 there were 1002 registered co-operative societies , including 430 consumer and 375 consumer-marketing societies. Of these 805 societies that operate stores , less than 600 appear to be active . 43 Financial success of societies has distinct spatial features (Tab le 3.25) . -�4 While one-fifth of societies make a net trading loss, the proportion varies :i,etween J,.1. 2 pe·r cent of societies in the Eastern Division 232

Jable 3.25

Trade performance of consumer and consumer-marketing co-operative societies

Locat ion (Division)

Central Western Northern Eastern a Sales/mon th (Number of cases) (138) (103) (127) (169) Mean ($) 995.6 758.1 780.4 1011.6 Standard deviat ion ($) 1523. 3 975.2 1149.9 2041.3

Per cent societies with net prof it 81 .4 79 .1 62 . 7 88.8

b Societies with net profit (Number of cases) (114) (91) (79) (151) Maximum profit (%) 16 . 35 16.83 31 .23 30.56 Mean profit (%) 5. 71 6.56 9.06 9.58 Standard deviat ion (%) 4.13 4.46 6.84 6.15

b Societies with net loss (Number of cases) (26) (24) (47) (19) Maximum loss (%) 12 .12 12 .06 91.60 29.45 Mean loss (%) 2.67 3.22 12 .62 5.98 Standard deviation (%) 3.12 3.09 17.99 6. 50

Society with accounts incompletec 3 13 3 2

d Inactive societies 12 28 14 9

a Calculated over mo st recent availab le accounts period of each society, which averages 12 .5 (Eastern Divis ion) to 14.1 months (Central) , generally during 1974-75.

b Expressed as percentage of gross turnover during account period .

c Usually without period of accounts ; excluded from average sales and profit/loss calculations.

d No account records ; assumed inactive .

Source : Department of Co-operatives, Suva . 233

and 37.3 per cent in the Northern. M�n thly trading volume also varies by region . It is greatest in the Eastern Divis ion ($1012) wh ere consumer-marketing societies dominate retail trade , and lowest in the Western Divis ion ($758) .

Comparable trading volume and profit /loss accounting is not availab le for Fij ian-owned stores (or for any other ownership group) , but as co-operative stores operate in similar economic and soc ial environments and , aside from input s from the Department of Co-operatives are run much like Fij ian shops, trading data are suggestive of aspects of Fij ian shop operation . It is likely , however, that a larger proportion of Fij ian shops do not make a net profit ,·tbat turnovers are con$iderably less and that the stores do not survive as long.45

When a co-operative society is established in a village , other retail outlets, at least those operated by the society members, usually close . Greater capital reserves and insti­ tut ional support contribute to resilience of co-operative stores but their funct ion remains similar to that of village stores - to operate as a min imum-profit connnunity pantry. A not untypical consume r co-operative society is that at Waisere , a village with thirty residents near Serea in Naitasiri Province . The society has ten members. Between October 1976 and January 1977 the store bad sales of $263, one-third in the month of December. About 10 per cent of sales were to non-members . In the same period , $87 worth of stock was purchased, four-fifths wholesale from the nearby Wainimala Co-operative As sociation and the remainder retail from a Suva supermarket.

Trends in retailing

Retailing in Fij i is not as static as the above review of types of outlet may suggest . Two main trends are un der way that have and are altering the nature of grocery retailing. One is the change in the location and form of outlets. The other is the increasing involvement of Fij ians as storeowners - as either in dividuals or through the co-operative movement .

Processed food retail out lets were in itially established when in ternal connnunications were not well developed an d consequently they were widely dispersed in areas of settle­ ment. Most connnunities were served by broadly-stocked general stores; there was little funct ional specialization , 234

even in towns. With improvement in transport an oppo site trend is apparent. In areas accessible to service centres the small general store is becoming less common , and wh ere it does exist it serves a more limited role. In rural areas people make occasional trips to towns , often in conj unct ion with produce market ing or to obtain basic services , on wh ich some grocery needs are purchased. Near and in towns , general store trade is in creas ingly of 'breakfast trade ' items; other groceries, including staples, are purchased in central shopping areas . Through aggressive retailing practices and by taking advantage of improved connnunicat ions, expatriate and locally owned supermarkets are expanding their trade hinterlands.

The second trend is the in creasing involvement of Fij ians in retailing. Co-operative society stores represent a maj or new type of outlet in terms of organizat ional structure , motivation of establishment an d locat ion . Over 1000 co­ operative societies have been established since 1947, about three- quarters of which operate general retail shops . It is envis aged that another 200 or so primary societies will be created, many of which will be con sumer or con sumer-marketing societies (Fij i, Department of Co-operatives 1975 :3). Co­ operative stores have flourished since the early 1950s when a ban was placed on trade in green coconuts which effectively destroyed the basis of trade of the many Ch inese stores located in Lomaiviti and Lau . In their place came co-operative stores, especially in the Eastern Division wh ere half the retail stores are operated by co-operative societies (Table 3.4) .

As sociated with the rise of co-operative society stores has been the in creasing involvement of Fij ians in the pro­ cessed food trade . Throughout Fij i, 30 per cent of retail food stores are operated by Fij ians. If co-operative stores are treated as Fij ian - consumer society memb ership is largely Fij ian - almost one-half of retail shops are Fij ian-owned and operated.

Many co-operat ive society and Fij ian shops have replaced withdrawing Chin ese and In dian traders , but others have been established in areas not previously served by retail shops. Comprehensive data on long-term trends in the number of licensed retail or wholesale-retail outlets are not available . It does appear from the limited informat ion available , however, that the number of outlets in some areas , such as Navua town 235 an d district , in creased rapidly through the 1960s to attain a level that has since been maintained. Much of the expansion in store numbers in Navua is accounted for by the establishment of Fij ian outlets . In other areas , especially where there is a considerable Indian population , the numb er of outlets has been relat ively constant. In the ten years since 1966, in the Nausori district for in stance, the number of retail stores averaged 277, ranging between 244 (1969) and 298 (1973) .

The in creasing participat ion of Fij ians in retail acti­ vities refocuses attention on a theme of this study : the differential role of the races of Fij i in food distribut ion . Throughout the study of the retail grocery system, as well as elsewhere , attention is drawn to different forms of parti­ cipation by Fijians , Indians , Ch inese an d Europeans. It is useful to close the an alysis of the retail system by explicitly examinin g the place of race in retailing, particularly of Fij ians.

Race and trade

Throughout the ab ove analysis of the processed foods distribut ion system, the key organiz ing concepts of the form of business organizat ion , that is upper and lower circuits, andrace of participants have been ever present considerations. The value of the upper/lowe r circuit .dichotomy in analysing forms of organizat ion is apparent , especially as it has strong racial correlations . Upper circuit activities are almost exclusively the realm of European firms . Indian and Chinese traders are common in the lower circuit, but also occupy an ' intermediate ' circuit . It appears , however , that an inab ility or unwillingness to delegate key managerial positions to non-family members limits the development of Indian , and to a lesser extent Chinese, businesses to this circuit. Fij ian activities are almost exclusively within the lower circuit.

Fij ian participation in the pro cessed food trade is almost all in retailing in rural areas (see Table 3.4) . Located in Fij ian -· residential areas , usually villages , Fij ian stores primarily supply the daily food and basic household requirements of a local , exclusively Fij ian clientele . En try requirements can be minimal (Table 3.26) . The store may operate from a house or a building constructed by family labour for insignificant cost by using local surplus materials. 236

Table 3. 26

Amount and source of in itial capital of Fij ian and Indian retail shops In itial Year shop capital Source of capital Connnen ts established ( $)

Fij ian shops 1957 124 Borrowed from within Fo r stock; sales from home . family .

1971 Son working overseas .

1972 70 Trucking and handicraft For stock; sales from home . b:1siness.

1973 30 Savings . Building already establish­ ed (p rev. shop had ceased trading) .

1974 2,000 Savin gs and sale of For building and stock . cooked food in market .

1975 ? Capital used to add a room on to house and for stock 1975 1,300 Family savings and fund­ $500 for building, $800 for raising activities among stock ; soon after, another relatives . $900 for two freezers .

1976 187 Wage-earning sons. To build room on to house and for stock. Indian shops 1944 Savings from employment Converted house verandah; and hawking . later expansion with store­ generated funds .

1955 6,000 Savings ($1000) and corn- Building and stock. mercial bank loan .

1957 5,514 Own and family savings ; $314 for stock; store connnercial bank loan . building and land purchased.

1967 Savings from employment Room added to house. and family loans.

1969 12 , 400 Family (sold property) . Building.

1971 2,750 Savings . Renovation of family build­ in g; cost of stock not available.

1975 ? Savings from employment . For stock and building ren t ($300/month) .

1975 40o+ Own and family loan . Building materials from wage and hawking income ; $400 borrowed from brother to purchase stock.

Source : Interviews with storeowners in the Cen tral Division . 237

A retail licence, if purchased at all , costs no mo re than $12 for shops in rural areas . Fifty dollars or so buys an adequate basic stock. Re stocking is conducted on the basis of what sells well and the amount of cash at hand.

Fij ian retail sales techniques are passive . There is no advertising, no price wars; if it were not for the require­ ments of the Prices and In comes Board, prices on basic commodities would probably not be marked . Stores are usually established and operated to generate income or to provide a convenient source of groceries for the local community . Little attempt is made to expand trade through imp roving services or stock or by relocat ion to more profitable sites. The store is often the only commercial interest of the owner, apart from a little in come from agriculture, but a store is not regarded as vital to the own er 's economic survival . Stores are often established to 'test ' involvemen t in business . Necessary income such as school fees may be generated by a store but failure to do so is not disastrous ; the family survives , and probably with much less worry and effort than accompanies the operating of a store. Fij ian stores are rarely the source of capital for other commercial enterprise. No cases were en countered where a Fij ian owned more than one shop ) although a few had small trucking bus inesses an d similar en terprises in addition to a store .

It is not that there are no successful Fij ian retailers. Measuring success in terms of net profit , persis tence of enterprise or even trade expansion , there are some . Invariably, however, successful Fij ian retailers are not operating in their home communities but in areas where they have weak social ties. Reflect ing the general patterns of internal migration , many of the successful store operators are from Lau and Lomaiviti and are located in the greater Suva area . Precise information on the role of ' non-local ' Fij ians in retailing is not available , but one suggestion of their dominance comes from our efforts in selecting Fij ian store­ owners for interview . Eight Fij ian retailers operating around Suva were selected for in terview at random from a list of twenty-five ; the list was compiled from business licence records . It turned out that six of the eight had migrated from Lau and one from Rotuma ; the eighth was a descendant of the Solomon Islands plantation labour force that was brought to Fij i at the end of the last century . The Lauan storekeepers and Lauan farmers on resettlement schemes in the Central Division, describe themselves as 'the Indians 238 of the Fij ians', meanin g that they must succeed in commerce as they have little to go back to in Lau and limit ed access to garden lands in their new homes . The Solomon Islander­ descended storeowner described his people's plight and the mo t ive for establishing a shop :

We have ab out sixty-six mo re years to go on our land lease. We do not lmow what will happen to us then . They [the land-owners ] might turn around and say, 'Oh, you had better find another place. ' But we have no money . Unless we have jobs and are work ing in towns and get money to look after our families , the only place for us to get money is to try to run a business . And so we [the brothers in the family] teamed together to do some thing for our future people , especially our children .

Not all Fij ian shops operating away from the owner's home are successful. The eight shopkeepers in terviewed had been operating for an average of 5.4 years : an equal numb er of Indians who were also interviewed averaged 12 .9 years. The Fij ian storeowners did not have a history of entrepre­ neurial activity. Organization of these 'non-local ' Fij ian shops does not differ greatly from other Fij ian stores . Having few social ties with the area in wh ich they operate, however, they do not face consistent or unrefusable pressure from customers for credit . Also, the feeling of dependence on store income , or at least on monetary in come., is a force for perseverance.

Fij ian storeowners face sign ificant mo tivational and attitudinal impediments. It is contrary to Fij ian custom to demand payment for goods or to refuse credit. 'Credit problems ', one storekeeper said, 'are largely the result of the Fij i custom by which we cannot turn away people empty handed' . Th ere is little prestige in owning a store and a storeowner opens himself to jealousy and gossip for 'un-Fij ian' behaviour .

In spite of the difficulties , Fij ians own a large numb er of shops. Almost one-third of the retail grocery shops of Fij i are owned by Fij ians and another fifth are co-operative society stores (see Tab le 3.4). By share of trade, however, the contribution of Fij ians is not great . In the Central Division , it is likely that considerably less than 10 per cent of the grocery trade is handled by Fij ian shops; 239

co-operative society stores would account for a similar share . In other words, almost 50 per cent of retail and wholessle-retail grocery shops in the Division are Fij ian or co-operative society stores but these handle under 20 per cent of the grocery trade . The Fij ian share of the combined grocery wholesale and retail trade would be considerably less than this one-fifth of trade, and Fij ians do no direct importing - apart from the Fij i Co-operative Association , if it is treated as Fij ian .

The disproport ionately small share of trade in Fij ian hands raises some important questions. Why , for in stance, after thirty years of activity and the tm.deniable en trepre­ neurial in terest of in dividual Fij ians, do co-operative societies have such a small share of the grocery trade and essentially no share in non-agricultural commerce? The explanat ion may in part be motivational and att itudinal , but equally basic are structural considerations.

A most significant trait of Fij ian connnercial activity is its isolation. While existing in the same commercial system as the retail and wholesaler interests of Indian, Chinese and expatriate firms , the Fij ian is in may ways very much apart . Fij ian retail shops are usually located in isolated rural areas . Not only have Fij ian shopowners not had connnercial training, unless perhaps they have worked for a co-operative society, but also they are very unlikely to learn even rudimentary business principles by operating a village store. The limited trade activity in villages and general ab sence of commercial competition of fer no incentive for traders to innovate, let alone to become mo re aggressive . As trade is entirely with Fij ians - mo st of whom are relatives of the storeowner - who have largely predictable and unchanging demands , commercial in itiative is further inhibited. Isolation is felt in other ways . In terms of supply, Fij ian stores are set apart from others . They are rarely visited by salesmen or wholesalers , even when they are located on roads . Mo st storeowners travel to supply centre to purchase foods , usually for cash since wholesalers are often unwilling to grant credit and many Fij ian storeowners are unwilling to receive it . In addit ion , storekeepers are all but ignored by government, training or funding agencies.

As well as aiding Fij ian economic progress, the co­ operative movement has contributed to the creation of the present situat ion. Co-operat ive societies were promoted as 240

the avenue of Fij ian entry in to connnerce . Through their isolated rural locations , connnunity service role an d the possibility they afford for the denial of individual respons­ ib ility, many have become community pantries for necessary processed foodstuffs . Fij ian demands for mo re effective in tegration in to the economy can be deflected with the an swer that co-operative societies exist to this end. Wh ile it is true that the Department of Co-operatives and societies are regarded as Fij ian , they do not appear to have helped Fij ian in dividuals to advance economically , an important aim for many as their involvement with stores suggests. The view that Fij ians should 'advance ' through co-operative societies effectively curtails any assistance that mi ght be given to individual Fij ian en trepreneurs.

Chinese and Indian businesses involved in the grocery trade stand apart from Fij ian commercial en terprise but this is not the only common feature they share . Both are involved at importing, wholesaling and retailing levels. Indian and Chinese grocery retail establishments comprise a range of forms and scales of operation . Condit ions of en try are no more rigorous than for Fij ian and many commence trading on a similar scale (Table 3.26) , but there are differences. While Chinese an d Indian retailers are ignored by the government as much as are individual Fij ian traders , they have a stronger commercial tradition , and although many Indian and Chinese shops do net expand beyond the scale of mo st village stores , vertical and horizontal expansion is more characteristic .

The business organization of the smaller Indian and Fij ian shops does not differ greatly . Labour and other inputs are family-based and credit is an important feature of sales policy; investment and working capital are greater than for Fij ian shops, but are still often limited. Greatest differences occur in the mot ivation of store operat ion . For mo st Indians , the store is a full-time business - if not for the owner, who may farm or hold another job, for others in the family who will keep it open throughout the day. Sales practices may be as passive as those of Fij ians , but most are mo re than service outlets for the connnunit ies in wh ich they are located. Many are located on roads and have some through-traffic trade . Few Fij ian stores sell much more than basic requirements , but In dian grocery stores are often general purpose outlets, carrying a basic range of homeware and drapery. Many Indian shops in towns have non-grocery funct ions. Whereas retailing is practically the only 241

commercial activity of Fij ian storeowners , it is often just one of a range of activities of Indian families . Re tailing usually remains the basis of business , with taxi services, car repairs , wholesaling, market and other selling, commercial agriculture and wage labour of ten being carried on by memb ers of the one family . Expansion out of food retailing in to other activities can stretch family resources but rarely to the extent that organization principles change. Finance for expansion is generated by the other activities , although successful Indian entrepreneurs do gain access to bank loans . Labour remains intensive and family�based , and income continues to be saved and invested rather than consumed .

Chinese grocery retailing is in some ways similar to that of Indians. Operating largely out of rented premises in urban and peri-urban areas , Chinese retailing is based on family labour and internally-generated capital. Perhaps reflecting more central store locations and more active clientele , Chinese stores often carry a wider range of stock than Indian shops, and certainly more than Fij ian shops, although groceries remain the maj or item. Chinese retailers usually expand business by moving to a more central location , increasing stock range or extending vertically in to whole­ saling. Th ere is less development into ancillary operat ions than occurs among Indian traders, a main reason being the dependence of Chinese on family labour. Non-Chinese are never employed save in menial tasks where help is warranted by volume of business .

Chinese involvement in the retail and wholesale-retail trade of the Central Division is but a shadow of previous levels, even though they own 13 per cent of the stores and probably still handle about one-fifth of the business. Chinese shops, once predominantly rural in location , are now heavily urban : 135 of 178 Chinese-owned retail and wholesale-retail stores in Fij i are in urban areas . Many Chinese sto reowners moved from rural locations as the ban on green coconut trading and the rise of co-operative society stores cut into their trade . The Chinese population has decreased greatly in recent decades through emigration so that many families today have in sufficient memb ers to support current activities, let alone expansion . More than in any other racial group , Chinese male children are encouraged to work in the family business. 242

Commercial activity in Fij i is strongly in fluenced by factors of location . Fij ian endeavour is severely handi­ capped by spatial isolation, .and the strength of Indian , Chinese and expatriate European activity is partly due to locational advantage . Given the spatial characteristics of the national economy - strong rural/urban dichotomies and the concentration of wealth in urban areas and the sugar­ producing region - the future of racial involvemen t in commerce takes on a certain air of inevitability. The number of Fij ian shops may increase in response to populat ion growth , but whilst Fij ian shops remain in rural areas serving Fij ian communities , the Fij ian share of the retail trade will decline and there is little possibility of entry into wholesaling or importing . Businesses in urban centres will continue to control a large share of the grocery trade at all levels, as well as of other commerce, in_ which al ready there is negligib le Fij ian involvement. Without sign ificant change in the spatial features of the economy or the location of Fij ian en terprise, Fij ians are unlikely to achieve significant involvement in commerce . There are maj or differences in economic fortunes an d commercial activity between sub-racial groupings that are as much a product of spatial factors as is the plight of Fij ian connnerce� these differences, however , are beyond the scope of this report. 45

Conclusion

Three objectives of the study as they relate to the processed foods distribut ion system - delineation and analysis of the system, the role of indigenous and non-indigenous people in it , and government policies relat ive to the system - have been covered above an d require no restatement. The fourth objective of reviewing the implicat ions of government policy on indigenous involvement, however, is an appropriate conclusion to this chapter .

The govenunent of Fij i, apparently , has two chief concerns relative to the dist ribution of processed foods :

(a) to ensure that prices paid by consumers are reasonable and inflat ion in these is minimal ; (b) to broaden the participation in the distribut ion system by all 'local ' people , but particularly by Fij ians . 243

Th ese concerns are pursued throughout the commercial sector, including the grocery trade , and a review of government policy in this latter field is a case study of broader policy .

The concern of limiting costs to consumers and dampenin g inflat ion is handled through the activities of the Prices and Incomes Board (PIB) . The PIB has been effect ive in establishing prices for bas ic commodities, but it has not achieved a second obj ective of countering perceived mon opo­ listic trends in commerce. Indeed, price controls in gro cery importing and wholesaling appear to have encouraged monopo­ listic trends in the supply of basic connnodities . If the trend towards the withdrawal by me rchants from wholesaling price-controlled commodit ies continues, it might be necessary for the government itself to become involved in wholesaling. In the light of this development , there is some urgency that the consequences of allowab le price markups on basic connnodit ies be examined.

One difficulty with price controls on imported basic commodities is that a critical stage of price determinat ion take s place outside Fij i. The extent to wh ich external pricing practices affect the costs of goods in Fij i is not precisely known . As it is likely that knowledge of the external system obtained by an enquiry or even through an ongoing government office will be incomplete and dated , one possible way to minimize negative consequences of such practices is to establish a para-governmental buying house . This would serve the dual purpose of familiarizing the government on an ongoing basis with international trade pract ices and would also, if operated efficiently , provide meaningful competit ion for the larger importers . The Fij i Co-operative As sociat ion could be a foundation for such an organizat ion , but to be effective it would need to operate on a mo re compet itive connnercial basis than at present and also to expand its trade , perhaps by servicing non-co­ operatives, in order to obtain quantity discounts similar to those received by its potential competitors .

In relat ion to the second concern of in creasing indi­ genous involvement in connnerce , specifically in the processed foods trade , three points can be made . First , the main avenues for encouraging Fij ian participation in connnerce - the Department of Co-operatives, the FDB and the Fij ian Business Opportunity and Management Advisory Service - have not developed extensive Fij ian participat ion . The role 244 given by the government to the co-operative movement , for in stance, as the mean s of Fij ian participation in connnerce has in fact contributed to the isolation of Fij ian connnercial endeavour and has compounded the locational disadvantages that Fij ians face in successfully entering commerce. Alter­ native steps to promote Fij ian participation in commerce could include :

(a) in creasing Fij ian participation in critical areas of commerce by encouraging the location of Fij ian commerce in urban areas and the involvement of Fij ians in grocery importing and wholesaling as well as in non-grocery trade . This could be done by reserving particular connnercial functions for Fij ians and installing Fij ians in strategically­ located shops with finance and , as needed , in itial managerial assistance from the Fij ian Development Bank or a similar institution ; and

(b) recognizing that co-operative consumer societies serve a particular ftmction - that is , the low-cost supply of established consumer demands of Fij ian villagers - and cannot be expected to meet the economic aspirations of all Fij ians . Other equally valid forms of economic act ivity (such as retail store ownership) ought to be supported with easy access training programs and bus iness support services similar to those received by co-operative societies.

Second , wh ile increased involvemen t of Fij ians in connnerce is a prime national priority , significant differences in economic status within all racial groups should not be overlooked ; nor the fact that , as a group , grocery retailers earn much less per person than people in mo st other commercial occupations . Improved economic conditions for all people involved in the grocery trade could be achieved by placing emphasis on comme rcial and business courses at secondary school level and by establishing connnercial courses for others , which could include taking basic business instruction to the communities where storekeepers are active .

Finally , expatriate European firms dominate general importing and wholesaling and are very significant in retailing . As long as trade practices are fair, business efficient and profits not excessive , there is nothing inherently wrong with overseas involvement in connnerce . 245

However , the people of Fij i could be safeguarded by the government buying in to the main companies . In this way , the government would gain a degree of oversight of company operations and reasonable financial returns on the investment and the companies would not suffer unduly by having a significant government shareholding. Appendix

A Fij ian permanent vendor at Suva market : an edited version of an in terview with Vika Bainimar ama by Tevita Ba, 20 February 1977

[Vika Bainimarama from Mokani, Tailevu, is one of the more estab­ lished Fij ian perman en t vendors at Suva market . After leaving school at Class 8 level an d marrying a man from Rewa , Mrs Bainimarama worked for seven years as a �ous e-girl in Suva , for a short time as a wait ress and then at home sewing items to sell to neighbours . She started selling in the market in 1968. He r husband retired from his job as a carpenter in 1972 and has devoted his time since then to operating a carrier from Suva market and a mo tor-powered punt in the Rewa delta area , both of which were financed largely from Mrs Bainimarama 's market earnings . Both her husband an d son, who is just complet ing secondary school , as we ll as a number of relatives , help Mrs Bainimarama in the market from time to time. ]

I had been wat ching the foreigners [non-Fij ian ] vendors in the market and how they stayed there without too much moving around . Then one day I decided that I would give it a try ... �he first thing I was to sell was food. I cooked some food then wen t and sold it to women I knew at the market as a trial run . The food did not last an hour before it was all finished. They told me , 'Cook again for tomorrow' . And so I cooked again and it was then that I realized that there is a good life at the market with no one to boss us around except ourselves . After one whole year of selling cooked food I started selling crops because I was alone with no one to help me cook or sell food. That was in the year 1968 and I have cont inued to sell up to now. At that time I was in side the market , over at the side near the fish market where there was no roof. We tied sacks to stop the heat of the sun . Then in 1973 the place had a roof and then I moved outside . I moved out because I was selling watermelons. They were always piled out side so I came out and bought them from the growers from Nadroga . That ' s why I move d out and have a table outside the market.

When I started selling there were a few Fij ians selling whom I knew pretty well. They told me many things for advice : like the work depends on personality - talk well with customers and then they 'll care for you and your table; if we are always looking angry no one will dare come and buy. Many th ings like that they advised me .

I had few customers at first , which was rather discouraging. But when I started to meet people and to know them, mo re and more customers

flocked to buy from my stall. It took quite a bit of time to know and to attract customers in to my stall. But things turned out right and became promising as the years went by.

I also found it hard to adapt my self to the mood of se�ling. Firstly , I knew very little about selling. Secondly, selling was a new

246 247

act ivity with regard to my culture . I had this attitude of shame towards the job. I was also ashamed to meet people, especially Fij ians . I used to think that selling was a lowly-regarded kind of job. This also used to be the general attitude of Fij ians toward selling, espe cially selling at the market . But this attitude is improving now as more Fij ians are involved in selling at the market today .

My most difficult prob lem was that about arrangin g with farmers for the items I wanted to buy from them. I was not well acquainted with farmers at first. Hence I found it diff icult to contact the right farmers who would supply me not only with the commodities that I was interested in but also those items that were best to sell. When I came to know farmers from Nadroga, Tailevu and Naitasiri, my prob lems were eased.

Also the long hours standing in the market and especially on the concrete floor have adverse effects on my health. Somet ime ago I had swollen legs that caused me to limp badly .

I learnt from experience selling techniques and the best items to choose to sell. The price to impose on each item sold depends on the cost prices . If I bought certain items at cheap prices , I would set my selling prices for these items at lower rates too . The items I buy at higher prices from farmers are expected to have higher selling prices . My pricing would depend o� my cost prices. I do not set a target to achieve in one day . I would rather make sure that I get back the amount of money I spend each day to buy my connnodities from farmers and to get reasonable profits from the sale of these connnodit ies as well.

Some farmers impose unreasonably high cost prices. Th ese are usually inexperienced farmers. These types of farmers usually cause prob lems to vendors . Th ese same farmers would come with poor quality commodities. They do not even take great care in selecting the best items and in making decent bundles and arrangements of the items in baskets so that they are att ractive to look at . However, the experien ced farmers are quite helpful in that they usually make their commodities in good and attractive order , and their prices too are usually reasonable.

Sometimes I experience losses in my business. This has been caused by poor quality items which were packed unnoticed in bags I obtained from farmers . Sometimes I experience losses on items like rourou, bele and cassava which do not last long. If these items are not sold after one day , it is obvious that I would experience losses with them because they would stand very little chance to be sold on the next day . Items that can last long are dalo, wild yam, sweet potato, and ivi and fruits.

Sometimes people , especially the farmers , are dishonest in their dealings with me . This is mainly regarding the sale of their poor items to us. For in stance, when they sell cassava in bags , they put in at the bottom poor, small and even bad cassava. This we would not see but the big, good ones at the top . Only at the time we start selling, then we realize that much of the cassava cannot be sold as it is bad or of poor quality. Thus , this would be a straightout loss to my business.

There are many advantages in this job. We get more money and we receive 'cents ' every day. In other jobs it depends on payday - late during the week or fortnightly or 100nthly, not like in the market where 248

you may be get ting it every day. Also, in other paid jobs we are bossed around as we have bosses . In the market , al though there are the market masters , all they want is the price of the table - then all the rest is up to you - whe ther you go for a stroll , sell or go to sleep, all that is up to you. But with other paid jobs, if I come lat e I would expect some harsh words . Here , you come and do what you want . Buy your produce , leave it at your table. Your life and death is entirely in your hands. No one in the market will talk ab out what you do - it is your own business.

I am really in terested in the bus ine ss and I am trying hard to expand it as much as possible. Meanwhile my husband and I are working together to expand the bus iness. Apart from selling, we are running a truck and an outboard boat , both of which are getting encouraging returns by way of money into our business . Even now we are planning to start a shop.

In the meantime my son is mo st interested in my business at the market . On occasions he comes to help with selling. He also helps his father in our plantation at Sawani and the truck and outboard business. My son realizes that the business at the market will provide him with all his financial requirements. But I will let him de cide for himself with regard to his paid job. I myself would like him to take a well-paid job, but otherwise he can join me in running our business here at the market .

My husband is very in terested in my business. Through this selling, he has also been ab le to purchase a ' carrier ' , an outboard and has built our house . Profit from my business at the market has helped tremendously to obtain these items . In fact my business has branched out as part of it is that ' carrier ' and tboaou rd, both of which are doing business as well.

The number of Fij ian vendors in the market is increasing steadily. The number of vendors joining the selling business in the market , Indian , Ch inese an d others all have increased, but the greatest in crease is that of the Fij ian vendors . Many Fij ians who come to buy in the market have been realizing that the costs at which these commodit ies are sold in the market have been giving good returns to the sellers. Many more who visit the market are being moved by the work we are do ing and through discussions with us they are now realizing that selling in the market is a worthwhile job as it is most rewarding in that it quickly and quite easily brings to us that money wh ich we want and need. All people in town , too, more or less, must buy food from the market. In this sense the market vendors are doing great service to the people. Hence selling in the market is both a worthwhile and a respectable job. Many Fij ians are now aware of these uses of selling in the market . These aspects, of course , are ch anging the Fij ian ' s negat ive attitude to selling. Many are even encouraged to join the group already in the market .

There were fewer Fij ians in the market in the late 1960s and early 1970s than today . Con sequently, the few Fij ians who were selling at the market would regard their business activities then to be at their peak . Th is was because we were receiving many customers , many more than we are getting today . The more cust omers we had, the more items we had to provide for them. This usually resulted in big sales each day and we got encouraging profit s from them. Today , there are more Fij ians selling in the market . This has affected the string of my regular customers . Most of them have to go and buy from Fij ian vendors they are related to 249

or whom they know personally . So today , mo st of our customers are our own relat ives or people who know us personally.

The type an d quantity and quality have also improved and increased since I started selling. This may have been the result of the increase in the number of people doing farming activit ies . Of course the prices of the items when bought from farmers have also in creased. And so the retail prices we impose have in creased. Notes

Chapter 1

1 Food imports as a percentage of total imports by value for other Pacific countries are : Papua New Guinea 20.1 (1971- 72) , French Polynesia 22 .2 (1972), Fij i 22.3 (1973) , American Samoa 23.5 (1971-72) , New Caledonia 23.8 (1973) , New Hebrides 24.4 (1972) , Solomon Islands 27.4 (1973) , Cook Islands 28.0 (1970) , Western Samoa 32 .5 (1972) , Tonga 35 .2 (1972) , an d Gilbert and Ellice Islands 36.8 (1973) . Comparative 1970 figures for Au stralia and New Zealand are 3.1 and 5.1 (McGee 1975 :14) . 2 Reviews of the function of marketing in economic development in clude Moyer (1965) and Moyer and Hollander (1968) . 3 Following local nomenclature , 'Fijian' , 'Indian' , 'European ' and 'Chinese' are used throughout the report in an ethnic or racial sense to refer to people of such origins living in Fij i. Save where qualified as 'expatriate ', these people may be taken to be cit izens of Fij i. 4 'crop area' should not be confused with 'cultivated area' . In crop area , areas under mixed crops are multi-counted and those with seasonal crops are adj usted to reflect total annual product ion . See Casley 1969 (which although dated and in parts difficult to comprehend, is the only available broad survey of agriculture in Fiji). 5 sugar and coconuts have for long dominated export-oriented connnercial agriculture ) but at present the product ion of both is in flux. Product ion of sugar declined 13 .8 per cent in the period 1965-75. The quantity of sugar exported fell 25.2 per cent from 1970 to 1975 although the value increased 197 .7 per cent. The increase in value was largely due to high returns in 1974 and 1975: value of exports from 1970 to 1973 in creased only 7.7 per cent while export quantity fell 18.5 per cent (Fij i, Bureau of Statistics, 1976) . The future of the sugar industry is uncertain . New areas being opened for sugar production are to be operated in part by corporate estates (in contrast to present reliance on

250 251

smallholder con tract farmers) . Competition from other sweeteners could have radical con sequences on the in dustry. The production of coconut products is in a slump . Copra production declined by almost one-third between 1965 and 1975 and the amoun t of coconut oil produced in 1975 was the same as in 1965. The export of coconut products (oil , oil seed, cake and meal) fell 23.1 per cent from 1970 to 1975 while the value of exports declined 2.1 per cent (Fij i, Bureau of Stat istics , 1976) . In some coconut growing areas diversification is under way based on root crop an d live­ stock production for the local market. 6 Unless otherwise acknowledged, statistical in formation in this section is from Fij i, Central Planning Office , 1975; Fij i, Bureau of Statistics, 1976; and Fiji, Prices and Incomes Board, 1975, n.d. 7 The 1959 and 1965 surveys are only roughly compatible with the two other surveys . For survey methodology and results see Ward 1970:67-8 and Fiji, Bureau of Statistics, 1968, 1974. 8 For population figures for urban areas see Table 2.3. 9 1976 population data are preliminary returns of the 1976 population census as published in The Fiji Times , 7 October 1976 and 21 May 1977 . 10 For each group of foodstuffs , range of purchase frequency between the four income quartiles was calculated as a percentage of the average expenditure. 11 studies were conducted over one week in July 1976 in commun ities selected primarily on the basis of racial com­ position and location in relation to Suva. The surveys were designed to investigate the relative frequency with which different foods were consumed . No attempt was made to measure the amoun t of food consumed or the amount spent on different foods (or on all foods) . Only main foods consumed were recorded; that is , spices and seasoning vege­ tables like onions and chillies were discounted. Survey results are not strictly comparable with consumption pat terns revealed in the urban household income and expenditure surveys , but it is possible to compare results with urban findings and to extrapolate to other rural areas , so long as it is accepted that different in come levels , availability of home­ grown food and access to supply centres , are just some of the many factors affecting consumption patterns. Taking these factors in to account, more revealing in ferences can 252

be made than from the available data on apparent food consump tion and aggregate national nutrit ional in takes (Fij i, Central Planning Office, 1975:63) . In each connnun ity the enumerators , all students at the University of the South Pacific, randomly selected ten households . It was planned to record during seven conse­ cutive days the frequencies with which foods were consumed in each household . As it eventuated, one connnun ity was surveyed for six days and fewer than ten households were monitored in three cases . Methodology and results of the survey are recorded in Michael Baxter, 'Food preference patterns in the Central Division , Fiji', CASD, USP , October 1976. Works relative to the survey methodology are Belshaw (1957) and Finney (1965) .

Chapter 2

1 The Departmen t of Agriculture calculated in 1963 that the proportion of food grown for home (sub sistence) consumption was 95 per cent of root crops, 81 per cent of vegetables and 77 per cent of fruit (Biggs n.d. :4). More recent data are not available . 2 official rural markets are operating at Laqere, a Suva suburb , and Dreketi in Macuata, Vanua Levu. It is planned to establish others at Nabouwalu (Bua) , Seaqaqa (Macuata) , and Gau and Kero in Lomaiviti by 1980 (Fij i, Central Planning Office , 1975:227) . 3 Even the arrangements for export markets were for long left largely in the hands of producers. An attemp t by growers to establish overseas outlets results in the establishment of the Fij i Planters ' and Fruit Growers' Co-operative Agency Company, Limited in 1890. The company aimed to sell its members' agricultural produce in Australia and elsewhere, forward produce of members, make loans to memb ers for the storage and shipment of produce and purchase of agricultural machinery on behalf of memb ers. There is no record of what the organizat ion achieved or how long it survived . See Memorandum and Articles of Association of the Fiji Planters ' a:nd Fruit Growers ' Co-operative Agency Company, Limited (G.L. Griffiths, Suva , 1890) . 4 The Nausori market was quickly developed to meet military produce requirements by expanding the production and market activity of a large number of small growers through the 253

supply of seed , fertilizer and in secticides, fixed-price purchase agreements, and co-ordinat ion of produce collection and transport. Marketing arrangements were handled largely by the Fij ian Co-operative Marketing Association (founded in 1942 with a paid-up capital of £5000 . 5 rnformat ion on early markets is scarce but some may be cleaned from the annual reports of government bodies . Informat ion on the establishment of Lautoka market is from the 1952 Annual Reports of District Connniss ioners , Legis­ lative Council Pap er , 19 of 1953, p.34. 6 Annual Report (1949) , Connnissioner of the Central Division , Legislative Co uncil Paper , 1 of 1953, pp.1-5 . 7 Reasons for failure of co-operative marketing societies included poor administration and members ceasing to sell through their society, either because better prices were available elsewhere or because they baulked at paying a connnission on sales . Marketing co-operatives were formed largely in reaction to middleman trade : the Nadroga-Navosa Co-operative As sociation Limited was created in 1948 'to free the Fij ian farmers from the middleman ' (Departmen t of Co-operatives , Annual Report , 1949 , p.15) . During the early 1960s, some marketing co-operatives in Lomaiviti appointed full-time agents in Suva to sell their produce , particularly yaqona. Few schemes survived as yaqona could invariably be sold for higher prices , and for cash, to middlemen on the islands and becaus e society overheads were high. Today there are a numb er of marketing co-operatives that deal in produce, but while they are sign ifican t outlets for some producers, they have a minimal role in produce supply in general. 8 The involvement of the Fij ian Administrat ion in economic activities and particularly the activit ies of the Economic Development Officers is examined by O.H.K. Spate in his 'The Fij ian People : economic problems and prospects ' , Legislative Co uncil Paper, 13 of 1959 , especially pp . 40-7 . 9 see the 1951 and 1954 Annual Reports of District Connniss ion­ ers, Legislative Council Paper, 1 of 1953, 19 of 1954. 10 rn 1967 it was decided that weather and soil conditions , nematodes and bunchy top disease were ob stacles too severe for successful banana production at Lomaivuna so a mixed economy was adopted . Today , the minimal Lomaivuna banana production is for internal markets. 11 see Couper (1967 :243-5) for details of the Beqa-Suva and 254

Yasawa-Lautoka trade in 1964. Today the trade of Beqa producers is largely to Navua market rather than to Suva . 12 The experts were Manuela G. Maramba, a nutrition officer who wrote 'A plan of operation for a market study to be done in the Colony of Fiji' (MAFF , file 2/140) based on a 1960 visit , and the agricultural economist, D.R.N. Brown , who se report was 'Study of the economics and marketing of certain agricultural conmodities in the Fij i islands - Progress report , July 1964 ' (MAFF , file 620/2/39). The market officer was transferred to other duties one year after appoin tment and not replaced . 13 section 2 of the Suva (Markets) By-laws , 1937 , stated 'The markets shall be for the use of the general public for the purpose of selling vegetables, fruit , fish , tobacco , live poultry and such other connnodities as may from time to time be permitted '. Un like current regulations , the By-laws encouraged non-producer trading by giving a fee reduction (4d per day as opposed to 6d) to vendors who paid one week's fees in advance. The By-laws left room for licensed hawkers to sell produce in towns . Later legislation has defined goods that may be handled by hawkers as excluding produce sold in markets, e.g. Suva (Hawkers) By-laws , 1966. 14 under the Markets Ordinance, produce sales could be pro­ hibited within three miles of markets, but two miles was the distance adopted by the Market Regulations ; towns sub­ sequently adopted restrictions 'within town boundaries '. The Regulations prohibited sales of fish within one mile . 15 A separate 'market ' for handicraft dealers has been built (July 1977) in Suva; it is in tended that the handicraft vendors who are presently in the market will operate from the new building. Further conflict between food and non­ food vendors in Suva market has been avoided, and probably even that between producer and non-producer sellers has been partially defused, as more selling space should be available for producers. 16 0ne of the few detailed accounts of marketing prior to the mid-1960s refers to Lautoka marke t. There , in 1959 , approxi­ mately 44 stalls were occupied during the week, 29 of which we re held by six perman ent stall holders ; over 130 stalls we re occupied on Saturdays when there were 'Fij ian people

who come from Sigatoka and elsewhere [who ] •.• bring stocks of bananas and other things . They spread them outside [the market building] , and the market charges them by the

square foot [of space] •.• ' (Colony of Fij i, Legislative 255

Council Debates (1959) , Session of 1959-1960 , pp . 695-707) . Couper (1967) reported on aspects of in teTI1al marketing in 1964. 17 As government marketing officer, Josevata N. Kamekamica organized a survey of more than 4700 market vendors at all markets in Fij i in 1962-63. The report of survey findings ('Report on a survey of local markets in Fij i, 1962-1963 ', unpublished) cannot be located . A brief account of local produce marketing based on the survey material (Kamekamica 1966) is the only remaining record of the research . 18 Biggs noted that middlemen (non-producer vendors) were mo st numerous at Suva market wh ere there were 66 (Biggs n.d.: 57) . In 1976 there were more than double this number . 19 charges at Suva market wh ich were amon gst the highest in Fij i suggest the in tricacy of fee schedules . From 1 January 1977 fees were: produce stalls , $1.05 or 52c per day or $31.50 or $15.50 per mon th (depending on size and location) ; curio , kava and tobacco stalls, $1. 30 or 68c per day , $39 or $20.40 per mon th ; mo st shell fish , seaweed, sea urchins , etc. , 6c per basket or bundle ; mussels , clams , mud crabs, etc. � $1. 05 per sack; fresh water prawns, 6c per bundle ; sea prawns, lOc per pound, crabs, turtle , smoked fish , 6c per pound. 20 Customer patronage at the fifteen official urban markets was examined in a series of surveys in January 1976. An account of survey findings is in Raj esh Chandra, 'Market buyers and their purchase behaviour in Fiji: a preliminary analysis ', CASD, USP, Decemb er 1976. 21 In addition to general observat ion , in formation on urban produce markets an d vendors is from three surveys conducted by the Distribution Systems Proj ect . The main source of in formation is a week-long survey conducted in the fifteen main markets over the week co11ID1encing 26 January 1976. In the January surveys , data were collected by teams of in terviewers, mo st of whom were students at the University of the South Pacific. The teams varied in size relative to the day of the week and the numb er of vendors in each market ; in most markets the numb er of interviewers was at least doubled for Friday an d Saturday . Two different questionnaires were used. One sought basic socio-economic informat ion about each vendor. The first day during the survey week a vendor came to market , a 'personal data record ' was completed. Assistants to vendors were not in terviewed . The second questionnaire dealt with the 256

produce each vendor had for sale and was completed each day a vendor came to market. It was in tended to cover all vendors and their produce for each day of the survey week. In the smaller markets this was done, allowing for a small number of vendors who completed selling and left the market before being interviewed : 90 per cent or better coverage was achieved in Raiwaqa, Korovou, Tavua, Navua, Levuka , Savusavu and Vaileka markets; coverage in Nausori , Lautoka and Labasa was around 80 per cent; in other markets it was less. Coverage was lowest in Suva , the busiest market , where less than one-half of vendors were interviewed. In all markets, Monday-to-Thursday coverage was close to com­ plete but great increases in vendor numbers on Friday and Saturday made it impossible to maintain high coverage on these days in some markets. Con sequently , the significance of permanent and non-producing vendors relative to non­ permanen t producing vendors, mo st common on Friday and Saturdays , is over-estimated for markets where there was not complete coverage. A provisional report of the findings of the January surveys is Baxter, 'Market vendors an d their produce in Fiji: a tentative analysis ', CASD , USP , November 1976. Other surveys of market vendors were conducted in Jtllle and Novemb er 1976. The purpose of these surveys was to check the general validity of the January results and to examine seasonal variation in market behaviour . The surveys centred on the vendor 's race and other personal characteristics, place of residence , place of production and type of goods sold. The Jtllle surveys were conducted in Suva , Nausori and Navua markets on a Wednesday , Friday and Saturday of one week : 718 vendors were in terviewed in Suva , 426 in Nausori and 152 in Navua . In November, all vendors present in Ba and Labasa markets were in terviewed daily for one week. In Ba, 323 vendors were int erviewed and in Labasa 281 : both these figures represent about 90 per cent coverage . Where no data source is given in this section , it may be assumed the in formation is derived from the January 1976 survey . 22 The study of Fij ian producer-vendors in Suva market was conducted during Septemb er and Octob er 1976 and continued in 1977 . Findings are partially reported in Jenny Baines , 'Fij ian vendors at Suva market ', CASD, USP , December 1976. Data on non-producer vendors in this section are from Baines ' paper . 23 nata on price markups and non-producer-vendor income are from studies tllldertaken in Suva market, 7-11 Sep temb er 257

1976. Every purchase an d sale of produce by 30 vendors was recorded for at least three consecutive days and for mo st vendors for five days (Tuesday to Saturday) . 24 nuring the January 1976 market surveys , data were collected on the produce handled by all interviewed market vendors . In considering the data, allowance should be made for survey methodology , particularly the extent of coverage (see Note 21 above) and seasonal variation in produce availability. On the first day a vendor was interviewed, a 'daily produce record ' was made of the produce brought to, or bought at , the market that day and a similar record was completed each subsequent day of the week the vendor was at market . The quantity an d selling price of each type of produce , production location and the source of produce - the vendor or his family , an other grower, a non-grower or an importer - were noted. Informat ion is mo st complete for produce handled by producer vendors as it was relatively easy to record the total amount of produce they brought to market. Permanent vendors make mo st purchases early in the day but may buy during the day : it proved difficult to record these later purchases. 25 Main produce categories (and some typical items in each) referred to in this report are : Native Staples - taro (all varieties), cassava, sweet potato, green banana, cooking banana, plantain , breadfruit , dry coconut ; Native Vegetables - ota, rourou , bele, duruka , pumpkin leaves; Indian Staples - Irish potato , onion , garlic, carrot , rice , pulses , dry spices (e. g. cumin) , pumpkin , marrow, jackfruit ; Indian Vegetables - okra , eggplant , beans (all save 'French'), gourd, chillies , ginger, sago heart , bitter cucumber, angled loofah� In troduced Vegetables - lettuce , tomato, French bean, Chinese an d English cabbage , bell pepper; Fruit - pawpaw , banana, local citrus fruit , guava , Polynesian apple , soursop , ivi, man go , watermelon , pineapple, green coconut ; Imported Fruit - apple , pear , grape , plum , sultana, peach ; Tobacco , yaqona ; Oils - coconut, soybean , 'salad ' and peanut oils ; ghee ; Animal Products and Sea Food - poultry , fowl eggs , crus­ taceans, molluscs , dried and smoked fish , beche-de-mer, seaweed, turtle. Capit alized produce groups in the text ref er to the above categories . 258

26 nata were compiled from the residence-to-market distance of producer- sold products recorded in the January 1976 surveys. Vagueness about the production locat ion of produce sold by non-producers made it necessary to discount this produce from calculation s. Results of surveys conducted at Suva , Nausori an d Navua markets in June 1976 generally support production location data recorded in January. Pro­ duction distances for producer-sold products at Suva market in June and January are: dalo 36.9 km (23 .2 in January) , cassava 32 .8 km (16.3), dry coconut 39.6 km (34.8 ), banana 20.8 km (20) , rourou 34 km (28.5) , okra 17.7 km (46.4) , eggplant 15.6 km (14.5) , chillies 38.2 km (27), Chinese cabbage 14. 3 km (9.3) , cucumber 20. 1 km (20. 7) , pawpaw 28. 7 km (19.2) and lemon 45 km (11.6 ). 27 The development of vertical links in the trade of marine and fresh water products is a good example of the conse­ quences of resource control on product supply and pricing. The riverine, estuarine and coastal environments of molluscs, crustaneans , edible seaweeds , etc. are more or less exclus ively controlled by Fij ians , who exploit the resources much as subsistence-commercial farmers do their.products. The products are not wholesaled to non-producer vendors; amounts brought to market are limited and prices relatively inflexible. In contrast, sea fish entering the markets are caught mainly by Indians and much is sold through non­ fishing in termediaries; although by law fish can only be sold by licensed fishermen or retailers. Prices of fish are more responsive to supply and demand trends than are other seafood prices. Aspects of seafood production and supply in Suva , Ba and Labasa markets were examined in November and December 1976. 28 carriers are privately-owned trucks hired by an individual or a group for a specific journey or which collect fare­ paying passengers along a particular route . 29 A questionnaire survey was conducted over one week in Suva, Nausori and Navua markets on 21-26 June 1976 and in Labasa and Ba markets on 22-27 November 1976. It was in tended to cover all boats and vehicles other than buses used to bring produce to market over the survey period; this was generally achieved with the exception of early (before 7am) arrivals at all markets but Suva. Upon arrival at market, vehicle and boat operators were questioned about the vehicle 's ownership and journey origin and nature . Information was also compiled on the vehicle 's size, etc. and form and source of cargo . In all, 330 different vehicles making 259

532 journeys (Suva 352 , Nausori 68 , Labasa 24 and Ba 82 ) and 29 differen t boats (Navua 24, Labasa 5) making one journey each were covered. A preliminary account of the June survey is Baxter, 'Specialised transport of produce to Suva , Nausori and Navua markets : a tentative analysis of a survey conducted in June 1976 ', CASD, USP , Novemb er 1976. 30 carriers are important for producer-vendor tran sport when significant supply areas are poorly served by bus (e.g. in the Western Divis ion and at Navua an d Labasa) . The tran sport survey revealed that while the frequency with which vehicles are used by their owner remains fairly constant between markets, there is considerable variation in the rate of use for hiring and for collecting in dividual fare-paying passengers and their produce. Surveyed markets with good bus links to their hinterlands (like Suva and Ba) have a low incidence of carriers collecting in dividual passengers . Poor bus connections between producing areas and Nausori, Navua and Labasa encourage the use of carriers to collect fare-paying individuals rather than their hire for a specific journey. 31 rn addition to the number of vendors or vendor-day's , an other guide to the share of trade by race is the amount of selling space occup ied . Because they account for almost all permanent vendors, Indian sellers occupy 5-10 per cent more selling space than the proportion of vendors they account for on any rticpa ular day . On Saturday 27 Novemb er 1976 at Ba market , for instance, 35 per cent of vendors were permanent vendors who occupied 51 per cent of selling space . On the same day , 81 per cent of vendors were Indians occupying 86 per cent of selling space. 32 Few Guj eratis sell in the market (only 9 were enumerated in the January 1976 market survey) but their selling behaviour is markedly different from that of other Indian vendors. In commercial experience and family background, dependence on market sales for income , distance of residence from and means of travel to market , assistance in selling and the type of produce sold, Guj erati vendors have mo re in connnon with Chinese than with other vendors. 33 . . q uo t at ion· s are f ram inter· vie. ws wit . h f ive. FiJ. ian permanent vendors at Suva market conducted in January and February 1977 by Tevita Ba. For an edited version of one in terview see Ap pendix. 34 0ne Fij ian non-producer-vendor saw reliance on Fij ian 260

suppliers in this way : 'Fij ians who sell in the market [as non-producer vendors ] encourage Fij ian farmers to work hard and grow more food crops, as the farmers know that there will be Fij ian friends and relatives in the market who will take an d buy their commodit ies without becoming involved in any problems regarding negotiation , purchase and pricing.' (Interview by Tevita Ba, February 1977.) 35 The NMA was established within the Ministry of Connnerce, Industry and Co-operatives but was transferred to the Ministry of Agriculture , Forests and Fisheries in June 1972. The quotation of the Authority's aims is from Sect ion 5 of the Act . The first board of the NMA isolated the following obj ectives : (a) To provide a guaranteed market for specific primary produce at predetermined prices for different pro­ duction zones ; the range of produce to be widened as quickly as markets, in frastructure, staff and other resource capabilities allow. (b) To maintain a stea�y flow of reasonably priced, high quality produce to markets. (c) To assist in Government 's drive to combat inflation by dampening upward spirals in produce prices. (d) To work in concert with all organizations and agencies involved in rural development to ensure that pro­ duction and marketing plans are integrated. (e) To co-ordinate closely with extension services to ensure that production and marketing plans are integrated . (f) While not intending to be a profit-making body , nevertheless, it should act connnercially and aim to do no worse than a break-even posit ion overall. (g) Where marketing bodies or organizat ions already exist, they will be en couraged and assisted provided they are fulfilling a useful role. (h) Where no marketing infrastructure is present, to aim to set one up either by itself or in association with other agencies . (i) To vigorously investigate additional and/or alternative markets both locally and overseas in the interests of maintaining local price and supply stability. (j ) To conduct research in to ma rketing characteristics of various produce so that deman d and product ion pro­ jections can be made with a greater degree of accuracy. ('National Marketing Authority Report for the years· 1971, 1972 and 1973, Par liamentary Paper, 3 of 1975, p.l) . The obj ectives of the NMA are not nearly as broad as suggested 261

during debate over the Marketing Bill . The Minister for Commerce, In dustry an d Co-operat ives in moving the bill , for in stance, claimed that areas of activity would be produce , cattle and handicrafts dealing , warehousing an d storage. Other speakers saw the chance of the NMA earning the profits then made by middlemen . (Parliamen t of Fij i, Par liamentary Debates , Session of 1971, Part 1, pp .32-59 , 62-700) 36 · · · s a 1 es to in· divi. ·d ua 1 consumers wit h.in piJ· i are t h roug h NMA stalls at some markets. That in the Suva market had a gross monthly turnover in 1976 of around $3000. When the NMA established the stall in 1973, permanent vendors in the market feared that it would be used to undercut their prices. Concern was so great that a deputation from the Suva Market Vendors ' Association proposed that non-producer­ vendors would purchase all their supplies from the NMA if the Authority ceased selling retail in the market (NMA file 22/1) . The NMA stall has had a negligible effect on market trade , in part due to its poor location , thus feeling against the Authority has subsided. 37 NMA file 20/2 . 38 NMA file 12/4 . Numerous similar cases are encountered among producers and in NMA records : at Koro in 1976, for instance, the Authority refused to buy dalo from anyone other than three named 'reliable' suppliers. When estab­ lished, it had been the intention of the NMA to make regular shipping runs to the islands to collect produce (The Fiji Times, 14 September 1971) . 39 Market vendors' feelings of not being adequately served are seen in the following example . In Ap ril 1972 , 105 Suva market vendors , who we re predominantly Fij ian opr ducer­ vendors, petition ed the Suva City Council for a reduction of market fees , provision of sleeping facilit ies at the market for out-of-town vendors , reduced market fees for vendors coming from long distances, the conversion of the market 's kava saloon to a restaurant , co-opt ion of producer­ vendors to the Council 's Market Connnittee , and the provision of shelter for stalls and tables outside the market building. Council rej ected all but the last two requests and acted only on the final item (MUDHSW file 749/26/2 , 13 Ap ril 1972). An indi cation of the problems between vendors and council is that for at least the five years since 1972 , Suva market witnessed various conflicts involving the council, police and political activists; there have been similar problems at other markets. 262

40 MAFF file 54/1, 27 November 1975 . Only a very small pro­ port ion of produce handled by middlemen is in fact imported.

Chapter 3

1 As noted at the beginning of Chapter 2, some foods sold in produce markets (ghee , coconut oil, some dried and smoked seafoods and ready-to-eat snacks) are processed , and some produce is sold in shops, but otherwise there is little overlap in products handled between the produce and processed foods distribution systems . 2 An outline of the Min istry 's current interest in connnerce is in Fiji's Seventh Developmen t Plan (Fij i, Central Planning Office , 1975) , especially pp . 162-3 . The relative place of connnerce in development plans may be judged by the fact that wh ile over 40 pages in the Plan are devoted to agri­ culture , forests and fisheries , only 11 are given to connnerce , industry and co-operatives (2 , 7 and 2 pages , respectively) . 3 The broad long-term obj ectives of the Co-operative movement in Fij i have been defined by the Department of Co-operatives (1975 :2-3) as : (a) To provide co-operative trading facilities both in the retail and in the wholesale field. To provide a greater variety of goods and thus stimulate the desire to in crease product ion for the purpose of acquiring wealth to procure material comforts. (b) To provide co-operative saving and credit facilities particularly in the rural cane areas of the country . To stimulate saving in order to enable sign ificant productive lending thus leading to increased production . (c) To assist individual producers or fragmented groups of producers of copra and other agr icultural produce to exploit their individual productive and financial resources so as individually , and as a connnun ity, to benefit from the ext ent to which their influence can be felt in the marketing field. (d) To in fluence an upgrading of quality of local produce as the resul t of insight in to the consumer's requirements. (e) To assist in the transit ion from a subsistence to a connnercial economy by propagating modern and economic me thods of cultivation . (f) To ensure that the maximum return is coming back in to the hands of local producers , thus providing in centives 263

for raising their own living standards . (g) To provide a status symbol an d sense of personal and local ownership in the local people . (h) To provide the opportunity for developing people in the lower educational range to translate their acquired knowledge in to practical commercial value . (i) To train the people in the application of democratic principles to help them to build their own destiny. 4 Trading figures are from the Department's 1974 Annual Report (in press) . Trading figures are not for a twelve­ mon th period but for the in ter-aud it period , wh ich averages almost fourteen months . A large numb er of societies are inactive . See 'Retailing' (below) for further discussion of co-operative societies. 5 An official account of the operations of the Fij ian Business Opportunity and Management Advisory Service is not available . The judgment of the Service 's success is from sources both within an d outside the Service. 6 The annual licence fee un der the Business Licence Ac t, that is, applicable outside gazetted towns and cities, for a commission agent (the applicable category) is $50. In Suva , the fee is $100 and $50 for a manufacturer 's representative ; elsewhere, commission agen t licence fees are generally $50. 7 rt is likely that there are a few indent agents in Nadi and perhaps also in some other towns in addition to Suva an d Lautoka. In addition to the 'independent' agents , the expatriate European-owned firms have their own in dent departmen ts (as do a coup le of larger local firms , like Rab i Holdings Ltd) that indent goods only for the firm. These 'tied ' indentors are not taken into account in this discussion . Information on indent agents is derived from in terviews with a small numb er of indent agents in Suva an d Lautoka and a March 1977 telephone survey of all con­ tractable 'indent ' or 'commission ' agent s and 'manufacturer 's representatives' listed in the 1976-77 Suva telephone directory. 8 Some snack-food, aerated water and non-food manufacturers in Fij i have appointed exclusive 'manufacturer 's represent­ atives ' to represent their interests and solicit custom in Fij i. The trend towards exclusive local distributorships is strongest among non-food lines, at least one of wh ich has started a subsidiary to market its products. 9 In a survey of retail stores in the Central Division (May- June 1976) storeowners named 12 7 different companies as 264

supply sources , mo st of which operate in the Suva-Nausori area . Only 33 of these companies, however , were listed as suppliers for five or more stores. The 33 most frequently named suppliers accounted for 80 per cent of the 755 named suppliers (including repeated suppliers) . Morris Hedstrom, Burns Philp , a Chinese company in Suva and two Indian com­ panies in Nausori accounted for 36 per cent of named suppliers; Morris Hedstrom and Burn s Philp together accounted for 22 per cent. lO The only figures available on who lesaling returns are for Regional Co-operative Associat ions (Table 3.2) . Trading results of co-operative societies are not representat ive of other wholesalers because co-operative societies have service rather than profit mot ives , receive un charged support services from the Department of Co-operatives and a degree of 'captive trade ' from society members, and often operate in a non-competitive environment. Nevertheless, co-operat ive wholesale trading results are useful in that they do suggest returns for one form of wholesaling : four of sixteen association s made net losses and two broke even ; the average return of associations with a net profit was 4 per cent of gross sales . 11 Goods handled by co-operative societ ies are basic groceries (canned fish , salt, flour , rice , sugar, dry biscuits, etc.) and other 'essentials' like soap , matches , cigarettes, kerosene and confectionery . A number of associations have applied for a licence to sell liquor. A proposal to open FCA branches in Lautoka and Labasa or Savusavu is being considered within the Department of Co-operatives. It is unclear whether the proposed branches would operate as independen t importers and purchasers of local supplies or would be supplied, at least in imported goods , from the Suva head office . Information on co-operative wholesale activities is from interviews with staff of the Department of Co-operatives in Suva , FCA and a number of regional wholesale associat ions in Viti Levu , in addition to Fiji, Central Plann in g Office (1975) and Fij i, Department of Co­ operatives (1975) . The FCA is only one in a line of centralized import and wholesale schemes that have been associated with the co­ operative movement since its earliest years . In 1949, for in stance, the Fij i Co-operative Purchasing and Distribution Society was formed to allow memb ers to buy in bulk and so obtain discounts. (In 1951 the society was deregistered because it was 'unsuited to co-operative control' . Department of Co-operatives , Annual Report , 1951.) 265

12 Pricing practices are reputedly mo st arb itrary in the islands beyond the effective control of PIB in spectors . For in stance , in early 1977 the Lakeba Regional Who lesale Association was putting a 10 per cent markup on all grocery lines irrespective of permissible price control markups. 13 0ne ropeaEu n-owned importing wholesaler-retailer who imported all food lines was active in Suva in 1975-76. The business has since been drastically curtailed and today the firm's grocery trade (still all in imported items) is primarily retail and is only a small part of its total business. 14 Burns Philp (South Sea) Co . Ltd is a subsidiary of Burns Philp and Company Ltd of Sydney whose operation s are wholesale merchants ; shipping, travel and general agents; importers, plantat ion owners ; trustees; finance > steel , glass and liquor merchants, motor dealers ; hotel owners; distributors of drink dispenser machines ; electrical wholesalers and manufacturers; min ing investor; office machinery distributor. The company has 160 subsidiary companies, 126 of wh ich are in corporated in Australia, 17 in Papua New Guinea, 7 in Fij i, 5 in the New Hebrides and the remaining 5 in the UK, USA, Hong Kong an d New Zealand. In 1976 , total sales of Burns Philp and Company Ltd and subsidiaries were $343 million with a gross trading profit of $15.1 million ($9.5 million post-tax) ; net operating was $7.5 million , 76 per cent of which came from subsidiary companies. Total assets were valued at $339 mill ion and over 12 ,000 people were employed by the company and sub­ sidiaries. Dividend for the year was 15 per cent. In addition to activit ies of Burns Philp (South Sea) , the Fij i subsidiaries are involved in sales and service of motor vehicles and office equipment, glass, wrought iron and steel manufacture , manufacturers ' representat ion and importing services. Ninety per cent of the 1976 operat ing . profit of $0. 996 million of the Fij i subsidiaries came from Burns Philp (South Sea) and motor vehicle act ivities. Trading profit of Burns Philp (South Sea) for the year ended 30 June 1976 was $1.894 million before tax an d $1. 335 million after tax; net operating profit was $0. 565 million . Burns Philp (South Sea) is involved in wholesaling, retailing, motor vehicle distribution , shipping and travel agencies, manufacturing, agriculture , finance and investmen t and operates through 14 locations, in cluding 3 in Tonga, 2 in Samoa and 1 in Niue , in addition to those in Fiji. One of the Fij i out lets, at Ba , ceased operations early 266

in 1977 . Informat ion on performance of sub-company branches or divisions is not available. Over 700 people are employed in the merchandising division in Fiji, which company sources claim did not make a profit until mid-1976. (Currency is $ Australian . Source : Annual Reports for year ended 30 June 1976 of Burns Philp (South Sea) Co . Ltd, an d Burns Philp and Company Ltd.) 15 Mo rris Hedstrom Ltd is a sub sidiary company of W.R. Carpenter Holdings Ltd of Sydney . The holding company has 72 sub­ sidiaries , 27 in corporated in Australia, 19 in Papua New Guinea, 16 in Fij i, 4 in the New Hebrides , 2 in the UK and 1 each in the USA, the Bahamas , New Zealand and Western Samoa. The group 's trading profit for the year ending 30 June 1976 was $10.4 million on trading revenue of $160 million . Net operating profit of the group was $8.1 million , 80.2 per cent of which came from subsidiary companies, 13.6 from W. R. Carpenter Holdings Ltd, and 6.2 per cent from equity companies. The Fij i subsidiaries had a net profit of $3.9 million and tho se in Papua New Guinea, $2 .6 million ; Australian subsidiaries had a net loss of $0.3 million . The three equity companies operating in Fij i (Carlton Brewery (Fij i) Ltd, Wailekutu Meats Ltd, and Fij i Paints Ltd) had a net pro fit of $264 ,000. Principal activit ies of Carpenter subsidiaries in Fij i are merchandise wholesaling and retailing, automotive and avyhe earth moving equipmen t sales and service, rental cars , coconut oil manufacture, copra production , steel fabrication , foundry electroplat ing, ship and barge construct ion and repair, property development, insurance, electrical con­ tracting, joinery , upholstery , shipping, stevedoring, brewery , paint manufacture , butchery and small goods manu­ facture. Elsewhere , group activities comprise illuminated sign s an d lighting, finance and property development, in surance, textiles , vineyards , investment, merchandise indenting and conmiodity marketing (in Australia) ; merchan­ dise wholesaling and retailing and automotive retailing (in Tonga and Western Samoa) ; cocoa, copra, tea and coffee production , coconut oil manufacture , merchandise whole­ saling an d retailing , shipping agencies , liquid gas dis­ tribution , paint manufacture and in surance (in Papua New Guinea) ; and merchandise indenting, connnodity marketing and insurance (in the UK and New Hebrides). Morris Hedstrom Ltd, the company's chief merchandise wholesale-retail outlet in Fij i, made a net profit of $1.6 million from 11 outlets in Fij i (one of wh ich has since closed) and 6 in Western Samoa and Tonga. Woolworths Ltd 267

had a declared net profit of $383. Half ($2 mill ion) the net profit of Carpenter subsidiaries in Fij i came from Carpenters Fij i Ltd, a company dealing in land developing, finance activities and services at normal connnercial rates for other Carpenter subsidiaries. (Currencies is $ Aus tralian . Source : W. R. Carpenter Holdings Limited, Annual Report, 197 6.) 16 The chief relat ed activit ies of the processors are : pro­ duction of sweet biscuits, coconut oil , edible oils, margarine and fats (for industrial users) , soap , detergents and plastic products by Cope Allman (South Pacific) Ltd, one of the two dry biscuit producers ; operation of a bakery by the other biscuit manufacturer (Lee's Trading Co .); production of animal feed by Visama Rice Mill Ltd; Punj a and Sons in Lautoka packs tea and oil and mills rice in addition to being involved in general retailing and whole­ saling, soap manufacture, hotel ownership , transport and other activitie s. 17 . FSC sells sugar (and rice from the government-owned Rewa Rice Limited mill) on credit to cane farmers by delivering orders to producers at sugar sector offices but otherwise plays no active part in the internal dist ribution of sugar. The cost of sugar , rice and company-supplied farm inputs are deducted from the farmer 's next cane crushing income . 18 Although FSC apparently advises wholesalers that sack contents are approximately 224 lb , sacks are not weighed or repacked by wholesalers before sale to retailers . Some retailers are under the impression that each sack is 224 lb net . In fact , net weight is occasionally 10-12 lb under 224 lb . For the 1977 season , sugar at Penang mill was to be sold in sacks marked 50 kg net weight. Irregular sack weights have been attributed to inadequate weighing f aci­ lities. Installation of mo re efficient scales at Penang has made accurate sack weight possible. Similar scales will be in stalled in other mills in 1978-79. 19 rt is claimed that the mill ' s products in fact are consider­ ably more expensive than imported products. Early in 1977 when the mill ceased operations for a short period to un dertake expansion, some wholesalers and bakers were granted licences to import flour from Australia . These importers were landing baker 's flour in Fij i at $264 a tonne (cif) , when the most recent FMF ex-mill price was $330. In May 1977, the ex-mill price of flour was dropped to $243 after the company announced a substantial 1976 operating profit and then withdrew its accounts. The 268

government appointed an independent au ditor to prepare a set of accounts and conrrnenced enquiries in to the company 's activities. 20 n ai · 1 y mi"1 1 capacity. o f 158 tonnes was to b e in . creased to 220 tonnes by Ap ril 1977. Prior to expansion , the mill was able to meet local demand. Ap proximate pre-expansion monthly production was : sharps, 1650 tonnes ; baker 's flour , 600) ordinary flour , 600, bran, 400 (including 350 for export); pollard , 300; and about 130 tonnes of semolina, brown atta, wholemeal fl9ur and wheat germ. 21 The competitor, Lee 's Trading Co. , actually ceased production in May 1977 over a dispute concerning the recognition of an emp loyee 's union . It is unknown when the company will resume biscuit production. 22 The dairy's 'long-life ' milk is marketed by contracted suppliers who use trucks and other equipment supplied and maintained by Rewa Dairy . Butter sales are largely ex factory , a 5 per cent price dis count applying on purchases over 2400 lb , the min imum purchase un it being one 60 lb case . 23 cap . 154 (Licences), S.17, The Laws of Fiji (rev. ed. ), 1945 . 24 The Nausori (Hawkers) By-Laws , 1973, are typical of hawker legislation in their definit ion of hawker ('a person who sells goods from a handcart, box, basket , tray and the like , or from a boat , horsedrawn vehicle , bicycle, motor vehicle or any other type of conveyance; but does not include the holder of any licen ce under the Licence Ordi­ nance or any employee of such holder selling or supplying goods in respect of which such licence is held ') and delimitation of goods that can be sold: '(a) Any food or drink other than liquor ; (b) cigarettes , cigars , tobacco and matches; (c) curios and handicrafts, in cluding mats, baskets, fans, articles made of straw and grass skirts; (d) plants, fen1s and cut flowers ; (e) jewellery , ornaments and souvenirs ; (f) live poultry and eggs , (g) (i) books , (ii) filigree jewellery , imitation jewellery and ladies even ing bags, (iii) dolls dressed in traditional Fij ian costumes or portraying Fij ian police officers ; (iv) postcards ; (v) fruit and vegetables; (vi) scarves , cushion covers , beaded purses, men 's belts, with the word "Fij i" thereon the outside thereof; (h) wares of other Pacific islands origin , but not in cluding any produce ; article or thing made, manufactured, prepared in , or 269

originating from any territory outside Fij i, with the exception of the articles specified in paragraph (g) of this definition . ' Outside gazetted town and city boundaries , hawkers are licensed under the Business Licence Act, 1976, for $5 p.a. Elsewhere , the annual licence costs up to $15 (as it does in Nausori). 25 About 23 per cent of shops in Suva sell some groceries . In Nausori town , 68 retail licences were issued in 1976, 21 for shops dealing in foods tuffs. Ten of the nineteen retail-wholesale licences issued in Nausori were held by merchants selling foodstuffs . In Navua, one-half of the 31 retail licences and both the retail-wholesale licences were for grocery shops. All but one of the twelve retail licence holders in Korovou sold some groceries. 2 6 Memb ership of co-operative societies (especially consumer and marketing societies) is predominantly Fij ian . In the Central Division , one rural consumer co-operative with an exclusive Indian memb ership was located. The racial com­ position of memb ership of urban societies is more diverse than that of rural societies , but Indian, Chinese and European participation is slight. 27 Estimates of share of retail trade were derived from a number of sources for each 'store-type ' (primarily the studies and data reported in Tables 3.24, 3.25) and then extrapolated to the total number of stores (Tables 3.4, 3.5) . 28 The retail store survey sought basic information about the location , nature and function of retail food stores in the Central Division with the general hypothesis in mind that : 'Variation in food retailing through stores in the Central Division is in part related to store ownership features, particularly the race of proprietors, form of business operation and access to supply sources .' The survey was conducted with a questionnaire that covered store location , ownership , workforce , size and nature of stock, and sources of supply . Questions on turnover and pro f it rates were not included as it was felt better indi­ cations of these could be indirectly gained. Also , it was planned that a later study of a small number of selected stores (cf . Table 3.24) would investigate the topic more adequately than would be possible in the survey . The questionnaire consisted of two parts . The first dealt with store establishment, ownership , location , and physical form and general stock composition and supply, as well as 270

personal in forma tion on the own er . The second was a detailed inven tory of all foodstuffs on display and in storage . Stores with retail-wholesale licences were excluded from the samp le in Suva city, but elsewhere were included. Two two-person interview teams visited selected stores in May and early June 1976. The interviews were held with shop owner-operators (or man agers wh ere appro­ priate, as in co-operative society stores) . If the own er was not available or the store closed, an effort was made to return later: apart from stores that were not readily accessible, shops where there was no initial contact with the owner were visited twice . If the in terview could not be conducted on the second visit , a nearby store was sub­ stituted . A preliminary report of the survey results is , Michael Baxter , 'A survey of retail food stores in the Central Division , Fiji' , CASD, USP, November 1976. Unless otherwise noted, all data on Cen tral Division retail grocery stores are derived from the survey. 29 As i. d e f rom in. tervie. ws wi . t h store owners ( w h ose recounti. ng of long-previous business activities is not always accurate) , a useful source on store history is the business licence registers . However, each register covers about ten years and completed registers are soon discarded . At present, few registers covering the period before the mid-1960s are available. 30 Local government bodies have jurisdiction over hours of business activity. The Suva Shops Closing Order (8 July 1966) , for instance, establishes maximum hours of shop operation within Suva as 6am to 6pm, except lpm closure on a weekly half-holiday and 9pm closure on Friday . Exemption Orders excuse grocery and some other shops from compliance with the regulations. 31 For the purpose of the retail store survey , confectionery , ice-cream and soft drinks were not regarded as foodstuffs. It is likely that the 22 shops enumerated as selling 'only foodstuffs' also sold some of these items . 32 A 'food type ' is a generic food grouping without reference to variety , such as 'canned meat ', 'canned fish' , 'tea ' and 'sauce '. A 'food item ' is the connnon selling un it of a food (a can of mackerel, etc. ). Products stores in bulk and broken into smaller units for sale are converted to the number of composite connnon selling units to determine the number of food items . For in stance , since rice is usually purchased by the pound, a 100 lb sack of rice is 100 'food items'. Dry goods pre-packed in units larger 271

than on e pound - a practice connnon on ly in supermarkets - were converted to one pound un its for enumeration as 'food items '. 33 'Local foods ' are taken as frozen chicken , bottled milk, butter, ghee in 26 oz or smaller containers , canned tuna, local fish, dry biscuits, sugar , bread, coconuts, yaqona and a range of local fruits and vegetables like bananas and eggplant. Some of these foods are occasionally imported, e.g. butter , eggs and larger containers of ghee , and others are manufactured locally from imported ingredi­ ents (dry biscuits and bread) . 'Basic foods ' are beef, mutton , mackerel, milk, dripping, edible oil, ghee, flour , sharps, dry biscuits, salt, sugar , tea, potatoes, onions and garlic. Some other foods that are basic to diet (such as dalo) , are insignificant in store transactions. 34 Factors other than distance from Suva and accessibility affect the size and composit ion of grocery stocks (e.g. distance from actual supply centre , rural/urban location , and clientele) , these have not been analysed . 35 of the surveyed private stores, 54.6 per cent were owned by Indians, 24.2 per cent by Fij ians, 20 per cent· by Chinese and 1.2 per cent by part-Europeans. The 14 club stores and 68 co-operative society stores are regarded as Fij ian . Ownership of the 4 companies was divided equally between Chinese and Indian firms . 36 In the retail store survey, personal data of storeowners were collected only for individual owners of private stores or, where applicable, the dominant member of a partnership . Valid responses were obtained from all but 3 of 263 surveyed private stores. Occasionally a store is licensed under one name but is operated by another person (usually a member of the licence-holder 's family) . In the survey , personal data were collected for the licensed owner unless it was clear that it was a nomin al proprietorship . 3 7 Because of a slight urban bias in sample selection , racial representation among the surveyed stores is slightly more Indian and Chinese than for all food shops. In the Central Division , survey an d total representat ion by race is : Fij ian 41 .2 per cent of stores surveyed, 48. 7 per cent of all stores in Central Division ; Indian 42 and 33.4 per cent; Chinese 15 .1 and 12 .4 per cent. 38 Patronage of the supermarkets and shopping centres was investigated in Jtme 1976. In the three shopping centres, 272

844 customers were in terviewed over six consecutive days. Interviewers worked in teams that systematically covered all trading periods (from before 7am until 6.30 or 7pm) , Monday to Saturday midday . At the supermarkets, 824 inter­ views were conducted during consecutive periods (morning, midday , afternoon , mo rning) from Wednesday to Saturday in on e week. Informat ion recorded at both supermarkets and shopping centres in cluded the cus tomer 's residence location , race , posit ion in household, employment of household head , and a list of purchased items , price paid and method of payment. A preliminary report of findin gs of the super­ market survey is Michael Baxter, 'Patronage of supermarkets in Suva : report of a survey conducted in June 1976', (CASD , USP, February 1977) . Data from the shopping centres studies have been processed but not written up apart from the basic data in Table 3.23. Unless noted, information on supermarket and store custom is from the patronage surveys. 39 Me at pro duct ion . , im . portlllg . dan mar k eti . ng were not examllle. d in detail. Information here is from general observation and enquiries made with trade and government sources. Butchers in Fij i deal in local pork , goat and fish as well as imported frozen (and less commonly , chilled) beef, mutton, lamb , goat ar1 d fish. Aside from the expatriate super­ markets, butchers are the only retail outlets for non- frozen meat . At least two butcher chains wholesale meat to retail stores . The bulk of poultry supplies are in the hands of local producers who process and eefr ze birds prior to sale to retailers. One of the butchers , Wailekutu Meats Ltd in Suva , is 50 per cent owned by W.R. Carpenter Holdings Ltd. It is expected that the meat supply of Woolworths and Morris Hedstrom at least in the Central Division , will be met by Wailekutu Meats. 40 . Data on sa1 es activity. an d ot h er in. f ormatio . n concerni. ng the shop described here , and others that follow, come from two sources . The more important is a study of the sales activity of twelve shops in the greater Suva-Nausori area conducted in early December 1976. Over six consecutive days all transactions were monitored for between 60 and 80 per cent of the time the shops operated , the periods being selected so that all phases of daily activities were systematically covered. In each tran saction the item pur­ chased, amount paid, form of payment and the general residential location of the cust omer (neighbourhood, connnunity or passing regional trade) were noted. With such coverage, statements about the nature and volume of trade 273

over the week could be made as could proj ection s of trade to annual rates. The possible un iqueness of the mon itored week should be kept in mind. The second source of data is a series of interviews with the owners of the twelve stores . In terviews focused on the mo t ivation and sources of capital for store establishment , subsequent operations and the en trepreneurial activities of the storeowner . In formation that might enable identification of storeowners has been omitted (although mo st said they would not obj ect to being identified) : names are fictitious . Trade characteristics of the twelve monitored shops are sunnnarized in Table 3.24. 41 Krishna initially in tended to keep his job and have his wife look after the shop while he was working. It became apparent , however , that his wife could not manage the shop (she was 'too shy'), and so he resigned. 42 0ne reason Krishna started his shop in the face of estab­ lished opposit ion was that some customers , including his own family, had found the other shopkeeper 'a difficult man'. Krishna thought he could cater to the rej ected and dispirited customers. The other shopowner tried to talk Krishna out of starting a shop , and later accused him of 'stealing customers '. Some people buy at both shops but customer allegiance became fairly well defined within a year or so after the opening of Krishna 's store and has not changed greatly since. 43 societies were considered active if recent (19 73-76) audited accounts records are maintained by the audit section of the Department of Co-operatives in Suva . In addition to consumer and con sumer-marketing societies, in 1974 there were 147 thrift and credit , 5 marketing and 45 'other' (mainly land and housing) societies; there were also 93 'unregistered ' ('probationary '?) societies . 44 Trading figures are for all trading of societies (i.e. for store and marketing activity). The proportion of societies with stores that are consumer societies ranges from about 4 per cent in the Northern and Eastern Divis ions to 76 and 98 per cent in the Western and Central Divisions. The 1973-76 accounts period saw good prices for copra , the main item marketed by the Eastern and Northern consumer­ marketing societies, thus their trading volume and profit/ loss rates are probably better than usual . 45 B usi· ness l"icence regis. ters are a potenti. a 1 source o f information on period of store operation as licence fees are paid quarterly. However, registers more than about ten years old were not located. One indication of the 274

persistence of stores comes from the Central Division retail store survey . The average operation period of all stores is just over 12 years . Surveyed co-operative societies stores had been in existence an average of 8.3 years and Fij ian stores , 5.1 years . Re ferences

Aylmer, F. (ed. ), 1860 . A Cruise in the Pacific, from the Log of a Nava l Officer (2 vols) , London , Hurst and Blackett.

Belshaw, C. S. , 1957. The Great Vi llage : the Economic and Social We lfare of Hanuabada and the Urban Cormru:nity in Papua , London , Rout ledge and Ke gan Paul .

Biggs , H. C. , n.d. Report on the Marke ting of Agricultural Produce in Fiji , London , Ministry of Overseas Development .

Blackie , W. J. , 1949 . 'Agricult ural research and policy in Fiji' , Fiji Agricultural Journa l, 20(4) : 98-120.

Brewster, A. B. , 1937. King of the Cannibal Is les. A Ta le of Early Life and Adven ture in the Fiji Is lands , London , Robert Hale .

Brookfield, H. C. and Hart , D. , 1971. Me lanesia. A Geographical In ter­ pre tation of an Is land Wor ld, London , Methuen .

Burns , A. , Watson , T.Y. and Peacock , A. T. , 1960. 'Report of the commission of inquiry in to the natural resources an d population tren ds of the Colony of Fiji' , Legislative Council of Fiji, Council Paper 1, 1960.

Calvert, J. , 1870. Fiji and th e Fijians; and Missionary Labours among the Cannibals (3rd ed . ), London , Hodder and Stoughton .

Casley, D. J.L. , 1969 . ' Report on the census of agriculture 1968' . Legis­ lative Council of Fij i, Council Paper 28, 1969.

Cooper, H. S. , 1880 . Cora l Lands (2 vols) , London , Richard Bentley .

Couper, A. D. , 1967. The island trade : an analysis of the environmen t and operat ion of seaborne trade amon g three island groups in the Pacific. Ph .D. thesis, Aus tralian National University , Canberra .

Fij i, 1967. The Laws of Fiji (rev . ed . , 10 vols) , Suva , Government Printer .

Fij i, Bureau of Stat istics , 1968. ' A preliminary report on the 1968 urban household income and expenditure survey in Fiji' , Suva , Bureau of Statistics .

1972 . ' A report on the urban household income and expenditure survey in Fij i 1972' , Suva , Bureau of Statistics .

1974. ' Fij i household income and expenditure survey 1973' , Suva , Bureau of Statistics .

1976. Current Economic Statistics , October 1976, Suva.

275 276

Fij i, Central Planning Office , 1975. Fiji 's Se venth Deve lopmen t Plan 1976-1980, Suva , Central Plann ing Office.

Fij i, Department of Co-operatives , 1975. ' Policy and development plan for co-operatives 1976-1980' , Suva , Department of Co-operatives .

Fij i Development Bank , 1976 . ' Annual report for year ended 30th June 1976' .

n.d. ' Information regarding commercial and industrial loans to Fij ians ' .

Fiji, Parliamen t of (annual). ' The Trade Report ' . Parliamentary Paper .

Fij i, Prices and Incomes Board , 1975. Fighting Inflation in Fiji . An Account of the Work of the Prices and In comes Board During Year ended 30th June 1974, Suva , Government Printer.

n. d. In fW.tion in Fiji . A Further Account of the Wo rk of the Prices and Incomes Board, Suva , Prices and Incomes Board .

Finney , B.R. , 1965 . ' Economic change and dietary consequences among the Tahitians ' , Micronesica , 2:1-14 .

Geertz , C. , 1963. Peddlers and Princes: Social Change and Economic Modernisation in 'I'wo Indon esian Towns, Chicago .

Harvey , C. , 1943. 'A further note on the Nausori Fij ian co-operative market ' , Fijian Agricu ltural Journal, 14 (2) :36.

In ternat ional Labour Organisat ion (ILO) , 1972. Employment, In comes and Equa lity: a Strategy for Increasing Productive Employment in Kenya, Geneva , ILO.

Kamekamica, J.N. , 1966. ·Marketing of Fij i ' s agricultural products, M. Comm. thesis , University of Melbourne.

McGee , T.G. , 1973. ' In praise of tradition : towards a geography of an ti-development ' . Paper delivered at the 45th ANZAAS Congress, Perth.

1975. ' The persistence of the proto-proletariat : occupational structures and plann ing for the future of third world cities ' . Seminar paper , Department of Human Geography , Research School of Pacific Studies , the Australian Nat ional Un iversity, Canb erra.

Moss, F.J. , 1870. A Planter 's Experience in Fiji . Being a Concise Account of the Coun try, its Pre sent Condition and its Prospects as a Field for Emigration , Auckland, Jones and Tombs.

Moyer, R. , 1965 . 'Marketing in economic development', Occasional Papeps in In ternationa l Business, l, Graduate School of Business Adminis­ tration , Michigan State University , East Lansing.

and Hollander, s. c. (eds) , 1968. Markets and Marketing in Developing Economies . Homewood , Illinois , Richard D. Ir,win . 277

Patterson , D. D. and Do dds , K. S. , 1945 . 'Recommendations for the re organ­ isation and effective operat ion of the Departmen t of Agriculture , Fiji', Legislat ive Council of Fij i, Council Paper 29 , 1945.

Pechy , W. C. , 1870. Fiji Cotton Cu l ture and Planters ' Guide to the Is lands , London , Jarrold and Sons .

Santos, M. , 1973. ' Spatial organisation in the Third World : the two urban fields '. Seminar paper, Departmen t of Geography , University of Toronto .

Shephard , C.Y. , 1944 . 'Report on agriculture policy for Fij i and the We stern Pacific High Commission territories ', Legislat ive Council of Fiji, Counci l Paper 24 , 1944 .

Spate , O.H.K. , 1959. ' The Fij ian people : economic problems and prospects ', Le gislative Council of Fij i, Counci l Paper 13, 1959 .

Stockdale , F. , 1938 . 'Report ... to the Secretary of State for the Colonies on his visit to Fij i in 1938', London , Colonial Office .

Surridge , H. R. , 1940. ' A note on Lawaqa market ', Fiji Agricul tural Journal , 11(2) :32.

Thomson, B. , 1896. Fiji fo r Touris ts , London , The Canadian-Australian Royal Mail Steamship Line .

Thurston , J.B. , n.d. Fiji, The Navigation and Trade 1874-5 .

Ward , M. , 1970. The National In come and Balance of Payment Accounts of Fiji , Suva, Bureau of Statistics.

Webb , A. J. , 1884 . 'Fij ians and civilizat ion ' , The Victorian Review , July 1884 :247-55.

279

DEVELOPMENT STUDIES CENTRE PUBLICATIONS

Monographs No.l Labour and the Law in Papua New Guinea, D.W. Smith, 1975. 129 + vii pp. A$4 .00. No.2 The Ro le of the Papua New Guinea Defence Force , Paul Mench, 1976. 231 +xii pp. A$ 6.00. No .3 Education , Employmen t and Migration in Papua New Guinea, J.D. Conroy , 1976. 248 + xii pp . A$ 6 .00. No .4 Chimbu: Issues in Development, Diana Howlett, R. Hide and Elspeth Young with J. Arba, H. Bi and B. Kaman , 1976. 384 + xl ix pp. A$6 .00.

No .5 Professional Structure in South East Asia, ed . T.H. Silcock , 1977. 306 + xi pp . A$6 .00. No.6 Educat ion and Nation-building in Plural Societies : the We st Malaysian experience , Chai Hon-Chan , 1977. 116 + ix pp. A$6 .00. No.7 Bougainville in Transition , T.K. Moulik , 1977. 200 +xii pp . A$6 .00. No .8 Man and Environment in Eastern Timor, Joachim Metzner, 1977. 380 + xxi . A$6 .00. No .9 Micro-macro Links in Planning , P.B. Krishnaswamy , 1977. 116 + ix pp . A$6 .00. No.10 Appetite for Educat ion in Contemporary As ia, ed. K. Orr , 1977. 165 + ix pp. A$6 .00.

No .11 The Adaptation of Traditional Systems of Agriculture , ed. E.K. Fisk, 1978. 400 + x pp . A$9 .00.

No. 12 Taim Bi long Man i: the evolut ion of agriculture in a Solomon Island Society, John Connell , 1978. 271 +xii pp . A$ 9 .00. No .13 Practice Without Policy : genesis of local government in Papua New Guinea, D.M. Fenbury , 1978. 309 +xvi pp. A$9 .00. No. 14 Customs of Re spect : the traditional basis of Fij ian communal politics , John Nat ion , 1978. 165 + xx pp. A$ 8.00. No.15 Agricultural Proj ect Design andEvaluation in an Island Community , Alan Bollard, 1979. 228 + xiii pp. A$9 .00. No .16 In ternat ional Aid : some political , administrative and technical realities , R. T. Shand and H.V. Richter (eds) , 1979. 276 +xiv pp. A$ 9 .00. No.17 Forestry in National Development : production systems , con serv­ ation , foreign trade and aid, K.R. Shepherd and H.V. Richter (eds) , 1979 . 245 + x pp . A$9 .00.

No.18 Indonesia and Aus tralia : the politics of aid and development since 1966. Phillip J. Eldridge , 1979 . 224 +viii pp . A$ 8.00.

No .19 Corned Beef and Tapioca : a report on the food distribut ion system in Tonga, Epeli Hau'ofa , 19 79 . 169 +xii pp . A$ 8.00.

No .20 Pres tige and Profit : the development of entrepreneurial ab ilities in Taiwan , 1880-1972, Manfred Steinhoff , 1980.A$8.00.

No .21 The Political Economy of Political Development : a case study of regional development in the Philippines before martial law , Mohd A. Nawawi , 1980 . 79 + xv pp . A$6 .00. 280

No .22 Food in Fij i: aspects of the produce and processed food distri­ bution systems , M.W.P. Baxter , 1980. 276 + xx pp . A$9 .00.

No.23 Rur al-born Fij ians and ludo-Fij ians in Suva : a study of movements and link.ages , Shashik.ant Nair , 1980 . 94 + x pp . A$8 .00.

Occasional Papers No.l Motivational Factors Affect ing Bougainville Villager 's Money Earning Activity, T.K. Mo ulik , 1975. 27 pp. A$2 .00. Now out of print . No.2 Food Dependency in the Pacific : a preliminary statement, T.G. McGee, 1975. 19 pp . A$2 .00.

No.3 Burma's Rice Surpluses : accounting for the decl ine , H.V. Richter, 1976. 51 pp. A$2 .00. No.4 The Fading of Earthbo\IDd Compulsion in a Hong Kong Village : populat ion mob ility and it s economic implicat ions , Rance Pui­ leung Lee , 1976. 15 pp . A$2 .00.

No .5 Th e Defen ce Forces of Small States : a comparison of defence statistics in 26 states of Tropical Africa, Papua New Guinea and Fij i, Paul Men ch , 1976. 32 pp . A$2 .00. No .6 Hun ting and Gathering: the forage economy of the Siwai of Bougainville, John Connell , 1977. 28 pp. A$2 .00. No .7 The Limits of Economics for the Study of Development, R.M. Sundrum, 1977. 16 pp. A$2 .00. No .8 Public Housing in the City States of Hong Kong and Singapore , David Drakakis-Smith and Yue-man Yeung, 1977. 17 +iv pp. A$2 .00. No .9 The Island of Niue : development or dependence for a very small nat ion , E.K. Fisk, 1978. 12 pp . A$2 .00. No .10 The Regional Economy of Bougainville :. growth and structural change , M. L. Treadgold, 1978. 81 pp . A$ 2 .00. No.11 The Structure of Rural Supply to the Honiara Market in the Solomon Islands , M. A. Bathgate , 1978. 65 + iv pp. A$2 .00. No.12 Primary Industries in Papua New Guinea, Julius Chan , 1978. 11 + iv pp . A$2 .00.

No .13 Industrializat ion in Indonesia: development and prospects, Peter Mccawley , 1979, 88 pp. A$2 .00.

No . 14 Technical As sistance : towards improving the underlying framework , Rodney C. Hills, 1979. 30 pp . A$ 2 .00.

No .15 Social and Economic Aspects of Tidal Swamp Land Development in Indonesia, William L. Collier, 1979. 70 PP · A$2 .00.

No .16 Surveys of Free Re sour ces : is there a role in planning and proj ect implementation ? Ro dney C. Hills , 1980 . 30 pp. A$2 .00.

No .17 Replacing Impo rted Food Supplies to Port Moresby, Papua New Guinea, Geoff T. Harris , 1980 . 22 +vii pp. A$2 .00.

No .18 Elitist Involvement in Agriculture in West Africa, Rowena Lawson , 1980. 66 pp . A$2 .00. 281

No . 19 In ternational Civil Aviation in the South Pacific : a perspective , C. C. Kissling, 1980. 62 pp . A$2 .00. No.20 Economic Ac tivities of Women in Rural Java : are the data adequate? Hazel Moir, 1980. 44 pp . A$2 .00.

MADE Research Series

No .l The dynamic replacement problem in the rubb er industry of Sri Lanka , S.K.W. Jayasuriya, 19 76. 112 + xi pp . A$ 5.00. No. 2 Rational farm plans for land settlement in Indonesia , M.A. Wardhani, 1976. 136 + xiii pp . A$5 .00.

No .3 The Evaluation of Agricultural Loans : a case study of Development Bank financed cattle proj ects in the Markham Valley of Papua New Guinea , Desmond P. Brunton , 1980 . 157 + xii pp . A$ 5.00 Thesis Ab stracts

No.l Studies in the economics of agricultural development , D.M. Etherington and B. Carrad (eds) , 1976. 45 +vii pp . A$3.00.

No .2 Studies in the economics of agricultural development , D.M. Etherington and B.C.F . Boucher (eds) , 1979. 55+ix pp. A$ 3.00

MA in Demography Research Series No.l An Exploratory Demographic Study of the Nupe of Niger State: the case of Sakpe Village , Abubakar Katcha, 115 + ix pp. A$5 .00. No .2 The Value of Children Among Tea Estate Workers ' Families : a case study in a village of West Java , Indonesia, Asep Djadj a Saefullah , 1979. 128 + xiii pp . A$5 .00.

Teaching Notes Series

No .1 Basic Mathematics for Demographers , S.K.Jain , 1979 .A$5.00.

No .2 Beginning Population Studies, edited by David Lucas , 1980 . A$5.00.

Pacific Research Monographs

No. 1 Technical Training and Development in Papua 1894-1941, Tony Austin , 1978. 204 + xiv . A$8.00 . No.2 Villagers at War: some Papua New Guinea experiences in Wo rld War II, Neville K. Robinson,1979. 224 + xix. A$8.00. No. 3 A Thousand Graduates : conflict in university development in Papua New Guinea , 1961-1976, Ian Howie Wi llis , 1980. 362 + xiv. A$8. 00.

No. 4 Viceroy of the Pacific: a life of Sir John Bates Thurs ton , Part II of The Maj esty of Colour , Deryck Scarr , 1980. 336 + x.

No. 5 Brown or White? A History of the Fij i Industry , 1873 - 19 73 , Michael Moynagh , 1980. 311 + x. A$8.00. 282

Publicat ions are available from the ANU Press, PO Box 4, Canberra , ACT , Aus tralia. Please enclose money including postage with order . Postage rates are : Monographs : O ' seas A$2 .00, Au st .: NSW $1. 15, ACT 90c , SA, Vic. , Tas . and Q'land $2 .25, WA and NT $2 .60. Occasional Papers : O'seas 55c, Aust. 45c .

Subscriptions to the Monograph series are available for five issues for A$40.00 plus postage (O'seas A$10.00; Aust .: WA and NT $13 .00, NSW $5.75, ACT $4 .50, SA, Vic. , Tas . and Q'land $11.25. Sets of the first ten issues in three bound volumes are avai lab le at a cost of A$ 80 .00 plus postage (O' seas A$5 .10, Aust .: WA and NT $4 .60, NSW $1.55, ACT $1.10, SA, Vic. Q'land and Tas. $3.25). .. Distributed by ANU Press Michael Baxter has researched and worked on a variety of PO Box 4 Canberra ACT 2600 issues in rural small producer development, particularly Jn Austral ia Papua New Guinea, Fiji, Brazil and Mexico. He was edutated in Australia , Papu a New Guinea and the United States(Ph.D., ISBN 0 909150 03 6 Geography, University of California, Berkeley). Since leaving ISSN 1057-5767 Fiji on completion of the study reported here, he has been employed by the Worl d Bank in Washi ngton, works on rural development topics.