ReportNo. 8544-TON Kingdomof AgricultureSector Strategy Review

October10, 1990 Public Disclosure Authorized AgricultureOperations Division CountryDepartment V AsiaRegional Office

FOR OFFICIALUSE ONLY Public Disclosure Authorized Public Disclosure Authorized

Docmentof theWorld Bank

Public Disclosure Authorized Thisdocument has a resticteddistribution and may be usedby recipients only in the performnanceof their official duties. Its contentsmay not otherwise bediscosed without WorldBank authorization. ABBREVIATIONSAND ACRONYMS

Symbols and abbreviationsused in appendix tables:

0.0 Not equal to zero - Zero ... or .. Data not available n.e.s. not elsewherespecified

ADB Asian DevelopmentBank ACIAR AustralianCentre for InternationalAgricultural Research AIDAB AustralianInternational Development Assistance Bureau ANU AustralianNational University c.i.f. cost, insuranceand freight DCF Discounted cash flow f.o.b. free on board FAO and AgricultureOrganization of the United Nations FDL Fruit DistributorsLtd () PHIA FundacionHondurena de InvestigacionAgricola FIMCO Friendly IslandsMarketing Co-operative FSRE Farming Systems Researchand Extension IMP InternationalMonetary Fund IRHO InternationalResearch Institutefor Oils and Oilseeds MAFF Ministry of ,Fisheries and Forestry 1UV ManufacturingUnit Value NCDS National Centre for DevelopmentStudies ODA Official DevelopmentAssistance OIC officer in charge TBES Tonga Banana Export Scheme TDB Tonga DevelopmentBank SPC South Pacific Commission DPIV 1Ionga'sFourth DevelopmentPlan 1981 - 1985 DPV Tonga's Fifth DevelopmentPlan 1986 - 1990

Note on metric and imperialmeasures.

Although officiallyTonga uses the metric system of weights and measures, a mixture of metric and imperialunits is used, even within the same definition (eg lb./ha or kg/acre). Each crop tends to have its own data conventions. In this Report, data is generallyleft in the units used in Tonga for that commodityto avoid confusionin Tonga and to minimize inevitableerrors in conversion. Metric and Imperialtons tend to be used interchangeablyin Tonga and although this Report uses metric tons much of the data is probably the imperial ton. However, this error of 2 percent is well within likely data errors in a countrywhere statisticsare not well developed.

CURRENCY

The Tongan is the Pa'anga (T$). The Pa'anga is linked to the Australiandollar at the rate of T$l = A$l. - - FOROMFFICIL USE ONLY

TONGA

AGRICULTURESECTOR STRATEGYREVIEW

Table of Contents

Pate No.

PREFACE ...... *. .** *. iii

EXECUTIVESUMMAIRY ...... lv ...... iv

I. AGRICULTUREIN TONGA...... 1I

Introduction ...... 1 TheEconoy ...... , ...... 3 Agriculture in the Economy...... 6

TI. PAST AGRICULTURAL .ERFORMANCE ...... 8

The Tongan Farming System ...... 8 Major Stple Crops .. 10 ...... 1 ...... 11 Introduced Commercial Crops ...... 14 Livestock ...... 17 Forestry .18 ...... *..i Fishing 19 Food Imporpt ...... 20

III. FACTORS AMCTING AGRICULTURAL DEVLOPMENT ...... 22

introduction ...... 22 SomeGeneral tongan eatures...... 22 FactorsAffecting Agriculture ...... 26 Key Constraints to TonganAgriculture ...... 33 Conclusions ...... 35

IV. PROSPECTS FORGROWT ...... 36

Introduction . . . 36 Household Consumption. 36 Marketed Domestic Production ...... 37 Niche Markets...... * . . . 37 Other Exports...... *.. 38 Overall Trend ...... 39

V. A STRATEGY FOR FUTURE DVELOPET OF AGRICULTURBE ...... 39

A. ELENTS OF A STRATEGY ...... 39

Introduction ... 39 BuildinguponStrength.. 40 Concentrating Effort...... 40 Defining the Role of the Public Sector ...... 41

This documenthas a restricddistribution and may be used by recipientsonly in the performance of their offlcialduties. Its contents may not otherwisebe disclosedwithout WorldBank authorization. -il -

Page No.

B. PROGRAM PRIORITIES ...... 42

Building upon the FarmingSystem ...... 42 ExploitingMarket Potential ...... 43 StrengtheningPublic Institutions ...... 45

C. IMPLICATIONSFOR DONORS ...... 50

APPENDIX AgricultureValue-Added and Exports

BIBLIOGRAPHY

MAP

TECHNICALANNEXES (Availableon Request)

1. Tongan SmallholderFarming Systems 2. Major Tongan Crops 3. The in Tonga 4. The Profitabilityof Tongan Agriculture 5. Forestry in Tonga 6. Tongan Society,Economy and Land Tenure 7. Labor, Migration,Agricultural Productivity and Mechanization 8. AgriculturalValue Added and Exports 9. The Post Harvest Phase of Agriculturein Tongat Marketing, Quarantineand Quality Control 10. Extensionand Research 11. AgriculturalPolicy and the Fifth DevelopmentPlan 12. AgriculturalProject Identification,Design and Appraisal 13. ProposedOrganization Structure for MAFF - iii -

TONGA

AGRICULTURESECTOR STRATEGY REVIEW

PREFACE

This is the World Bank's first review of the agriculturesector in Tonga. The agriculturesector is still the predominantsector in the economy, but the priority it warrants, for both Governmentand donor support,has been called into questionin recent years, as its performancehas seemed to falter.

The report is based on the findings of a desk review undertakenby consultantsin April 1989 and a field mission that visited Tonga in November, 1989.1/ Both stages of the review were carriedout with financial assistancefrom the Governmentof ,and the field missionwas conductedin collaborationwith the Asian DevelopmentBank, which helped guide the initialwork of the mission and was responsiblefor the review of the Forestry Sector.

The chief objectiveof the resultingreport is to suggesta strategy for the future developmentof the agricultureand forestry sectors.In doing so, it emphasizesthe need to build upon and sustainthe successful traditionalfarming systems that have evolved in Tonga over the centuries,and suggests that Tonga should try to turn to its advantagesome of its special characteristics,such as its smallnessand remoteness,which have hitherto been regarded as severehandicaps.

During the course of the review,a considerableamount of data was assembledand backgroundanalysts undertaken. For the sake of brevity, the technicalannexes preparedby the missionhave not been includedin the final report,but are availableon request.

The reportwas discussedin draft form with Governmentministries during a follow up mission in June, 1990, and has greatlybenefitted from their comments. The very considerabletime devoted to the review by the Government,and in particularby the Director and Staff of the Departmentof Agriculture,Forestry and Fisheries,is gratefullyacknowledged.

1/ The review was undertaken by a team led by John Burrows (Principal Economist,World Bank) and comprisingRobert Allen (ForestEcologist, Asia DevelopmentBank Consultant),George Hoengangongo(Rural Adviser, World Bank Consultant)and a team of consultantsfrom Hassall & Associates,provided with funding from AIDAB. The consultantteam consistedof Barry Shaw (Team Leader), Ralph Arnison (Agriculturalist), Hugh Coulter (MarketingSpecialist) and Brian Hardaker (Agricultural Economist). - iv -

TONGA

AGRICULTURESECTOR STRATEGY REVIEW EXECUTIVESUNMARY

Agriculture and the Econowy 1. Tonga is a small group of islands, covering a mere 670 sq.km. of land, set in a vast expanse (360,000 sq.km) of sea, and remote from major markets. It has a population of about 95,000, which enjoys an unusually high level of health and education. There are an estimated 40,000 Tongans who have migrated to New Zealand, Australia and USA, but still maintain strong social ties with the country and provide an important export market for traditional . Remittances from abroad amount to about four times exports. Together with a large inflow of concessionary aid, these remittances have allowed Tonga to finance a high level of imports, without significant overseas borrowing, and to maintain its currency at a level higher than it would otherwise be. 2. With a per capita income of some US$800, Tonga fits into the lower- middle income group of countries, and there is little poverty. Recent economic performance has been good by South Pacific standards, with an annual growth rate of about 3 percent. Agriculture is the dominant sector, accounting for around 40 percent of GDP according to official estimates. However, the report argues that official data underestimate agricultural production and exports, and that value-added in agriculture may be 302 higher than previously estimated. This implies not only that agriculture contributes much more to the economy than usually assumed--and therefore warrants more attention by Government--but that GDP itself is probably significantly understated. 3. The report examines the widespread view that agricultural production is stagnating or even in decline; a view that is reinforced by official figures and the decline in major exports. While data are very weak, the report concludes that the apparent stagnation in overall value of production in recent years has been wholly due to the decline of major export commodities, particularly coconuts, and that the production of traditional food products, both for domestic consumption and for export, has been growing vigorously, with only modest help from public sector agencies or external donors. Past Performance and Prospects 4. Tongan agriculture is based on a highly productive smallholder farming system, supported by good soils and an almost sub-tropical climate. Root crops such as taro, , , and kava are %rown under coconuts and mixed with cash crops, trees and livestock. Crops are serially rotated over about five to seven years. after which land is fallowed, usually under grass and bush. Pests and diseases are not a serious problem in this traditional system, but have affected introduced crops. Although fallows are getting shorter, farmers have so far succeeded in maintaining soil fertility. S. Most farmers not only meet their own basic household food requirements, but also supply domestic and export markets. Much of the recent growth in the sector has been in traditional foods, such as taro and kava, particularly for export, where producers are successfully exploiting an ethnic "niche' market among overseas Tongans. Farmers have been able to respond to -v-

new incentivesby adding new crops such as vanilla to the system,while maintainingthe productivityof traditionalcrops. Two importantcrops, pandanus and paper mulberry,provide raw materialsfor an expandingtrade in Tongan traditionalmats and ceremonialtapa cloth, and afford significant incomesto women.

6. Coconut products,the "traditional"export of Tonga, have declined dramatically,as a result of low world market prices, processingproblems, high wage rates, and the better returnsoffered by other crops, particularly root crops. While the Governmentwishes to rehabilitatethe industry,the report contends that copra and other coconutproducts will not be viable as exports,unless the return to farm labor can be greatly increased.On the other hand, in its traditionalrole as the 'Tree of Life' in Polynesian society,the coconutwill continue to be grown as an intrinsicpart of the multi-storeycropping system and used for a wide variety of food and other domesticuses. Thus, the report sees the need for furtherresearch designed to increasethe productivityof coconutswithin the existingfarm system.

7. Another crop in decline, banana,has suffered from poor qualityand disease control, and faces tough new quarantinerequirements and an imminent loss of trading preferencesin New Zealand.However, many producershave turned to the domesticmarkets and, with good managementand attentionto overall post-harvestmanagement, the export industry could recover.

8. Domestic pigs and poultry are importantand there are some commercialpiggeries. Other livestock,including cattle (slaughteredfor Socialoccasions) and horses (used as transport)also fit into the farming system. But, given the high cost of feedstuffs,commercial livestock is unlikely to be viable, unless it is given some protectionagainst low-cost imports.A more viable option is to develop livestockas part of the wider farm system.

9. Trees are an integralcomponent of the farmingsystem, providing shade and windbreaksfor crops and livestock,food (coconuts,breadfruit), timber, fuelwoodand materials for handicrafts.Agro-forestry warrants greater attentionthan it has so far received.Natural forest reservesare very small and should be conserved.There is a small plantatAonon the island of 'Eua, but it is unlikelyto reach a commercialsize unless more land is allocated.

FactorsAffecting Agricultural Development

10. The Constitutionguarantees every adult male 8.25 acres of farming land. With increasingpopulation, this has been impossibleto maintain, and there are fears that land will become a severe constraint.However, with many landholdersoverseas or in town, land seems to be readilyavailable through officialand informalleases, and the report does not see it as an immediate problem. But landholdersprefer short leases (up to ten years),which may discourageinvestment and accelerateland degradationand deforestation.

11. Tonga's traditionallow-input agriculture requires little credit, except for labor and other cash inputs,but the demand is growing-withnew commercialcrops and greatermechanization. Credit is generallyavailable from the banks, and agriculturalloan defaults are not a problem. Interest rates have been very low, but have recently risen. - vi -

12. Many crops can be successfullygrown in Tonga, despitetazards such as periodicdroughts and strongwinds. However,new crops and innovationsare not always carefullyscreened for their ability to fit into the farming system.AP a result they may be soon abandoned,if they are not financially attractive,or may call for practices,such as land clearing,monocropping or increaseduse of chemicals,that threatenthe long term sustainabilityof the system.

13. The report contends that the most seriousconstraints facing agricultureare macro economic and institutional,rather than technical. Recent analysis undertakenby the Treasury suggeststhat the real effective exchange rate has appreciatedand is at its highest level since 1975 against all major tradingpartners.. This has reduced profitabilityof exports and made it difficultfor Tongan farmers to competewith importsof cheap food. Second, Tongan wage rates are very high--overTS2.00 (US$1.50)per hour-- w-aich,together with the option of overseasemployment and the good returns offered by traditionalroot crops, result in a high opportunitycost of farm labor.Agricultural administration suffers from skill shortages,and from the common small country problem of having to spread its limited resourcesover many differentactivities.

Growth Prospectsand Proposed Strate&y

14. The report concludesthat value-addedmight grow by about 3 percent a year up to 1994, and proposes a strategyfor achievinga higher growth rate thereafter.It suggests that Tonga should build upon its strengths, particularlyits robust and efficienttraditional farming system, and try to turn some of its supposedhandicaps, such as smallnessand remoteness,to its advantage.However, it recognizesthat the domesticmarket is small, and that the combinedeffects of the high exchange rate and labor costs will make it hard for Tonga to compete in major commoditymarkets. The report therefore sees both the need and the potentialfor Tonga to developnew markets for specializedproducts that offer high returnsto labor. This will call for some concentrationof effort and a redefinitionof the role of the public sector.

15. The report suggests three priorityprograms. First, it recommends measures for building upon and conservingthe existing robust farming system, includinga greater focus of farmingsystems research.Second, it outlines a program for developingTonga's potentialas an exporter of semi-tropical produce, includingimprovements in quarantineandu quality controland market promotionand support for private exporters.Finally, the report discusses prioritiesfor institutionalstrengthening within the public sector,and suggestswhich functionscould be undertakenby farmereor private businesses.

Implicationsfor Donors

16. Tonga already receivesa high level of overseas assistance. However,many externally-aidedprojects have been poorly designed and implemented.The report thereforesees a need not for more projects,but rather for fewer, better projects,which might be broader in scope. Donors could help in this, by focusing on key priorities,improving the quality of project appraisalsand trainingprograms, and by offeringmore assistancefor strengtheningplanning and preparingthe priorityprograms outlined in this report. - 1 -

TONGAt AGRICULTURESECTOR STRATEGYREVIEW

I.

Introduction

1.1 The Kingdom of Tonga comprisesan archipelagoof 172 coral and volcanic islands (36 inhabited)with a total land area of about 670 kmffspread over 360,000klu of sea. There are three main groupssTongatapu and 'Eua, Vava'u and Ha'apai. (259 kmn) is the main island and includesthe capitalNuku'alofa.

1.2 Tonga's social and economic system has evolved over about 5,000 years of habitationand is a pervasiveand powerfulunderpinning of life and society. More recently,Western overlayshave interactedto produce a sociopoliticalsystem an understandingof which is crucial to any analysisof Tonga agriculture. Tonga has been a ConstitutionalMonarchy since 1875 and, althoughunder the protectionof Britain from 1900 to 1970, has never been a colony. With few dialecticaldifferences, the is a powerful unifying factor,while English, as an importantsecond language,facilitates contactwith the outsideworld. Christianityand the churches exert a powerful influenceand church buildingsaccount for most fixed investment.

1.3 The politicalsystem comprisesthe King and the Royal Family, 33 nobles and commoners. Governmentis shared in a complexmanner through the Ministers (appointedby the Crown), the Nobles and the LegislativeAssembly. Land is distributed(under the terms of the Constitution)to all adult male Tongans by the Crown through the Governmentand the Nobles.

1.4 The 1986 populationwas estimatedat 94,535. Populationcensuses were conductedin 1956, 1965, 1976 and 1986 (Table 1) and show considerable declines in populationgrowth, especially in the outer islands. There has been considerableinter-islaw'd migration to Tongatapuand migrationoverseas. There are an estimated40,000 Tongans living overseas (mainlyin New Zealand, Australia, and mainlandUnited States)but the number could be higher, dependingon the definitionof "Tongan". Lower birth rates have also contributedto the decliningrate of natural increase. There are some inconsistenciesin the data and demographicexperts have suggestedthat there is under-enumerationand that populationgrowth is higher. Urbanizationrates appear high, but are impossibleto comparewith other countries,since the entire main island of Tongatapucould be consideredas peri-urbanand closely linked to the capital Nuku'alofa.

1.5 The proportionof the populationunder 14 and over 60 years (normallyconsidered dependents) is almost 50 percent; very high in comparison zo countriesat similarlevels of development. This partly resultsfrom the high outmigrationof the 20-35 year age group. - 2 -

Table 1s Tonga Populationand populationdensity

Region 1956 1966 1976 1986 ------_ Population

Tongatapu 31,264 47,920 57,411 63,614 Vava'u 9,918 13,533 15,068 15,170 Ha'apai 12,477 10,591 10,792 8,979 Eua 1,925 3,391 4,486 4,393 Niuas 1,254 1,994 2,328 2,379 ------Total 56,838 77,429 90,085 94,535

Annual percentagechange during 10 year period (X)

Tongatapu ... 4.4 1.8 1.0 Vavau ... 3.2 1.1 0.1 Ha'apai ... -1.6 0.2 -1.8 Rua ... 5.8 2.8 -0.2 Niuas ... 4.7 1.6 0.2 ------Total ... 3.1 1.5 0.5 ------Area (kn?) Populationdensity (person/km2)

Tongatapu 260 120.0 184.0 220.4 244.2 Vava'u 119 83.2 113.5 126.4 127.3 Ba'apai 110 113.4 96.3 98.1 81.6 Hua 87 22.0 38.8 51.3 50.2 Niuas 72 17.5 27.8 32.5 33.2 ------…-_------Total 649 87.6 119.3 138.9 145.7 …------_

Source: Kingdom of Tonga StatisticalAbstract 1987

1.6 Educationalstandards in Tonga are high, with virtuallyuniversal primaryeducation and literacy. In 1987 there were 31,000 pupils enrolled in educationalinstitutions in Tonga and a significantnumber doing university studiesoverseas. But smallnesscreates a difficultyof maintainingadequate and relevantcurricula and less relevantones from other countriesare often used. This in turn makes high school graduatesmore useful and employable overseasthan in Tonga. Attempts at creating local relevanceare often frustratedby lack of skills and severe diseconomiesof scale in curriculum development.

1.7 Health care indicatorsin Tonga, as reflectedin a life expectancy of 65 years at birth and a very low infantmortality rate of about 26 per thousand live births, are generallyfar superiorto those in most countriesat a similar stage of development. Food suppliesper capita are more than adequate,with an average daily calorie supply of 2,940. In fact; the high incidenceof obesity is a cause of concern to the Tongan health authorities. -3-

The Economy

1.8 There are no reliablenational accounts statisticsin Tonga. Verious estimateshave been made by governmentdepartments, the IMF and other donors, ir.cludingADB and the World Bank. The latest estimates,prepared by the Bank in collaboratirnwith the Tongan Treasury,are set out in Table 2.

Table 2: Tonga - Gross DomesticProduct by SectoralOrigin 1981/82-1988/89 T$ million at 1981182prices

81182 83184 84/85 85/86 86187 87/88 88/89 ------__------_ Agriculture 24.3 22.2 24.2 25.0 25.9 23.3 23.8 Mining and quarrying 0.4 0.4 0.4 0.4 0.4 0.5 0.6 Manufacturing 4.7 5.0 5.3 6.1 6.3 6.3 6.6 Electricityand water 0.3 0.4 0.5 0.6 0.6 0.7 o.F 2.3 4.1 4.4 3.3 3.4 3.7 4.L Wholesaleand retail trade 5.7 5.7 6.3 6.1 6.9 6.9 6.9 Transport,storage & comes. 3.5 3.3 3.4 3.5 3.6 4.0 4.4 Financeand real estate 2.2 2.1 2.6 2.6 2.7 2.8 2.9 Comm.. social and pers.servs. 8.9 10.7 10.8 11.6 12.0 12.4 12.7 ------GDP at factor cost 52.3 54.0 57.9 59.8 61.8 60.6 62.8 Indirecttaxes less subsidies 8.4 8.9 8.5 8.7 9.0 8.7 8.7

GDP at market prices 60.7 62.9 66.4 68.5 70.8 69.3 71.5 GDP deflator -- 18.5 20.6 44.4 55.4 72.6 77.6 GDP at current prices 60.7 74.5 80.0 98.9 110.0 119.6 127.0 ------Agricultureas t of GDP 46.4 41.1 41.9 41.0 41.9 38.4 37.8 ------__------_ Sources Table 2.1, World Bank PacificIslands RegionalEconomic Report (forthcoming).

1.9 GDP has been growingat over 3 percent a year since 1981/82 (except in the drought year of 1987/88),which is high for the South Pacific region.l/ GNP per capita in Tonga is estimatedto be US$800, placing it in the lower-middleincome. group of countries. Taken togetherwith the generally high level of social services and the supportivesocial structure,:this income level suggests that the incidenceof poverty in Tonga is very low.

1.10 The economy is heavily dependenton agriculture,which accordingto these estimatesaccounts for around 40 percent of GDP in most years and nearly half of employment. However, the mission believes that the sector contributes a larger proportionof GDP than the official figures indicate (see below).

1/ For a fuller analysisof the economy, see the Bank's forthcomingRegional Economic Report, from which the above data are drawn. - 4 -

1.11 Manufacturingaccounts for only 9 percentof GDP, mostly processing of coconuts and other agriculturalproducts. If the estimatedGDP contributionof agriculturalprocessing is added to that from agriculture,the overall contributionof the sector rises to nearly 50 percent.

1.12 Foreign trade deficitsare considerable,with exports equivalentto 10 percent of GDP, and imports (mainlyconsumer goods) 60 percent. As a result, importedgoods account for 60 percent of the consumer price index. Nevertheless,the external account is usually in surplusas a resailtof receipts from , shipping,remittances from overseas Tongans (about 30S of GDP) and a large amount of overseasgrant oe concessionalaid (about 201 of GDP). Tonga has not had to resort to external borrowingand debt service paymentsare low.

1.13 The Tongan Pa'anga is linked (at par) to the Australiandollar, which floats freely in the internationalmarkets. This narrow link to one single currencyhas caused wide fluctuationsin both the rate of inflationand the real exchange rate. Furthermore,there are significantconcerns regarding the competitivenessof the economy due to the exchange rata policy (see paras. 3.44-3.45). Figure3 1 and 2 summarizethe balance of paymentsover recent years. Around 40,000 Tongans live overseasand now play a major role in the economy (and transactionsin the externalaccountj through their remittances and purchasesof Tongan goods, mainly food and ceremonialgoods such as mats.

1.14 The public sector includeseight ministries (the Ministry of Agriculture,Fisheries and Forests is responsiblefor agriculture)and several governmenttrading agencies,including the Tonga CommoditiesBoar4, which processesand exports coconutproducts and some other agricultural commodities. The public sector is also the major formal sector employer. Current expenditureis around 30 percent of GDP, and developmentexpenditure (whichis almost wholly financedthrough external aid) equivalentto 20 percent of GDP.

1.15 There is one commercialbank, the Bank of Tonga, part owned by the Government(40 percent) and three overseascommercial banks. The Tonga DevelopmentBank was establishedin 1977, mainly financed from external concessionalfunds and owned by the Government(90 percent) and the Bank of Tonga (10 percent). Both banks are significantlenders to agriculturefor production,marketing and exports. A Central Bank was establishedin 1989. Until recently,interest rates have been very low (5-6.5percent for time deposits),but have recentlyrisen.

1.16 The 1986 census indicateda total availablelabor force of 24,324 people, of whom 2,184 (9Z)were classifiedas unemployed. There is a perceptionin Tonga that unemploymentis a significantproblem but rural ' Go000 c

i - i~~ AI I I 1 II 33 3 a |

3U 0e SUE 01111'11^1 i i CA

ft Br,~~~~~~~~~~~~~~~~~~~c 6-

labor shortages,even at high wage rates, and the high level of household food productionsuggest that functionalunemployment (as opposed to a census definition)is low. Wages are very high by PacificIsland standardswith rural casual labor commandingT$W.CO - $2.50/hour.

1.17 Informationon Tonga from official sourcesin Tonga and overseas and from numerous research studies,reports and project documentsis voluminous and far greater than generallyconsidered available. For a very small country there is thereforea very broad and detailedinformation base but one which is dispersedand difficultto access. Unfortunatelyofficial statisticsare often not producedor are unreliable. The Statistics Departmentis very efficientin publishingstatistics but has little conttol over their collectionand accuracy. Data deficienciescan lead to inappropriatepolicy decisions. Although this report only discusses agriculturalstatistics, there is a great need to improvegeneral data reliability. The Mission believes,on the basis of its own experiencein collectingdata in Tonga for this report,that this need not be a difficultor costly process.

Agriculturein the Economy

1.18 Accordingto the estimatespresented in Table 2, agriculturalvalue- added has been virtuallyconstant since 1981/82,and has representeda steadilydeclining proportion of GDP. However,after discussionswith farmers,exporters, shipping agents, credit institutions,MAFF, the Statistics Departmentand officialsin severalgovernment Departments, the mission concludedthat the official figures seriouslyunderstate the contributionof agricultureto the economy and exports of Tonga.

1.19 The reasonsfor such under-reportingappear to be:

o the official figuresappear to understatethe value of home consumptionof food, other agriculturalproducts and forest products;

* statisticsare often collectedtogether with data which could be used for taxationpurposes creating a strong tendency for the seller/exporterto understateboth volumesand values;

e some of the most accuratedata comes from the CommoditiesBoard, which is mainly responsiblefor crops which have declined (coconut productsand bananas) so that the overall statisticstend to be unduly influencedby these declines;

o collectionof data on domesticmarketed productionis poor;

• there has been a very substantialcise in the volume of private sector exports, includingsales through relativesoverseas, which may not be always capturedby the official statistics;

* there is no regular data collectedon f.o.b. values of private exports, and thereforeno check on the standardprices used by the CustomsDepartment in its returns to the StatisticsDepartment. - 7 -

1.20 Since there was reason to suspectthat the role of agriculturehas been seriouslyundervalued, the missionmade its own estimatesof the value of agriculturalproduction and exports. These are set out in detail in the Appendix,21and summarizedbelowt

EstimatedValue-Added (T$ million, in 1988 prices)

1984 1988

Productionfor home consu.ption 30.6 31.5 Domestic sales 9.3 13.5 Exports 13.5 7.5 Total 53.4 52.5

1.21 The Mission estimatedthat agriculture'scontribution to GDP in 1988 was some T$52 million, or about T$30 million at 1981/82prices. This is some 252 higher than the figure given in Table 2. Given the mission's limitedtime and the scarcityof data, these estimatesof value-addedin agricultureare necessarilyrather crude. But if the mission estimatesare even of the correct order of magnitude,it is clear that agriculture'scontribution to the national economyof Tonga is much greater than previouslyrecognized. This suggeststhe need to rethinkpolicy in regardto asricultureand the levelof priority allocatedto it by the Government.

1.22 These estimatesconfirm that, in overall terms, the agriculture sector has stagnatedin recent years. However, this stagnationhas been whollydue to the decliningtrend in commercialexports, particularly coconut and banana exports,whereas the production,consumption and export of staple foods have all been rising, largelyoffsetting the decliningtrend in commercialexports.

1.23 There is also reason to believe that the period of stagnationis over. Not only has the decliningtrend in commercialexports now largelyrun its course, but there ts a broad consensuswithin Tonga that agricultural productionis again increasing. For example, farmersinterviewed by the mission consideredthat incomes from agricultureare generallyrising, and exporters,credit institutionsand shippingagencies all agreed that exports are increasing. Buyers and sellersat the producemarket in Nuku'alofa said that there have been significantincreases in volume sold in recent years, and pointed to the increasingarea taken up by produce on major market days and the rise of roadside sellingthroughout Tongatapu.

21 The mission'sestimates of value-addedin agriculturecover all crops for which there are reasonabledata. They includeestimates of household productionof crops, livestock(pigs, poultry and cattle), fish, fuelwood,building materials, kava, tapa, pandanus and mulberry;an estimate of local marketed production;and revisedestimates of agriculturalexports. For a detailed explanationof the methodologyused, see Appendix. Il. PAST AGRICULTURALPERFORMANCE

The Tongan Farming System

2.1 Tongan farmingsystems are complex and have proved to be robust and productive. They are fundamentallyagroforestry systems utilizing bush or grass fallow followed for severalyears by a series of crops intercroppedwith coconutsand other trees to create a rotational,multistoreyed fallow system. The productivityof the total system is higher than if any of these crops were monocropped. Each farm normally has parts of the holding at each stage in the cycle.

2.2 The traditionalstaple crops of taro, yams, plantains,coconuts and cassava dominate agricultural production, household consumption and exports, with fishingand some livestockas additionalsources of food. Tongans have evolved a highly productivecomplex farmingsystem which exploits the good soils and climatewithout much fertilizerand with very low labor inputs. This farmingsystem forms part of the wider social system, in which food is a major form of ceremonialexchange and achievesemotional and social importance which sometimesexceeds its nutritivefunction. Feastingand consumptionof traditionalfoods are an integralpart of any Tongan social occasion.

2.3 In contrast to cereal societiesbased on wheat or rice, the very high status of Tongan traditionalfoods means that demand increasesas incomes rise and lifestylesimprove. This has resultedin a remarkablystable demand for food for householdconsumption and the domesticmarket, where the traditionalfoods are in high demand and fetch very good prices. In addition, large numbers of Tongans overseasprovide a solid base for the export of Pacific food crops, Tonga's major export. Although there are significantfood imports (mainlyflour and meat), levels of food securityare high in comparisonto most other developingcountries.

2.4 A very importantfeature of the systemwhen compared to cereal systems is that seasonalvariations in labor requirementsare relativelysmall and timing of planting and harvest is not crucial. These factors enable a farm family to meet most labor needs from within the household.

2.5 Another featureof Tongan farming systemsis that new crops have been grafted into it very readily. Tongan farmers seem quite willing to try new crops and show degrees of risk taking uncommon in farmersat this stage of development. They have also demonstratedan aptitude for developing productiontechnologies suitable for each crop.

2.6 This Report does not use the common terms "subsistencesector', 'subsistencefarmer', 'subsistencecommodity", "non-monetary sector', 'comiercialsector', 'commercialfarmer' and many like them, since they are conceptuallyinadequate, misleading and of littlevalue for operational purposes. Writers on the Pacific Island Countriesfrequently refer to farmers as involved in a 'transitionfrom subsistenceto commercialfarming' as if this was some discrete,desirable and inevitableone-way process. In fact most farm families in Tonga: - 9 -

* produce a large part of the food they eat and the basic materials they use (e.g. firewood,thatching, mats, dyes) through crop cultivation,tree crops (e.g. breadfruit,coconuts), livestock production (pigs and poultry) and fishing;

* sell a proportionof crops for cash on either the domesticor export markets;

* sell a proportionof their produce directly to the public through local markets and roadside stalls;

* exchangeor present food for social, religious,family and ceremonialpurposes (includit, shipmentsto relativesoverseas);

• purchase food for consumption,either in the form of imported products (eg flour, meat) or local products,as a supplementto their own production;

o produce crops solely intendedfor sale (eg mulberry,kava, vanilla, copra, pumpkin,watermelons); and

3 receive cash income from wage employment(permanent or casual), either directly or through remittancesfrom abroad, or are members of farm co-operativelabor groups.

2.7 Thus there is virtuallyno "pure" subsistenceor commercialfarmer in Tonga except a very small minority in smaller outer islands,where marketingopportunities are few. Further, even the definitionof "farmer'is not clear. There are many full time wage earners in the non-farm sector, includingpublic servants,who own land and produce food for their own consumption,while some even have full-scalefarms or livestockenterprises.

2.8 Agriculturein Tonga primarilymeans smallholderagriculture, there being relativelyfew larger farms. But small-scaleTongan farmersare not Opeasants".31 The vast majority are operatinga farm within a decision frameworkwhich takes account of all the above activitiesin a highly sophisticatedway, using a technologicalmix which has evolved over centuries and which continuesto evolve to meet each new crop and challenge. This is not simply a romanticview based on what appears to be an attractivelife style to over-urbanizedWesterners, but based on a great deal of factual infornationand analysiswhich is presentedthroughout this Report and in the Annexes.

2.9 Anothermisconception is createdby the fact that Tongan farmers generallyuse very few cash inputs in relationto the value of their output.

3/ 7he word 'peasant'is not used in Tonga or the Pacific Islands and is a term of derision. Much misunderstandinghas been caused by a common belief that if one remainsa farmer in a low income country, one may be doomed to a lifetimeof 'peasanthood'and that only an industrialor service sector can provide any economic growth or improvementin living standards. This in turn can lead to a self-fulfillingdowngrading of agriculturalinvestment and policy decision-making. - 10 -

Theirs is a 'low input",but not 'low output" system. Major improvementsin output are possible,but these do not generallylie in highr inputs of fertilizer,pesticide, herbicide, credit or labor, but in improvedmanagement and better selectionof cultivars.

2.10 There is, of course, considerablevariation in farmers' performanceand management skills and in their valuationof leisureand time spent on other activities. There is thereforewide variationin farm income levels. Nevertheless,the importantfact remains that every Tongan with access to land (and that effectivelymeans every rural family)can, with moderate labor, produce sufficientnutritious food for family requirementsand social obligations.

Major Staple Crops

2.11 Tongan farmers are growing increasingquantities of major staple crops for household,domestic and export consumptionthrough continual adjustmentsto farmingsystems. Tongans and other Pacific Islandersoverseas provide a continuingand stable market for taro, yam, cassava and sweet potato. As noted elsewhere,Pacific Island countriesare unusual in that their stapleshave a positive income elasticityof demand so that as incomes rise, people purchasemore of these traditionalstaples. This is especially true of early yams which, as a very high status commodity,realize higher prices per kilogram the larger the yam. In Auckland a very large early yam may sell for as much as NZ$25.

2.12 In general, these crops are very successful. Centuriesof genetic selectionand evolutionof farmingsystems have combined to producehigh yields,while the country'sisolation has kept them largelyfree from pests and diseases. Sweet potato has had a disease problem,but it is hoped a new cultivar (now being released)will be disease resistant.

2.13 The increasein productionand sale of these crops has occurred with much less help from MAP" than has been provided for introducedcommercial crops, such as vanilla or pumpkin. Farmers have increasedproduction, created and managed their own marketing system and have demonstratedtheir willingness to work hard and to innovatewhen price incentivesare right. They also demonstratea higher level of entrepreneurialskill than Tongan farmers are normally creditedwith.

2.14 The most strikingaspect of the major staples is their high returns to labor; early yam, for example,gives a return of over T$10.00per labor hour, under averagemanagement, or more than four times the rural wage rate of T$2.00-2.50/hour.Figure 3 includesreturns per labor hour for major crops. It is noticeablethat introducedcrops such as cavendishbananas, vanilla, and pumpkins are lower ranking,while copra returnsonly about T$0.23/hour. - 11 -

Coconuts

2.15 Coconutshave been central to the Tongan economy for centuriesand more recentlyhave been the basis of its major agriculturalexport as copra, then coconut oil and desiccatedcoconut. Too frequentlythe coconut is seen as the producer of copra for export. But Tongans are estimatedto directly consume the equivalentof about 4,800 tons of copra per year, or about 230 nuts per person. The coconut is thereforeplanted and maintainednot just for copra, but as a multi-outputpalm which is planted and maintainedfor a variety of uses. It should be seen, first and foremost,as the "Tree of Life", which is universallygrown to provide a very wide range of important products apart from copra, includingfood and drink for home consumption,pig feed, thatch,mats, firewoodand timber.

2.16 Coconut export productsinclude copra, coconut oil, desiccated coconut, copra meal and (more recently)whole green and dry nuts. Figure 4 shows the trends in export values since 1974. The decliningtrend from about T$7.5 million of exports in 1974 to about $T3 million in 1987 is masked to some extent by the effect of hurricaneIsaac in 1982. The effect of price on growers' supply of copra is also evident from the very high correlationof the price to growers and their suppliesof copra. It is very clear that, given better income earning opportunities,Tongan farmersare not prepared to supply copra when prices are low. The Mission interviewedmany copra suppliersand the lowest price they would supplywas around T$350-400per ton.

2.17 Returns to labor from copra are about one tenth of the rural wage and the lowest of any recordedrural activity in Tonga. Farmers can get better returnsfrom almost any other income earning activity. For example, women interviewedby the Mission on Ha'apaiwere making considerablymore money for an hour's labor from producingmats from pandanus than the family could get from copra. Tongatapugrowers have stopped supplyingcopra at current prices, and deliveriesfrom other islandsare now very low.

2.18 Several replantingschemes have been implementedand many new palms have been planted,but this has not had any impact on suppliesof copra for export. Any plan to encouragereplanting or introducenew varietiesmust take account of the extremelypoor returnsto labor that the farmer earns from copra productionand the many alternativeuses of coconut, and fully assess farmers' incentivesto replant.4/

4/ Another way of looking at the incentiveto produce copra is to assess the additionalnet return to the farmer from processingcoconuts into copra, rather than consumingthem in the home or feedingthem to pigs. Assuming that the nuts have alreadybeen grown and collectedfrom the fields, only the costs of processingand transportingthe copra are taken into account. In this case, copra production,at 1988 prices, would provide a return of about T$l.3Olhourof labor (less,where transportcosts are high). Even this return (whilehigher than the discountedreturns from new plantings)would only be attractiveto farmerswhen they cannot earn the market rate for their labor. - 12 -

tice 3s RETURNSP LABOR lOOROf IAJOR CROPS

. -

So

i# 0 ~ g|

Sore ka -13-

Table 3: Annual Averaae Coora Prices to Suppliers ($1 per tofne) .______.._..__ee___...._____.f..__.._...--f------Year 1982 1963 1984 1985 1986 1987 1988

Annal average price at Commodity Board receival Depots: 227 323 510 371 220 244 298

Source: ComuoditiesBoard note: Monthly prices are includedin Annex 3

Figure 4: VALUE OF COCONUTAND COCONUTPRODUCTS EXPMRsv

TVOaPMR YEAR

we \g \ \W/ \ ~~~~~~~~~~~~~~~Whwolea

coonutZ

e ~~~~~~~~~~~~~~~~~~~Ap- =otameal

Source: Annex 3. - 14 -

2.19 The report concludesthat copra productionin Tonga will be uneconomicin the longer term, given levels of rural wages, the good returns from other crops and the costs of processingand exporting. The main reasons for this conclusionare:

* world prices for copra-basedproducts have been showing a decliningtrend and previous prices are unlikely to be repeated;

e the margin between coconutoil and copra is declining,forcing exportingcountries to become more efficientin their processing; yet Tonga is a small producerwith very high processingcosts;

* current low exportswill mean that Stabex transfersto Tonga will decline to virtuallyzero in the next few years and will depress producer prices even further;

e there are considerabletechnical problems with the desiccated coconut factory;

* there may be years of high prices resultingfrom supply problems in major exportingcountries, but it is uneconomicto maintain processingcapacity purely in the expectationthat this will occasionallyhappen;

* previous coconut replantingschemes have shown that farmers,even if they do replant coconutsunder the scheme,are unlikely to supply copra for export.

2.20 It is appreciatedthat the coconutmill is a large employer and that many outer islandsdepend on copra as their only source of income. While. there may be a case for maintainingcopra exports to assist the outer islands, there appears to be no case for continuedprocessing, especially once Stabex transfersend. Coconutswill continue to be grown, as a multi-purposecrop, and copra productionwill increasewhen prices are high. In the meantime, there is a growing, if small, export market for green and dry whole nuts and this should be encouraged.

2.21 Milled timber from coconuts could be a reasonablyviable industry. Despite a recent increase,the domesticprice of sawn coconut timber is still held well below the cost of importedtimber. A higher price could encourage the industryto make a greater contributionto import replacement. However, coconut sawmillingon Tongatapumay alreadybe near the maximum sustainable level, and the future potential supply needs to be assessed. Moreover, coconut timber has differentproperties from many of the importedtimbers and cannot replace them for all uses.

IntroducedCommercial Products

2.22 Introduction. The performanceof introducedcommercial products such as cavendishbananas, watermelon, pumpkin, other fruits and , livestockand plantationforestry is very mixed. For the domesticmarket, various introducedcrops such as tomatoes,carrots, and Irish potatoes are grown, but their productionand sales are not as great as the major staples. Some are marketed to hotels and supermarketsfor consumptionby foreigners. - 15 -

2.23 Vanilla. Vanilla has been successfullyintroduced on Vava'u. There are said to be about 2,000 growers with an average of 1/2 ha each. But this may be the number who have tried vanilla; many have dropped out of production, leaving a core of experiencedgrowers and a reasonably sound industry.

2.24 MAFF input has now slowed and the industry looks as if it will maintain its own momentum. Nevertheless,it is importantto ensure that quality of green bean and curing is maintained. Exports are about T$1.0 m to T$1.6 m per year, but unofficial unrecordedexports probably take the total to 10-20 percent more than this. The crop requires hand pollinationand this has become an important, and culturally acceptable, farm activity for women.

2.25 Returns to labor are around T$5.00/hourunder good management but fall to less than T$1.00 for growers with poor management. Such growers should not be encouraged to remain in the industry. There are serious disease concerns,particularly recently identifiedvirus infections,which will require the production of disease-freeplanting materials and improved plantation hygiene on the part of farmers, supported by the extension service.

2.26 Export bananas. In the 1960s, Tonga exported 10,000-20,000tons of cavendish bananas; since then exporto have fallen to 3,000 - 4,000 tons or even less. Black leaf streak disease virtually destroyed the industry and New Zealand, the major market, bought elsewhere. In 1989, exports were temporarilyhalted because of fruit fly found in a consignmentof Tongan plantains and exports will be negligible.

2.27 There has been a restructuringof the industry,with registrationof growers demonstratingbetter management and disease control. But unfortunatelythe improvementof disease control and consistent quality is unlikely to occur before the New Zealand banana market is deregulated in 1990. (Tonga previously enjoyed preferentialtreatment).

2.28 Tongan bananas are smaller and sweeter than their competitors from Ecuador and ,but quality is not consistent and shelf life, even from one carton, varies. As a result they are discounted at the wholesale level in New Zealand, the only market. Getting the plantation and post-harvestmanagement of bananas right could regain a good export industry. There are now growers with the skills to manage adequately,but the organizationof the industry and post-harvestmanagement is not yet developed to the point where the industry is viable.

2.29 Revitalizationof the banana industry is possible. But it is mainly a managementnot an agriculturalproblem. The technology exists in Tonga to put the industry back on its feet.

2.30 Pumpkin. Since 1987, Tonga has been exploiting the November- December period in the Japanese pumpkin market when prices are high. This began through the initiativeof a private firm in New Zealand, and in 1989 was organizedby the Tonga Growers' Associationwith a great deal of MAPF help and government financialassistance. Exports have increased rapidly,and in 1989 amounted to over 3,000 tons. In 1990, a second associationhas been formed, and growers are planning to produce 5-6,000 tons for export.

2.31 While outwardly the industry is successful and has attracted a great deal of grower support and attention, in fact there are problems. Despite a good growing season in 1989, the real net return from the whole crop - 16 -

was negligible.and without the TS500,000susbsidy injectedby Government, many growerswould not have covered their costs of production. Moreover,the cultivationof pumpkin has, for some farmers,created conflictswith the traditionalfarming system. Stands of coconutshave been felled to eliminate shade, and some areas have been monocroppedwith pumpkin for severalyears, with as yet unknown effects on soil quality.

2.32 The recent experiencewith pumpkins illustratesboth the opportunitiesand potentialhazards of such niche markets. It has demonstratedthat Tonga can penetrateeven such a sophisticatedmarket as 's, and that Tonga farmers can respondvery quickly to market opportunities. On the other hand, it also shows the importanceof quality control, and the difficultieslikely to be faced by growers in opening up and sustainingsuch markets in the future.

2.33 Watermelons. Watermelonshave been one of the most profitable crops in Tonga, giving estimatedreturns to labor of $8.55lhour.They have a high value on the domesticmarket and have been exported to New Zealand for considerablereturns. Unfortunately,fruit fly larvaewere found in one private consignment,and all fruit importsfrom Tonga (includingbananas) were banned. Recently,this ban has been lifted.

2.34 With good managementand a level of consistentquality control (to ensure these problems are not repeated),the market could be regained. As with bananas, the problem is not one of agriculturalpotential but of management.

2.35 PassionFruit. The current passion fruit projectwas begun privatelyin 1982, then encouragedby MAFF. Growers were encouragedto grow the crop two years before the factorywas finally completed,which created disillusionmentbut fortunatelythe crop sells well on the local market. The factoryhas been processingabout 60 - 70 tons (127 tons in 1987t8) but was designed to process 800 tons. There were 77 registeredgrowers in 1989.

2.36 There are plans to restructurethe processingplant and diversify into other juice processing (tomato,pineapple and pawpaw) and this may prove worthwhile. It has had significantamounts of overseasassistance which was neither timely nor well planned.

2.37 Other Fruits and Vegetables. There is a very wide variety of both local and introducedfresh fruit and vegetablesavailable on the Tonga domesticmarket, which provide a moderate income source for growers. These include such items as peanuts, onions, tomatoes,pineapples, watermelons and fresh coconuts. The value of 'localsales has probably been considerably underestimatedin the past. Touristhotels, restaurantsand supermarketsbuy local exotic vegetables. This market will continuebut will be unlikely to grow quicklyunless Tongan tastes change more rapidly or tourism increases. - 17 -

Livestock

2.38 There are over 80,000 village pigs, about 200,000village poultry, 10,000 horses, 14,000 goats and more than 9,000 cattle.51 The Vaini Research Stationhas pigs, poultry, goats, and cattle as a source of breeding materlal for exchange,crossing and distributionto local farmersto improve livestockquality.

2.39 Pigs play an importantrole in traditionalceremoniee for gifts and exchange at feasts for weddings and funerals. Every rural household raises pigs in a semi-openrange system around the villagewith supplementary feedingof crop residues,scraps and such food as coconut kernel. All villages have an abundantnumber of scavengingfowls, chiefly used for meat.

2.40 Cattle have become an importantpart of this system,mainly tetheredon garden fallow land, and have entered the social system as most are butcheredfor feasts. Horses are very common in villages and are an important beast of burden. The use of the horse and cart is now in decline due to the increasinguse of mechanization. Goats are shepherdedin villageherds or tethered in small lots. Village livestockproduction is mainly for food and for social occasions.There is an active local demand for live animalsfor social functions.

2.41 It is importantto see livestockwithin the Tongan social and economic system. Livestockare the principalform of householdand farm sr.vingand investmentfor the majority of households. A recent survey found them to average 75? of the value of rural householdassets includ!ngtransport equipment,fishing equipmentand agriculturalequipment. In the poorer outer islands,the proportionincreases to 89X.

2.42 Intensivelivestock raising has had a limited success. Tonga is largelyself sufficientin eggs, and until recentlywas self sufficientIn chickenmeat (broiler)production. Broiler productionhowever, has now completelycollapsed due to competitionfrom imports,principally chicken pieces from the USA, and pig productionhas been only partly successful. There are six dairy herdst one on Vava'u with a total of 250 milkers and 270 dry cows, and five on Tongatapu. Given the small scale of operationand lack of locally-producedfeedstuffs, it will always be difficultfor intensive livestockproduction in Tonga to competewith imports,particularly the cheaper products.

2.43 Livestock,especially pigs and cattle, are more often a social investmentthan a commercialenterprise. There are a few private large cattle enterprises,but even these are usuallyunwilling to sell cattle since they add so much to their owner'sprestige. From the smallholder'sviewpoint, cattle have been importantas a source of meat for ceremonialoccasions, an enhancer of prestigeand as a "bank" for emergencies.

2.44 Severalattempts have been made to encouragecommercial cattle; but these are unlikely to succeedwithout an assured supply of fatteningstock (see below). There is little land for cattle and they are normally tethered

5/ 1985 AgriculturalCensus. - 18 -

on fallow. This system enhances fertility,but the stockmust be hand watered which, with high labor costs, raises productioncosts.

2.45 Improvementof cattle and cattle numberswill depend on the availabilityof livestock. Livestockdistribution from Vaini Research Station was only 44 weaners in 1987 and 35 in 1988. To build up a centralherd to provide for adequate numbers of cattle for distributionwill be very difficult as no large tracts of land are available. Small nucleus herds would also be difficultto establishwhen compared to alternative,higher valued, land use.

2.46 Improvingcattle productivityin Tongan farm systemswill need to take accountof socio-cultural,technical and logisticaspects of cattle productionand will relate particularlyto nutritionand the place of cattle and pastures in garden fallow. Cattle numbers tend to rise and fall with rainfalland feed availRbility. Cattle will thereforeremain a significant, but not rapidlygrowing, part of the farming system with some commercial sales. There is a mobile buyer/butcherwho markets meat in the town after carcass inspectionby MAFF.

Forestry

2.47 Tonga has a very limited forest cover. Olantationforestry is largely in a developmentalphase and the establishmentof a timber resourceis of high priority. Wood will be the principalcrop from the single plantation forest (currently389 ha) on the island of 'Eua. The remaining4,000 ha of natural hardwood forest also containsa considerablewood resource.

2.48 A stated reason for developingthe forest plantationon 'Eua is to ultimatelyreduce Tonga's dependencyon sawn timber imports. Pinus Caribaea has been chosen as the only satisfactoryspecies for the exposed, relatively infertilesites. This pine will produce a medium strengthtimber, primarily to provide framing timber for building. On she'tered,fertile sites Toona australisand Cedrela odorata show far superiorgrowth to any other species. The major use for these two hardwood speciesshould be the productionof wide clear board suitable for interiorpanelling, exterior weatherboard and furniturefor the domesticmarket. The internal rate of return for planting the poorer sites in Pinus Caribaea is probably low (2.01 to S.AX). The variationin internalrates of return reflects the impact of different managementcosts: these costs must be contained. Forestry should also be allowed to develop as a land use on some of the more productiveland.

2.49 It is difficultto convince politiciansin most Pacific Island countriesof the long-termbenefits from forestry. The availabilityof land for plantationforestry on 'Eua is a critical issue. If the plantationis to reach a commercialsize, land will need to be allocatedoutside the area already zoned for the plantation. There is also some risk in investingthe total plantationeffort in one area. For example, cyclonic storms have caused considerableplantation damage. It is recommendedthat species trials be undertakenelsewhere as a basis for alternativeplantation development. This would require areas to be zoned specificallyfor forestryuse.

2.50 Logging of natural hardwood forestshas been exploitativeand there are no managed stands of such forests for either timber or fuelvood production. The extent of unmodifiedexamples of natural forests continueto - 1. -

declinebecause of agriculturalencroachment. The few natural forest remnants on Tonga contain some nationallyand internationallysignificant conservation areas, and the Tongan governmentneeds to act with urgency to protect such forests as a national heritage.

2.51 The potentialfor developingagro forestry is considerable, bearing in mind that nearly every farmer cultivatesunder the partial shade of trees. Apart from the value of trees as shade or windbreaks,there are at least five areas that warrant greater attentionin agro-forestry:the various uses of the coconut tree, the productionof fruits and nuts for both home consumptionand sale, the potentialfor high-valuetimber specieson smallholdings,the supply of pandanus,mulberry and other raw materials for handicrafts,and the productionof firewoodor charcoal for domestic consumptionand the increasinglyimportant urban market.

2.52 There has so far been no systematicprogram to develop agro forestryand little funding. It is importantthat a coherent strategy should be developedfor agro forestryas part of the developmentof the farming system. Tonga is well placed to do this, since forestry is well integrated into MAFF and forestryextension officers form part of the extensionteam at districtlevel. More work will be required,however, to determinebetter ways of integratingttees into the farming systems.

Fishing

2.53 The fisheriessector was not part of the terms of referenceof this study. In many ways, this is an unfortunateomission, since coastal. fishingis an integralpart of the householdfood productionsystem and fishingcompetes with farm activitiesfor householdand hired labor. Although there has been considerableassistance to offshorefishing, there has been relativelylittle study of coastal fishingand the collectionof molluscs and other seafoods. It appears there is a diminutionof inshoreresources on more populatedislands and that over-fishingand collectionis a problem.

2.54 Fish is a very importantpart of the Tongan diet. For the more distant islands,with fewer economic opportunities,fishing is an essential food source and forms a large part of incomes. Given the greater coastal resourcesper person of these islands,it is possiblethat the prospectsfor fishing for markets in Vava'u and Tongatapuare greater than for marketed agriculturalcrops or livestock. The omissionof fishingtherefore also createsa regionalimbalance in this reportwhich is recognizedby the mission.

2.55 Offshore fishing is a very importantand successfulactivity and has receivedsubstantial overseas assistance. Incomesearned by boat owners and workers are high by Tongan standardsand commercialfisheries produce around T$J million yearly with substantiallymore from artisanalfisheries. There are significantexports, but most is consumed locally. Aquacultureis a -ecentbut growing sub sector,which is receivingincreasing attention from MAYF. - 20 -

Food Imports

2.56 The level of food imports has been a concern in Tonga for many years and is currentlyabout T$18 million, amountingannually to about T$200 per head and about T$l,000per family. Food importsnow account for over 60? of the value of total imports,compared with about 352 in the early 1980s. Major food importsare set out in Table 41 meat accounts for 392 of food imports,cereal products262, dairy products 1lZ and sugar and confectionery 6 percent.

2.57 Much of the importedmeat are cuts for which there is little demand in producingcountries, such as brisket,mutton flap, chickenpieces (from layer hens) and canned meat. The average landed price of the popular beef and mutton cuts is about TS1.25/kg (about 95 US cents) and poultry pieces are T$1.40. No local productioncan competewith these prices, especially given the need to import feed at a landed cost of about TSO.50/kg.

2.58 There is no likelihoodof these commodities(and dairy products) being produced in Tonga at importedprices. It is an indicationof the strengthof the Tongan demand for traditionalfood crops that those crops command higher domesticprices than major importedfoods.

2.59 Flour is also a major importwith about 5,000 tons importedyearly (landedcosts T$0.44/kg),or about 55 kg/year/person,or a little over lkg/week. Family consumptionaverages 0.75 kg/day, and village surveyshave recordedconsistent consumption by some adults at 1 kglday. Flour is consumed as bread, pancake and various cooked batters and cakes. (It is linked with the import of almost T$0.5 million of baking powder and yeast annually).

2.60 These high levels of consumptionof flour, sugar and meats are cause for concern,especially since obesity is a major health problem in Tonga.6/ Most other Pacific Islands'food importsare around 20-302 of total imports. This high and rising level of food importsand the growth in demand for importedfoods reflect strong domestic denmandcombined with low or no tariffson foodstuffs. In part domestic demand is strong because of the 'Dutch diseasea effectsof large aid and remittanceinflows, which drive up wages and autonomouslyaugment purchasingpower. Vhether there should be tariffson low cost food importshas welfare implications,since these foods are eaten more by low income earners.7/ This is a policy issue on which MAFF needs to have a clear position,however (see para. 5.25).

6/ Tonga NutritionSurvey.

7/ The various facets of food policy in Pacific Island food policy is discussedin B. D. Shaw 'Food and NutritionPolicies for South Pacific Countries:Determinants of GovernmentPlanning" in R. R. Thaman and W. C. Clarke (eds) Food and National Developmentin the South Pacific, Universityof the South Pacific, Suva, 1983. - 21 -

Table 4a Food Imports 1988 (T$)

Beef Beef brisket 442 314,868 Mutton Mutton flap 892 2,999,131 Pork 13,553 Poultry Pieces 702, Whole chicken 30X 1,093,056 Other meat Sausages, bacon, ham etc. 619,304 Corned meat Tinned corned beef 1002 2,061,768

TOTAL MEAT 7,101,680 Dairy products Butter 292, Milk powder 222, Ice cream 191 Canned and WUT milk 142 2,002,592

Fish Tinned fish 71? Dried/smoked/salted 172 320,066

Wheat & flour Flour 892 2,468,004 Cereal products Biscuits 662, 2,109,668 Rice 159,276

TOTAL CEREAL AND PRODUCTS 4,736,948 Fruit and juice 500,014 Vegetables 339,522 TOTAL FRUIT, JUICE & VEGETABLES 839,536 Sugar & confectionery Sugar 642, Confectionery 332 Chocolate 242 1,082,202 Beverages & spices Cocoa 46X Coffee 15t 512,380 Animal feed 408,559 Margarine, fats & oils Marg/dripping 692

Vegetable oils 202 578,788 Other food Baking powder 622 449,849 TOTAL FOOD IMORTS 18,163,577

Animal feed 408,559 ------__------Sources Annual Foreign Trade Report 1988, Statistics Department, Nuku'alofa

Notet This data does not correspond to SITC classifications as used in Tonga's official statistics. It includes fats and oils and animal feeds which are normally omitted. - 22 -

2.61 This also has importantimplications for hurricaneresponse by donor countries. Large quantitiesof flour were sent to Tonga by Australia and New Zealandwhen the local farmingsystem could probably have coped with traditionalfoods (see also later sectionon hazards).

2.62 In spite of the high level of food imports,food security is high. The case of hazards is discussedbelow, where food security is greater than normally thought. The ultimate test would be that should food importsbecome unavailablefor any reason,Tonga could feed itself from availableresources and technologywith some diversionfrom exports.

III. FACTORSAFFECTING AGRICULTURAL DEVELOPMENT

Introduction

3.1 Tonga is a small, remote,archipelagic country with limited land-basednatural resources,distant from world markets and with high rates of outmigration. These would appear to be insurmountableproblems which combine to make the future of agricultureas a growth sector very bleak. In the early 19808 there were many conferences,books and studiesof small and/or island states. These focusedon these attributesas "problems'but generally did not see possible advantagesthat such countriesmight have.8/

3.2 For example, Tonga's remotenessis responsiblefor its relatively disease free status, its strong and cohesive social traditionsand language. Its smallnessfacilitates internal communication and the provisionof a high level of services to the population. Its high levels of migrationhave createdoverseas markets for Tongan produce, a remittanceflow that has enabled increasedstandards of living, and a two-way flow of populationthat has reducedisolation, generated excellent overseas shippingand air links, and made the averageTongan today part of a 'globalvillage'. This chapter discussesthe range of factorsthat affect Tonga'sagricultural performance and prospects,attempts to dispel some of the myths that have been propagated in the past, and finally draws attentionto some of the more serious constraintsfacing Tongan agriculture.

Some General Tongan Features

3.3 Tongan Culture, Social System and Economy. The Tongan culture and social system has profound implicationsfor agriculture. Official documents refer, from time to time, to the problems of developingcommercial agriculture given the features of the Tonga social system. Yet the Mission was impressed by the entrepreneurialactivities of farmersand traders and the way in which groups and individualshad often devised unusual solutionsto problems. This is evident in the methods of exportingcrops where the farmer often flies to the destinationto meet the containerand then markets it.

8/ These aspects of small island states are discussedin greater detail in B. D. Shaw "Smallness,Islandness, Remoteness and Resources:an Analytical Framework"in Regional DevelopmentDialogue, Special Issue, United Nations Centre for Regional Development,Nagoya, Japan. - 23 -

3.4 Several aspectsof the social and economic system are particularly relevant to

* '.andis individuallyowned and inherited,giving an incentiveto individualperformance (unlike most of the Pacific Island countries, where land is communallyowned), and entrepreneurshipis encouraged by a combinationof friendlycompetition and peer support;

e group labor activity is common for men and women and is used as a means of generatinghigher output per labor unit;

• Tongan food productionsystems require virtually no post-harvest storage or marketingtechnology; it is thereforemore difficultfor Tongans to conceptualizethe need to improvepost-harvest technology;

* the productionsystem enables the farmersand other rural people to provide the time and resources(often at short notice) to maintain the very importantsocial structure;

o high levels of food and personal securitycreated by the economic and social system leave little need for saving and investment;while this impedes investmentin agriculture,the system supports agriculturethrough local and overseasTongans' high levels of expenditureon traditionalfoods;

e the strong social links combinedwith migrationmean that many Tongans have visitedother countriesand many overseas Tongans return home to visit families;thus there is an unusually knowledgeablepopulation with a very good understandingof life and economies overseas;

e the recent rapid movement of farmers into the domestic and internationalmonetary economy,however, has not been matched by an understandingof farming as a business; too often the "Tonganway, is blamed for the failureof projectswhich was really the result of poor financialplanning;

e to a large extent, the Governmenthas an accepted role as protector, mentor and guardian of knowledge. Extensionworkers and other MAFF staff also assume some of this mantle. This places great social responsibilityon MAFP and its workers.

3.5 To sum up, there is considerablesocial cohesion in Tonga with one common language (plus a wide use of English),no dominantmigrant group and a common social system which is well understoodby the Tongans themselves, though not always by foreigners. This social system confers considerable advantageson Tonga in terms of communicationand agriculturalextension. It also results in today's exchange of food crops exports and remittancesbetween residentTongans and their relativesoverseas.

3.6 Tonga's GeographicalRemoteness and Small Size. Tonga consistsof many islands spread over an enormousexpanse of sea. Externally,this means that Tonga is very remote from world markets. However, shipping and air - 24 -

serviceshave improvedconsiderably in recent years. Containersof fresh food crops are marketed to New Zealand,Australia and elsewhereat a cost and regularitynot thoughtpossible severalyears ago and Tongansare taking great advantageof this.

3.7 Internally,the dispersedgeography creates problems of communication,transport and services. Inter-islandtransport is commonly inefficientand poorly coordinatedand cargo may be transhippedmany times. Communicationsand transportbetween Nuku'alofaand Auckland are better than to some islands. Fortunately,living standardsin many of these islandscan be maintained at reasonablelevels (thoughnot as high as they would prefer) through the sea and land resourceson which they have survivedfor centuries. They will not starve,but they want more income.

3.8 Tonga is trying to develop a regionaldevelopment strategy to overcome these problems. Yet planningunits in all Tongan ministriesare alreadyoverstretched. Some of the approachesin this report might be used to developappropriate strategies for these remoter islands.

3.9 Tonga is a very small country in terms of populationand administrativecapacity. This creates its own problems. Small states are characterizedby a very large share of output from a few sectors. The compositionof commodityexports is correspondinglynarrow while importscover a very broad range of products. Foreign trade per capita is high, the economiesare open and highly dependent;shocks from changes in international prices of importsor exports are felt quickly and are pervasive. This rende..s appropriatemacroeconomic management more necessary,including export price stabilizationpolicies.

3.10 The range of skills required in governmentat all politicaland administrativelevels are the same in small states as large ones. But they are spread over fewer people who generallyare less well trained. Job mobility is high and one person may have several responsibilitieswhich, in a larger country,would be done by an entire public servicedivision . Travelling abroad and dealingwith foreignofficial visitors takes an inordinateamount of time. Skills specializationtherefore becomes difficultto justify or achieve. The ratio of governmentsalaries to other cash economic activity is high and governmentemployment is the major focus for well educateduniversity or school leavers.

3.11 Smallnesscombined with remotenessencourages a narrow and underdevelopedprivate sector. In agriculture,this implies few market intermediaries,diseconomies of scale in processingand marketing,and a strong tendency toward parastatalmarketing and processingorganizations. The largestcommercial-industrial organization in Tonga is the Tonga Commodities Board. Domestic market smallnessalso limits industrialdevelopment and makes it highly unlikely agriculturalinputs will be locallymanufactured. Imported agriculturalinputs fluctuatemarkedly in price with the open, dependenteconomy (whileoutput prices fluctuateand often in opposite directions),so farmers tend to use fertilizerand pesticidesonly on high return crops. These limits to industrialdevelopment are crucial for agricultureby making it the only productionsector likely to provide employmentand incomes in the absence of large amounts of other resourcesor service industries. - 25 -

3.12 On the other hand, smallnessand remotenessconfer certain advantageson Tonga. Some of the more intractableproblems of low-income countriesmay be easier to resolve in small states. For example,the effects of policy changes can be monitored and correctedby using an 'economicand social laboratory'approach; the economy is so small that unforeseeneffects can be identifiedquickly. In agriculture,there are a number of advantages that should be exploited:

Communicationwith farmers. Smallnessin terms of numbers of farmers makes it easy to communicateor extend new technologies. The extension service is already in personal contactwith many farmers, and the compact size of MAFF shouldmake it feasibleto apply a coordinatedapproach to the developmentof farmingsystems, as suggestedlater.

Flexibilityof Response. Tongan farmershave demonstratedthe flexibilityof their farming systemsto changingcircumstances e.g. the pumpkin industry,export of food crops in containers,the passionfruit industry. Response time is very short so farmerscan take advantageof opportunitiesas they arise.

Quarantine/QualityControl Capability. Becauseof its smallness, insularityand isolation,Tonga has an opportunityto impose tight control over quality and quarantine,both for export of agricultural produce and the restrictionof pests and diseases into the country. Tonga has in the past been fortunatein being relativelyfree from major pests and diseases of plants and diseases,but severalincidents in recent years have demonstratedthe need for ever increasingvigilance in inspection.

Brand Promotionor ProductDifferentiation. A key element in the developmentof market niches is product differentiation.This can be effectedby emphasizingthe differenceswith competingproducts (e.g. differencesbetween the variety of staplesare importantto the different Pacific Island communitiesin overseasmarkets), or by developinga brand name with which consumerscan identify. For example,Tongan bananashave been sold under severalbrand names, such as 'Blue Palm' and 'Ifo', to promote their faced stiff competitionfrom Ecuador and Philippinesin the New Zealandmarket.

3.13 External Aid and its Management. Aid funds to small remote countriessuch as Tonga tend to be very high per capita and tend to support currenciesat levelswhich their productionand trade could not otherwise sustain. This createsproblems for Tongan agriculturein competingwith importedfoods and getting a good return for their exports. High levels of aid in relationto experiencedTongan staff creates problems for MAFP and other Ministries.

3.14 Since aid levels per capita are very high, capital is not a constraint. But considerabletime is taken in seeking,negotiating, spending and reportingon externalassistance. "Aid fatigue'is common. Thus there is too much emphasison projects and too little on broad planningand economic management. Projects tend to be reactive to donor suggestionsrather than proactiveand carefullyrelated to countryneeds. Adaptive approachesare frequentlyrequired, but their importanceis not always recognizedby donors, - 26 -

who do not wish to spend the considerabletime involved in appraisingsmall projects. Smallne8screates high costs for donors of offeringand designing projects.

3.15 Small remote countriesfind themselvesbalancing on a fragileedge of wanting greater independenceand income yet not finding it within the establishedpatterns of developmentof other countries. As a result they become susceptibleto those sellingquestionable benefits. Tonga has reviewed more than one proposal for the disposal of internationaltoxic wastes. Ministers and politicians,eager for solutions,are ready to listen to those with plausible suggestions. This is quite understandable,given the apparent lack of other possibilities.

FactorsAffecting Agriculture

3.16 Although the land area is small, soils are very suitable to the market garden-agroforestrysystem which Tonga has developed. Rainfall is lower and more variable than Pacific Island countriesfurther north. Sea and coastal resourcesare enormous and are being exploitedthrough traditional fishingand collectionas well as offshoremethods. There are no mountains and very little natural forest cover, nor are there known economicmineral resources. Seabed resources,such as minerals,are unknown, but there is a probabilitythat in the distant future, seabed mineral resourceexploitation may be feasible. Manganesenodules are believed to be on the seabed.

3.17 Tonga's climate is basically sub-tropicalwith mean daily summer temperaturesin the high 20s and daily winter means in the low 20s. Mean annual precipitationis between 2200 mm in the north and 1700 mm in the south, with more than half fallingbetween January and April. Cyclonic storms and drought create problems;cyclones of hurricaneforce have been recordedsix times this century,most recentlyin February 1990. While the hurricanesand fierce storms which cause property and other damage are best known to donors and other countries, droughts are probably a greater economic hazard. There is virtuallyno surfacewater except specific impoundments,although undergroundaquifers exist on major islands.

3.18 Soils. The islandsof centralTonga, includingTongatapu, most of the Ha'apai group, and parts of the Vava'u group are low coral islandswith a covering of volcanicash. A few islands in the Ha'apai group and most of the larger islands in the Vava'u group are of raised coral formations. 'Eua has a ridge of hard limestoneabout 300 m high, flanked by limestonealong the western side. On the steep eastern slopes volcanic rocks are exposed. Volcanic ash over coral leads to fertilesoils. Soils on the exposed limestoneof 'Eua are comparativelyinfertile.

3.19 The most widespreadand importantsoils in Tonga are dark brown to red volcanic soils derived from fine-grainedandesitic ash depositedover coral rock. These soils have excellentphysical properties, being well structured,friable in the upper horizons and well drained. They are very versatilesoils and suited to a wide range of crops. The remainingsoils (about 10 percent)are derived from coral rock and are coarse textured,poorly structured,frequently shallow and generallynot suited to any crops other than coconuts. These soils are usually found close to the sea and on the - 27 -

north and west sides of the larger and older islands,they are proportionally predominanton the smalleror younger islands.

3.20 The chemical status of Tongan soils can vary considerablywithin each type with, generallyspeaking, slight deficienciesof nitrogen, phosphorus,sulphur and occasionallypotassium occurring. The variability relates to farming intensityand practices--particularlythe frequencyof burning and mechanicalcultivation which removesorganic matter from the soil and its content of nitrogen and sulphur is lost to the atmosphere.

3.21 Land. There has been considerablediscussion in Tonga about the land tenure system and its impact on agriculture. The 1875 Constitution guaranteedevery adult male an area of farmingland of 8.25 acres (an 'ant). In recent years, this has proved impossiblesince land is no longer available. Populationdensity has approximatelydoubled in the past 30 years to about 150 persons/km2 currently. Because land/peopleratios have declined dramatically, there is an assumptionthat this is creating land shortagesfor agriculture. As noted elsewhere,the mission could find no evidence that farmers cannot gain access to land for expansionif necessary;land is availablethrough formal or informal leases for amountsof rental which vary accordingto the quality of land and the location. On Ha'apai a farmer had paid T$500 as a lump sum (plus gifts) for use of an 'api for 10 years; on Tongatapu $6,000 was paid for a similar lease.

3.22 While land is available,there is a problem since it is only on short term leases,although legallyleases can be up to 50 years. Many farmers complainedof this difficulty,which clearly discourageslandholders from investingin the land. Governmentmight use its influenceto encourage land owners to lease their land for longer periods. Some problemsof soil degradation,reduction in tree cover and other environmentalproblems are also startingto appear, and the report recommendsa program of environmental monitoring,especially on Tongatapu,so that the effects of changes in land tenure and agriculturalpractices (especiallyany move from rotationsto monocropping)can be assessedand policy changedaccordingly.

3.23 Human Resources. High levels of educationand health are sometimes thought of as constraintsto agriculture,since they assist Tongans to migrate to a number of countrieswhere their skills are valued. This is constantly referred to as a problem for the Tongan economy and, while there are costs, the Mission believes that the benefits for the rural sector have thus far outweighedthe costs. Rural wages are very high and this is commonly thought to be the result of outmigrationof rural labor, but it may equally be because of the demand for Tongan food crops from those Tonganswho have migrated.

3.24 On the other hand, these same levels of social servicesoffer an impressivehuman resourcebase on which to build; few developingcountries have farmers as well educated as Tonga. In 1985, in a populationof 10,121 farm family headss9/

* 2? have some form of agriculturalqualification, mostly tertiary;

9I Tonga: AgriculturalCensus, 1985. Bureau of Statistics,Nuku'alofa. - 28 -

* 52 have other tertiary qualifications

* 472 have a secondaryeducation

* 44Z have primary education

* 22 have no schooling

3.25 Tongans also have entrepreneurialskills, althoughthese are temperedby a lack of appropriatebusiness and financialacumen, and the demandsof the social system on those who are seen to have money. The future will requiremore attentionto business and managementskills, in both the privateand public sectors. Many Tongan projects demonstratethe drastic combinationof good ideas and risk takingwith bad planning and financial management.

3.26 The productivefarming system assists risk taking in agriculture, since no farmerwill starve if the venture fails. Adaptive risk-taking, giving new crops a try, is very common and has been the basis for many of the minor crops now incorporatedin farming systems. That the Tongan system leaves room for individualand private sector initiativeis demonstratedby several recent developments,including the growth of vanilla production,the developmentof a niche market for pumpkins,and the growingexport of traditionalfoodcrops.

3.27 Technology. Tongan agriculturaltechnology is often referred to as simple. On the contrary it is highly complex and continuallyevolving. The graftingof commercialagriculture onto the system is a major and underrated achievement. Tools, on the other hand are simple,and with labor costs very high, improvementsin mechanizationwill be necessaryin the future,but in a manner that does not destroy the farm system.

3.28 Agriculturalinputs such as fertilizer,seeds, equipmentand other items necessary for commercialproduction are sold through a number of importersand stores. There are periodic supply problemscommon in a small countrybut no farmer mentionedmajor supply difficultieswhen interviewed. The low-inputnature of the Tongan farmingsystem does not requirelarge purchasedinputs. While there are some supply problems in outer islands,the demand and need there is correspondinglylow.

3.29 Capital and Credit. Low-inputagriculture, as common in Tonga, needs very little credit. At times, the Governmentis concernedabout the low demand for credit and sees it as an indicatorof backwardness(especially in the agriculturalsector). But this is a common Pacific Island phenomenon: many of the highest yielding agriculturalactivities require no credit. Many of those that do requirecredit may, in fact, show returns too low to adequatelyservice loans (e.g. pumpkins).

3.30 The Tonga DevelopmentBank (TDB) has a large portfolioof agriculturalloans and has little problemwith loan defaults. This is another advantage of a small country; borrower are usually known to the Bank even before they borrow. Social pressures are used to ensure repayment, and security is rarely seized. The Bank has receivedconsiderable external support from IFAD, IDA, UNCDF and the ADB. They will lend for most farm - 29 -

inputs,but are reluctantto lend for labor. Because farmershave income from many sources,they normally requirepayments to begin as soon as possible,and before the crop is harvested,but only after working out the farmer's projectedcash flow. Security is taken over livestock,houses, cars trucks or tractors. Interest rates have recentlyincreased to 102 for loans less than T$l,000 and 11.52 for larger loans. Loans for vehicles are 132. They are also lendingto marketerssuch as FIMCO and the CommoditiesBoard. They are lookingat providingcredit for buyers and marketersof agriculturalproduce. In 1988 TDB make 2,681 agriculturaland fishing loans totallingT$4,104,000 (Table 5). The TDB has a Women's DevelopmentFund for loans up to T$3,000 which is mainly used for financingmulberry production,and mulberry purchases for tapa production.

3.31 The mission concludedthat credit availabilityis not generallya problem,although there have been some periods of tight liquidity. It suggests that the major constraintto agriculturaldevelopment and credit use is not the shortageof funds, but inadequatebusiness and financial understandingamong farmersand marketingproblems. However, this view is not necessarilygenerally shared.101

Table 5s Tonga DevelopmentBanks AgriculturalLoans (T$) ------Vanilla & Kava 1,654,210 Root Crops 1,103,452 Fruit & Vegetables 562,824 Fishing 416,640 Livestock 170,881 Mulberry 101,795 Agriculturalland 91,300 Other 3,600 TOTAL AGRICULTURE& FISHING 4,104,702 TOTAL OTHER LOANS 5,330,397 TOTAL LOANS 9,435,099

Sources Tonga DevelopmentBank Annual Report, 1988.

3.32 The Bank of Tonga is heavily involved in agriculturallending and artifactproduction from pandanus in Vava'u and will extend this serviceto Tongatapu. On Ha'apai the Bank is lendingfor mulberry and other agricultural activities. On Vava'u it is also lendingto farmersand buyers to finance food crop exports. Although they are often criticizedby politiciansfor not lendingsufficiently to agriculture,the two Tongan banks are servicing agriculturereasonably well.

3.33 Although individualsaving may not be a Tongan habit, group saving by financinginvestment through relativesoften takes its place. A tractor

10/ In discussingthe draft report,MAFF in particularfelt that the Mission had understatedthe difficultiesfaced by smallerfarmers in gaining access to credit. - 30 -

purchasemay be financedby requeststo relativesoverseas, and it is possible that a significantproportion of remittancesgoes to some form of farm investment. It is easier to get money for a specifiedpurpose from a relative than to make a general request.

3.34 Hazards. Tonga is prone to droughts,strong winds and, at irregularintervals, cyclonic stomss. Droughtsand storms are said to have created greaterproblems since 1980. Five cyclones of hurricaneforce have been recorded this century and in differentparts of the country. Strong winds cause damage to many crops, especiallybananas. The 1982 hurricane which hit Ha'apai and west Tongatapuwas a media event and attracted internationalattention and aid of many forms, includingfood (especially flour).

3.35 Mission members discussedthe problems of hazardswith many farmerswho all agreed that droughtwas the greaterproblem in terms of agricultureand food supplies,and that recent high prices of Tongan foods are related to droughts. Tonga has virtuallyno surfacewater, and therefore irrigationis not likely to be feasibleon any scale. However, there might be scope for further developmentof small scale water impoundmentschemes for such purposesas householdgardening and livestockwatering. The possibility of irrigationfrom groundwateron Tongatapuis also often discussed. There could be dangers in this, since the aquifer is thought to be fairly fragile, Before exploitingthis, a thoroughhydrogeological examination should be made.

3.36 Traditionallythe Tongan food productionsystem coped with hazards through its capacity to store food in the ground for long periods,helped by some food crops that can be harvestedwithin 2-3 months. Thus the main effect of hazards is on commercialcrops, rather than food supplies,and donors should only supply food when there are clear reasons for doing so. Depending upon the nature of the emergency,it may be more useful to assist with the distributionof plantingmaterials, especially short maturing crops like sweet potato, at the same time as MAFF monitors the food supply situation.

3.37 One recent problem is the reductionof tree cover as monocropping of crops such as pumpkins increases. This is reducing the shelter effect and requires a monitoringprogram combinedwith increasingagroforestry plantings.

3.38 GovernmentResources. Tables 6 and 7 show that MAFF funding for current expenditurehas fallen as a proportionof total expenditure. However, this reflectsthe 44? increase (in real terms) in expenditureon government general administration,which should be a cause for concern in a countrywhere such expendituresare alreadyvery high. A further large salary Increase approved for 1988/89will increaseadministrative costs further,and is likely to reduce the amount availableto financeoperational expenditures. MAFF fundinghas remained roughlyconstant in real terms and as a proportionof expendituredevoted to economic services. - 31 -

Table 6* GovernmentCurrent ExpenditureFiscal Years 1982183-1987187 (T$ million, current prices)

…------__------82183 83184 84185 85186 86/87 87/88 ------_------__------

Public administration 5.1 6.0 7.0 9.2 12.2 12.3 General administration 2.9 3.6 4.3 5.9 8.2 8.1 Fiscal administration 0.5 0.6 0.5 0.7 0.9 1.1 Law and order 1.7 1.8 2.2 2.6 3.0 3.1

Social and communityservices 5.0 5.3 6.1 7.6 8.5 9.8 Education 2.3 2.4 2.9 3.7 4.2 5.2 Health 2.2 2.3 2.6 3.2 3.3 3.6 Pensions and Gratuities 0.4 0.4 0.4 0.5 0.6 0.7 Other 0.1 0.2 0.2 0.3 0.3 0.3

Economic services 5.5 5.0 6.0 7.1 7.4 7.7 Agric., forestry & fishing 1.5 1.5 1.7 2.0 2.1 2.3 Tourism 0.2 0.2 0.2 0.2 0.2 0.3 Transportand works 3.2 2.8 3.9 4.7 5.0 5.1 CommunicationsLa 0.6 0.6 0.2 0.2 0.1 0.1

Other /b 0.2 0.4 0.8 1.6 1.7 2.5

TOTAL 15.8 16.7 19.9 25.6 29.7 32.3 ------___------Sourcet C. Browne Economic Developmentin Seven Pacific Island Countries InternationalMonetary Fund, Washington,1989.

/a Excludes post office expenditure lb IncludesStabex transfersto the CommoditiesBoard; excludes amortizationon public debt and appropriationsfor the development budget and sinking funds.

Notes This is based on actual expenditure,not budgetaryallocations. - 32 -

Table 7. MAFF Funding Relative to Total Expenditure & Expenditureon Economic Services

…------__------82/83 83/84 84/85 85/86 86/87 87/88 ------_------__------_- Proportionof MAFF funding to total currentexpenditure (2) 9.5 9.0 8.5 7.8 7.1 7.1

Proportionof MAFF funding to expenditureon economic 27.3 30.0 28.3 28.2 28.4 29.9 services (Z)

MAFF expenditureat 1982 1.5 1.4 1.5 1.5 1.3 1.4 prices /a ------__------

La Deflated by Consumer Price Index Sources Derived from Table 6

Table 8s Agriculture,Fisheries and Forestry:Development Budget 1980/81 - 1984/5 (T$) ------__------_ 80/81 81/82 82/83 83/84 84/85 ------__------_ Agriculturegeneral 649,128 934,826 1,044,222 1471,818 1,669,834 Livestock 62,226 71,539 53,459 41,665 24,665 Forestry 62,705 108,270 117,255 72,660 81,485 Fisheries 3,544,607 278,648 2,005,460 759,436 1,913,539

Total 4,318,6666 1,393,28 3,220,396 2,35,569 3,689,523

Total, all sectors 21,176,248 14,743,246 14,233,148 19,166,102 19,500,040

Proportionof MAFF DevelopmentExpend- iture to total. (2) 20.4 9.4 22.6 12.2 18.9 ------__------_ Source: Fifth Five Year DevelopmentPlan. Central PlanningOffice.

3.39 Projectsare funded throughthe DevelopmentBudget which is virtuallyentirely funded from external donor sources. As can be seen from the tables,development expenditure is much greater than recurrentexpenditure and changesconsiderably from year to year as projects start and end. These extreme variations,largely caused by the dependencyof the developmentbudget on external funding,cause considerableproblems within MAPF in balancing ongoing extension,research, regulation and planningwith the very irregular yet urgent demandsplaced on the staff by externallyfunded projects. - 33 -

3.40 Whether these levels of funding for agricultureare appropriateis difficultto say without a much more rigorousassessment of the sector'sneeds and MAFF's capacitythan was possible in this review. There is some sentiment within Governmentthat agriculture(and MAFF) have failed to perform and that resourceallocations should thereforebe reduced. However, given the earlier conclusionthat the performanceof the agriculturesector has been much better than generallybelieved and that agriculturecan be a major source of future economic growth, the sector may in fact warrant an increasedallocation of resources.provided they can be effectivelyused. Some suggestionsfor increasingthe efficiencyof resourceuse through institutionalstrengthening are given in the final chapter.

Key Constraintsto Tongan Agriculture

3.41 The mission believes that the major constraintsto agricultureare not primarilytechnical, but economic and institutional;nor do they result from Tonga's culture,its small market, or the remotenessof the country.

3.42 Labor costs. Tonga's labor costs, with rural casual labor at T$2- 3 per hour, considerablylimits the choices availableto farmers. But at the same time, the high rural wage is a reflectionof the productivityof both traditionalcrops and other crops. Some farmersform traditionalwork groups and appear to get higher labor output through the co-operativeeffort. The governmentis encouragingthese groups and should continueto do so.

3.43 Future agriculturalstrategy will thereforehave to find ways to relievethe labor constraint,by improvinglabor productivity,while sustainingthe farming system and soil structure. MAFF's mechanization program is in considerabledemand, but mechanizationcan create problems for coconuts and interferewith the operationof the farming system. Full cost pricing of tractorhire is essential if mechanizationis not to be overused. In the meantime,research should begin now to developmethods of cultivation (e.g. use of ground covers,minimum tillage)which reduce labor without mechanization,and to test types of tractor and equipmentto minimize damage to the system and maintain soil fertility.

3.44 Exchange Rate and Tariff Policy. Tonga'sagricultural export performanceis closely linkedwith the exchange rate policy. In particular, two aspectsof the exchange rate managementneed to be carefullyconsidered. First, as noted, the link between the Pa'angaand the Australiandollar has caused wide fluctuationin both the rate of inflationand the real exchange rate. The Governmentneeds to considermoving towards an exchange rate system that better reflectsTonga's trading partnersand developmentsin competitor countries. Second,the appropriatenessof the level of the real exchange rate should be assessedin view of the medium-termobjectives to promote exports. Recent analysisundertaken by the Treasury suggeststhat the real effective exchange rate has appreciatedand is at its highest level since 1975 against all major tradingpartners.

3.45 It is encouragingto note that the authoritiesplan to institutea managed float of the Pa'angaagainst a basket of currenciessometime in the future. However, a substantialdownward adjustmentof the real effective exchange rate would be required to enable Tonga to competein export crops such as copra or with msot imnorted food crops. An alternativewould be to - 34 -

considertariffs or other incentivesto agricultureand agricultural processing,especially in the establishmentphase.ll/ Whether or not any such policies are considered,it is importantthat these macro-economiccurbs on agricultureshould be recognized,and should be borne in mind when the performanceof the sector is being assessed.

3.46 MarketingConstraints. The constraintmost often mentionedby farmersand others associatedwith agricultureis marketing. Post-harvest technologyand marketingwas elementaryin earlierTongan systems and perhaps this accounts for the strengthof the farmingsystems and the weakness of marketing. Domestic marketingin Tonga is reasonable,and a considerable volume of produce is sold. But it could be greatly improved throughbetter market facilitiesand more urban and peri-urbanmarket centers.12/ There needs to be better planning of domesticmarketing, but only to further encouragethe good work already done by privatemarketers, not to become involved in the process of marketing.

3.47 There have been a few remarkablemarketing achievements, particularlywith ethnic niche markets,where many farmers, and now traders, have developedtheir own marketingsystem by utilizingthe strengthsof the Tongan social and economic structure. However, Tonga faces problems in trying to penetratemore demandingand competitivemarkets. With a small productionbase, the costs of marketing a new product are high, and neither the private sector nor the public sector has the expertiseto mount a major overseasmarket promotioneffort. Combinedwith exchange rate effects, this createsa pricing and marketing frameworkwhich makes any marketing initiativesvery difficult.

3.48 The challengefacing Tonga, as it tries to open up new markets, is to improvethe efficiencyof the whole post-harvestmarketing chain. The problem is that if one part of the chain from producer to final consumer fails, the whole system can break down, prejudicingthe entire industry. This is what happened in the export of watermelons,where a successfulexport market was lost (and the entire horticulturalexport potentialendangered), when fruit fly larvae were discoveredin one consignment. Quarantineand produce inspectionto meet the requirementsof importingcountries are an increasinglyimportant part of the total post-harvestchain in agriculture.

3.49 There is a tendency for Tonga to view other countries'quarantine requirementsas somethingnegotiable, so that when there are apparent problems there is the hope they might be changed. While there have been some capriciousdecisions by importers,these are extremelyrare. Tonga must be at all times aware of all new and likely requirements,since often these are known in advance and can be planned for. There is little doubt that major markets such as New Zealand,Australia and the United Stateswill impose even

111 The Tongan broiler industry is an example of an industrywhich would probably have survivedif it had been given the degree of protection considerednormal for industrialenterprises.

12/ The Governmenthas recentlydecided to expand market facilities. - 35 -

stricterrequirements in the future,not only to keep out pests and diseases but also to control pesticideand other chemical residues.

3.50 There is a need in Tonga to improveproduct management on the farm during production,the selectionof produce for harvest, the packaging and storageat harvest,and preparationand handling for export. Quality should be consistent. Experiencein other countrieshas shown that investment in good managementand post-harvesthandling of produce yields a profit in terms of increasedprice through buyer confidencein the quality of the product. Som3 marketingproblems (e.g.wharf strikes In New Zealand)lie outside Tonga'scontrol; but many are within its control, and should become part of an ongoingpost-harvest management program, as suggestedin the final chapter.

3.51 InstitutionalCapacity. As the key developmentministry responsiblefor agriculture,forestry and fisheries,MAFF carries a heavy responsibility,and should be adequatelyequipped to do its job. It must be recognizedthat the Ministry is (and will continue to be) a small agency, which cannot be expected to carry out the many functionsthat would be performedby an agriculturaldevelopment ministry in a larger country. However,MAFF is at present unable to carry out some of its more basic functions,as discussedfurther in ChapterV. At the same time, donor needs for projectcounterparts are placed on a small staff and can be very demanding. There is also a tendency for MAFF to do things (e.g. spray bananas)which could be done by farmersor the private sector.

3.52 The Tonga CommoditiesBoard (TCB) is a statutorymarketing authoritycharged with exportingagricultural produce with the objectiveof ensuringthe best price for the grower. It also operates the Copra Price StabilisationFund. Besides its agriculturalfunctions, it is also engaged in constructionand the retailingof constructionproducts. In the late 19709 it was profitable,but since 1980 there have mainly been losses. Provisional figuresshow losses of T$l.8m for 1988 and T$0.5m for 1989 and the Board is now basicallyinsolvent. It operatesthe coconutoil mill and the desiccated coconut factory.

3.53 In its early years, TCB was the only agriculturalproduce exporter and constructionoperator. Its powers are very wide and, in theory, it has the task of quality control of exportedproduce but it has not done this except for its own marketed produce. It has come under increasingpublic criticismin recent years, and has not respondedcompetitively to the growing, and more efficient,private sector exportersof agriculturalproduce. In addition,by seeing itself basicallyas an exporter of coconutproducts and bananas, it has locked itself into the weakest performingexport products. The longer term prospectsfor its coconut products,given its costs and decliningworld markets, are poor.

Conclusions

3.54 Several conclusiontsemerge from the profile and performanceof Tongan agriculture. All those crops that have done well share a number of crucialattributes: (a)they fit into the existing farm system and culture; - 36 -

(b) they give an acceptablereturn in comparisonwith other crops and economic opportunities;(c) they are not too risky or very long term, with possible incomemany years hence; and (d) they have assuredmarkets.

3.55 When thinkingof a new crop, it is importantfor the farmer--and MAFF or donors--tothink these featuresthrough very carefully. Some features depend on each others for example achievingan acceptablereturn on a crop depends on securinggood markets. Studiesof project and other reportsby the Mission showed that the degree to which farmerswould supporta new activity was often overstated.

IV. PROSPECTSFOR GROWTH

Introduction

4.1 The analysisof past agriculturalperformance and the constraints to agricultureshow that Tonga has a considerablepotential for increasing agriculturalproduction and growth provided it recognizesits constraintsand works well within them. It must not think only of its problems but look carefullyat its advantagesand constraints.

4.2 The key characteristicsof currentlysuccessful activities and those which will be developedin the future ares

* returnsto farm labor must be well above currentwage rates; labor intensiveactivities must be avoided unless returnsare high;

* exportswill be as fresh produce or with processingthat can be done on the farm, is not expensivein relationto product returns, or can be done centrally,without the need for substantial economiesof scale;

* producemust be storablewithout quality loss within the normal limits of transport;

* producemust have high value in relation to transportcosts;

* qualitymust be accordingto consumer requirements,including presentationand packagingand quarantine.

4.3 Vanilla, kava and root crops exports (includingcassava, which is peeled and frozen for export) are good examples of Tongan prottuctswhich meet these requirements. Some aspectsof this process have been weak in Tonga, but present indicationsare they could be improvedover the next few years in line with the strategy suggestedlater in this chapter.

HouseholdConsumption

4.4 From a national viewpoint,great importanceis attachedto domesticallyconsumed produce. This is mainly food, but also includespigs, poultryand cattle, fish, coconutproducts, breadfruit, kava, building materials.firewood, pandanus, and mulberry. Whether it is purchasedor home _ 37 -

grown matters little at the national level, but at the farm level it is importantto maintain food securitythrough household production and consumption. Experiencein Tonga and other Pacific Island countriesshows that people readilymove back to increasedhousehold consumption if there are significantchanges in the price or availabilityof imported food. Similarly, if there is a long term drought, less food is supplied to the market, prices rise and people rely more on home production.

4.5 Future trends in the consumptionof householdproduced food are difficultto predict since informationis very weak. The rising level of food imports in the past suggeststhat householdconsumption may have fallen. On the other hand, there is no evidencethat Tongan familieshave cut back on their consumptionof traditionalfoods, and the relativelyhigh average daily calorie intake suggests that importedfoods may add to, rather than replace, the comsumptionof staples. It is thereforeassumed that averageper capita consumptionof home produced items will remain constant--andhence that total home consumptionwill grow at the same rate as population (less than 12).

Marketed Domestic Production

4.6 Given the increasingvolume of sales of local agriculturalproduce in Nuku'alofaand Vava'u, there is no doubt that domesticsales will continue to increase. The future expansionof the domesticmarket will be sensitiveto factors, such as populationgrowth, broader macro economicpolicies and the potentialexpansion of tourism,that are beyond the scope of this report. Some of these factorscould encouragelocal production (for example,a major devaluationor higher tariffswould shift consumptionin favor of local foods by raising the cost of importedfoods). However, the scope for future expansionis severelyconstrained by the small size of the market, and the limited potentialfor furtherviable import substitution.

4.7 For the purposes of this report, it has been assumed that GDP expands by a little over 3 percent per year in real terms, that there will be some expansion in tourismand possibly a minor devaluation. Under these assumptions,the local market for domesticproduction is projectedto grow at 4 percent a year over the 1988-S'period.

4.8 The domesticdemand for wood is likely to be about 10,000 cu.m.overthe next decade and will continue to outstrip supply. About 70 percent of Tonga's sawn timber requirementis now importedand, with falling productionof indigenoushardwoods, local sawmillsare largely dependentupon coconut. With importedsawn timber costing twice the price of domestic timber, there is no lack of incentivefor production. There is also a good demand for traditionalforest products,such as bark from the paper mulberry tree (for tapa cloth) and the koka tree (for dye), wood for carving and firewood. However, a major effort will be needed to take advantageof this market.

Niche Markets

4.9 Exports to Tongans. The strong and stable demand for root crops in New Zealand,Australia and the should continue to be a major market for agriculturalexports. The growth of the market will depend on the number of overseas Tongans and Tonga'scapacity to competewith other Pacific - 38 -

producersof ethnic foods. In the longer term, it can be expected that overseas Tonganswill they lose their taste for ethnic produce generationby generation. On the other hand, there is a tendency for Pacific Islandersto eat more of the preferredstaple as their incomesrise. The long term growth trend is estimatedat about 5 percent a year in real terms.

4.10 If there is ever a fall in remittancesfrom abroad, it is to be hoped that it will be at least partly offset by increasedfood purchases. However, at current exchange rates and interest rates, there is every incentivefor exportersto leave their money overseas,or to remit only part back to Tonga. Tonga should encouragetraders to purchase for cash in Tonga for export so that more of the proceedsare returned throughthe banking system.

4.11 Vanilla prospectslook good in terms of productionand prices. The price trend is expectedto be stable in real terms in coming years, but year by year prices depend on Madagascar'sproduction and there will be short term fluctuations. Sales are denominatedin US$ so that the gradual depreciationof the Pa'angaagainst the USS has also helped raise prices to growers. A core of experiencedgrowers and processorsis now emergingwhich should improve the prospectsof the industry. There is thereforelikely to be moderate growth of perhaps 4-5 percent in real terms, provided high quality can be achieved.

4.12 Pumpkins. The future of pumpkins is uncertainand much will depend on the prices growers get for their crop in 1990. If yields and qualitycan be improvedand marketingcosts reduced,a group of good growers may emerge and there would be an opportunityto secure a permanentfoothold in this market. But this will requirethe growersassociations to be quickly strengthened,and it will be necessaryto look into the impact on soil quality of continuedmonocropping with pumpkin.

Other Exports

4.13 Coconut Products. The future of copra-basedproducts is not promising,given the low returnsto labor from copra productionand the probabilitythat internationallong term prices will continue to fall or stay at current levels. Copra will continue to be produced in the outer islands, and could temporarilyre-emerge as a significantexport if there is a major recessionin the Tonga economy or if world market prices surge, as a result of disruptionsin major producingcountries. We thereforeproject no increase in coconutexports over the next few years. In the longer term, the prospects for coconuts as an export crop will depend on whether new technologiescan be developedthat offer a higher return to farm labor. Exports of whole nuts will continue,and the market should be pursued aggressively,but the amount will remain small in relationto potentialproduction.

4.14 Bananas. Although long tenm internationalbanana prices will probably continue to decline slightly in real terms, the export industrycould be rejuvenatedif the industrywere reorganizedto produce consistent,high quality fruit. The followingchapter suggeststhat Tonga should concentrate on improvingquality control and managementduring cultivation,harvest and post-harvest. If this is successful,bananas would be the first crop to benefit and could become a steady export earner supplyingthe smallersweeter - 39 -

bananas on the New Zealandmarket. Since that market is about to be de- regulated,this quality improvementis the only hope for the export industry. There are growerswith the knowledgeto do this and it is feasiblewithin a year or two. This could mean exportsof about T$2 million a year.

4.15 Other crops. If the quality improvementprogram recommendedbelow is successful,there would be a good prospect for other niche markets to open up in two to three years' time. The potentialis difficultto predict, but even one valuable new export line every five years or so could provide the basis of sound growth for the agriculturalsector in the future. The mission projects a growth rate of 10 percent a year through1994, but this assumption is subject to a wide margin of error.

Overall Trend

4.16 In overall terms, value-addedin agricultureis projectedto grow by a little under 3 percent a year in real terms over the next six years. The basis for this projectionis:

ProjectedValue-Added (T$ mill.in 1988 Prices) 1988 1994 Growth Rate

Productionfor home consumption 31.5 33.2 0.9 Domestic sales 13.5 17.1 4.0 Exports 7.5 11.3 7.1 Total 52.5 61.6 2.7

In the longer term, the prospectsthrough the remainderof the centuryclearly depend very largely on how successfulTonga is in developingnew exports (throughexploiting its potentialcomparative advantage in niche markets) or regainingsome of the lost copra exports (throughthe introductionof more efficientprocessing techniques).

V. A STRATEGYFOR FUTURE DEVELOPMENTOF AGRICULTURE

A. THE ELEMENTSOF A STRATEGY

Introduction

5.1 The report concludesthat there is potentialfor accelerating agriculturalgrowth in Tonga to improve incomesof farmersand benefit the country as a whole. This requiresa strategywhich has three main elements in our view: (i) to build upon the strengthsof Tonga's farming system and exploit the advantagesof its smallnessand remoteness;(ii) to concentrate efforts of governmentand donor agencies alike on key priorities;and (iii) to redefine the role of the public sector in the developmentprocess, and strengthenits capacity to perform its major functions. - 40 -

Buildingupon Strength

5.2 This report has suggestedthat Tonga has much working in its favor, includingits robust traditionalfarming system,the potentialto produce a wide range of crops, a well educated and enterprisingfarming community, and its rich culture and stable social system. The first and most obvious strategy is thereforeto build upon these strengthosfor example,by promoting developmentactivities that complement,rather than conflictwith, the existing farming system and social order.

5.3 We have also suggestedthat some of Tonga's supposedhandicaps--such as its small size, remotenessand the extent of overseasmigration--have in fact benefittedagricultural development in some ways in the past, and could be furtherexploited in the future. On the oth r hand, there are real handicaps to agriculturaldevelopment, particu' rly the high exchange rate and wage rate, which may be beyond the control of 4griculturalplanners but which must be recognizedin formulatinga strategy. It is, for examplef futile to develop programs,like some that have tried to rehabilitatecopra production in the past, that ignore the realitiesof high labor costs.

ConcentratingEffort

5.4 Tonga has considerableagricultural potential. However, to successfullyexploit this potential,while overcomingthe various constraints noted earlier,requires careful planning,which usually requiresmost of the followingplanning steps to be undertaken:

o ensuring that a thoroughmarketing study is done;

a gettingdata on actual yields (from Tongan trials or from other countries)and making realisticestimates of yields under Tongan conditions;

e doing financialanalyses of farm operations,in addition to crop budgets, to estimate profitabilityand returns to labor;

checking that all new activitiesare feasible,given Tonga's economic constraints,and compatiblewithin existingagro-forestry systems;

* identifyingthe essentialfactors for successof the new activity.

5.5 The fact that Tonga is a small country,and its projectstherefore typicallysmall, does not lessen the need for carefulplanning and screening of projects and activities. On the contrary,since even one bad project in a small countrycan harm a large proportionof the farmingcommunity, it is importantto ensure that any interventionsin the existing farming system are carefullyanalyzed before they are approvedand followedthrough diligently during implementation.Past experiencehas indicatedthat projectshave sometimesfailed when all but one of the above steps were undertakenwell.

5.6 But it will be difficultfor the public sector,with its limited capacity,to do too many things at once. This suggeststhat government departments,and in particularMAFF, should concentrateeffort on their most - 41 -

crucial functions,leaving other activitiesto be undertakenby the farmersor the private sector,or to be left until later.

5.7 There is also a need for donors to concentratetheir project aid. The mission found that there have been about 50 discreteexternally-financed projects affectingagriculture and forestry since 1980.13/ This implies 5-6 new projects and 10-12 ongoing projects each year. These imposeda heavy burden on MAFF, which was often unable to provide agreed counterpart resources,or those resourceswere so stretchedas to become ineffectual. Reportingrequirements have also placed a heavy load on MAFF staff.

5.8 It is importantfor Tongan politiciansand those officials negotiatingexternal assistanceto realize that aid is not 'free",and that there are cases where a small aid project in Tonga has absorbedmore Tongan resourcesin negotiationand administrationthan the projectwas worth. As a general guideline, it is suggested that it would be better for Government and donors alike to concentrateon implementinga few, well-designedand high- priorityprograms successfullythan to dissipateefforts over many, poorer projects.

Defining the Role of the Public Sector

5.9 Tongan farmershave proven to be largelyself reliant in the past, and have demonstratedthe capacity to try out new technologiesand pioneer new marketswith only limitedhelp from governmentagencies. There is no reason to change this admirablepositions we do not see the need for major public sector investmentprograms to develop agriculture. On the other hand, the agriculturesector is clearly facingnew challengeswhich will require the public sector to play a new and differentrole--not to intervenedirectly in productionand marketing,but rather to play the more sophisticatedrole of facilitatorof development,advisor to the farmer and guardian of the environment.

5.10 The priority for action in the immediatefuture is thereforeto strengthenthe capacity of MAFF--as the key public sector agency--tobetter support the farmer in facing the challengesof the future. Given the need to concentrateeffort, it is suggestedthat MAFF should focus its main attention on five major regulatoryand servicefunctions:

* technicaland market researchand the disseminationof technology and advice to farmers;

* protection services,quarantine and quality control,both domesticallyand for export;

~o agriculturalpolicy analysisand advice;

* economic and financialanalysis of agriculturalactivities;

13/ Most of these projectswere administeredby MAFF, and some were the responsibilityof other ministries,but had agriculturalcomponents or implications. - 42 -

* environmentalmonitoring, including soil and water quality, tree cover, and use of chemicals;

The actions needed to strengthensome of the most importantof these functions are discussedin a later section.

5.11 At the same time, some of the present activitiesof MAFF and the Tonga CommoditiesBoard can be scaled down or phased out altogether. The guidingprinciple might be that governmentagencies should not undertake functionsthat can be carried out by the private sector or by the farmers themselves,either individuallyor through groups and associations. Thus in MAFF, it would be desirablefor extensionstaff to concentrateon their extensionfunctions, and to leave farmersto perform tasks (e.g.spraying) which are now done for them. An exceptionto this principleis when natural hazards, such as hurricanes,floods or droughts,call for special,but short- lived, assistancefrom Government. It is also suggestedthat mechanization services should be phased out over time and private contractorsencouraged to provide them.

B. PROGRAMPRIORITIES

Buildingupon the Farming System

5.12 The robust agro-forestryfarming system that has evolved in Tonga provides a good foundationon which to build future development. To do this successfully,it is suggested,will requireboth Governmentand the donor communityto recognizethe importanceof the system, to refocus researchand extensionactivities to support it, and to ensure its sustainability.

5.13 The first need is to reaffirmthe importanceof the traditional farming system to the Tongan economy and to recognizethat it is a system. It is the agro-forestryfarming system as a whole (which includesthe crops, livestockand trees that are its components)that should be the focus of work with farmers. In the past, a farm was seen more as a collectionof individual crops than as an integratedwhole. Any new technologyor crop should always be judged initiallyby its impact on the system. Foreign researchersand project designersshould use field visits and some of the existing publicationson the Tongan farming system to assess the interactionof their project on the system. Such an approachhas been noticeablyabsent in most project documents,particularly those targetedat coconuts and cattle.

5.14 Extensionand other MAFF staff should continue to be systems oriented and learn from farmersas well as give advice. MAFF should look at donor project proposalsand documentsfrom the viewpointof the farming system and the project'simpact. These commentsmay sound trivial:they are not. Most people working in agriculturethroughout the world think of only a few products at a time, not a system.

5.15 Second, there is need to design and implementan integratedfarming systems developmentprogram that can both preserve and increase the - 43 -

productivityof existing smallholderfarming systems. MAFF has already embraced the concept of a farming systemsapproach to smallholderdevelopment, and most of the componentsof a farms system approach are in place. Some gaps need to be filled:

o to institutea small farming systemsresearch program to study farm systems,montor the impact of new initativeson such systems, identify farm level problems,design methods of solvingthem, and liaise with extensionin introducingand testingappropriate techniquesto farmers;

o to collect and ananlyze data on the economicsof the whole farm, as an extensionof the excellentcrop budget analyses that have been developed (and extensivelyused in this report).

5.16 HAFF also lacks the organizationalstructure to forge the necessary links between farm surveys,economic analysis, research and extension. Yet a small country such as Tonga has an exceptionalopportunity to successfully forge these links into a strong chain of technologicaldevelopment, and this should be a priority for MAFF in the future.

ExploitingMarket Potential

5.17 Tonga is a very small countryand can use this feature to advantage in finding small, specializedor periodic marketswhich requirea relatively small quantityof a specificproduct, produced to very specificstandards. There are a number of differenttypes of niche markets:

o Ethnic markets: Tongansoverseas prefer food from their own country, and demand a particular"Tongan" quality. There is also a wider Polynesianmarket that Tonga can tap.

* Quality markets: Premium qualitymarkets can often be exploited, provided there is consistentlygood quality control from growing to marketing. For example,very good managementof vanilla production and curing can yield a premium of about T$20/kg.

* Seasonalmarkets: There may be seasonalopportunities which Tonga can exploit. The export of pumpkin to Japan during the November- December period is an example.

5.18 Tonga has done very well with the ethnic nlichemarket and should continue to do so. However establishingsales in the non-ethnicmarkets is a completelydifferent matter. Unlike the ethnic markets, which are relatively assured, these markets are very competitiveand demanding,and tend to come and go as other countriesenter or leave the market. They thereforecall for very specificknowledge about the precise standardsrequired by the market, the seasonalnature of demand and prices, transportand handling costs, appropriatepackaging, and the importingcountry's marketing system and quarantinerequirements. More importantly,they require that a system be in place to ensure that Tonga can consistentlymeet the demanding (and continuallychanging) requirements of these specializedmarkets. - 44 -

5.19 The mission recommendsthat a post-harvestmanagement program be developedover the next few years to help Tonga exploit its potentialmarkets. There are two separatebut closely linked and importantcomponents:

o QuarantineControl and Inspection. The first priority is to implementMAFF's current proposalsto upgrade the QuarantineSection to a Quarantineand InspectionDivision--a task that warrants external assistance. An importantfunction of this new division will be to monitor the quarantineand chemical residue requirements of importingcountries and ensure that new requirementsare quickly incorporatedinto the production-marketingprocess. This will require settingup a serviceat the point of shipment. In this way a small countrycan keep up with changingmarket requirementsmore easily than a large country.

O QualitgControl and Assurance. Tonga's success in developingand sustainingspecialized overseas markets will depend on its ability to maintain uniformquality standards. This will require cooperationfrom everyone in the marketingchain, includinggrowers, traders,processors and shippers,but Governmentmust take the lead by laying down standards,providing coordination, registering growers and exporters,and, where necessary,enacting legislation. A final check on quality can be added at the point of shipmentonce the new QuarantineDivision adds a produce inspectionfunction to its phytosanitaryservice.

5.20 There remainsthe importantfunction of market informationand promotionservices. As indicatedearlier, MAFF should not take on direct responsibilityfor marketing,but should providemarket informationand assist the private sector in devlopingnew markets. Since it is apparent that individualfarmers seldom have the capacity to successfullypenetrate the very demandingniche markets, two alternativeapproaches are suggested:

o The first is to encouragethe formationof growers associationsor cooperativeswhich, by aggregatingtheir resources,can overcome the limitationsof the individualsmall farmer. The recent establishmentof growers associationsto exploit the Japanese pumpkinmarket is a very positive development. But if they are to succeed in a competitivemarket, these young associationsmust have professionalmanagement and access to market information,and there is a good case for governmentdepartments and donors to support them during their formativeyears. Such supportmight provide training to associationmembers, financea share of the costs of employing qualifiedstaff (or provide technicalassistance), and meet the costs of overseas travel.

O The second might be to attract overseasbusinesses with the technicalknow-how, market knowledgeand business skills to pioneer new industries,either on their own or in partnershipwith Tongan companiesor growers associations. MAFF may need to play a promotionalrole in undertakingpreliminary studies of possible export markets, seeking out potential investorsand facilitating their establishment. - 45 -

StrengtheningPublic Institutions

5.21 MAFF. The above discussionof the prioritiesfor the public sector in promotingaccelerated agricultural development has implicationsfor organizationand management,particularly in MAFF. In discussingthese, it is importantto avoid making unrealisticproposals for administrativereform that Tonga neither needs nor can afford. Most of the importantdecisions and actions concerningthe future of the sectorwill continue,as in the past, to be taken by Tonga's farmersand other individualsin the private sector. In our judgement,MAFF should remain a small agency, but it must be able to perform its basic functionswell, even if it has to shed some of its previous activitiesto make way for new and more urgent functions.

5.22 In December,1988 MAFF submitteda comprehensiveproposal to the GOT Working Group on SalariesRevision and AdministrativeReview for strengthening and restructuringthe Ministry. The proposal,which has not yet been approved in full, called for the upgradingof certainposts (particularlyheads of divisions),salary revisionsand a major reorganization.The proposednew structurewould have nine divisions,after the creation of four new divisions (for planning,livestock, quarantine services and administration).

5.23 The proposalwas prepared by the Directorand senior staff of MAPF after much study, and would go a long way towards equippingthe Ministry to perform its key functions. A few structuralproblems seem to remain:

• The spread of control of the Director (with nine divisions) is broad, suggestingthat, where appropriate,some divisionsmight be merged or disbanded. For example, the MechanicalServices Division might be merged with Administration(pending progressive privatization),and the FisheriesDivision may be upgraded to a separatedepartment.14/

• The technicaldivisions have no direct control over functionsin the Outer Islands;instead a generalistin each island supervisesall staff and reportsto an Outer Island sectionin the Administrative Division. This structureis dictated by the scarcityof experienced staff, but in time MAlF should introducea single structurefor the whole country.

• Some functionsseem to be misplaced:for example,marketing is a responsibilityof the AdministrativeDivision, while the Specific Projects section falls under Extension. Both of these functions might be more appropriatelybe located in other divisions.

5.24 MAFF may wish to considerwhether these problems need to be addressed,either by adjustingthe presentproposal or at some future date. However, the precise structureof the Ministry is not, in our view, a major issue. The Ministry is so compact, comparedwith those in larger countries, that almost any structurecan be made to work, provided that two conditions are fulfilled: (i) the key functionsof MAFF are reflectedin the

14/ This has been agreed since the review was undertaken. - 46 -

organizationalstructure, and are performedby qualifiedand experienced staff; and (ii) there is good managementand coordinationbetween functions.

5.25 Arising out of the earlieranalysis of priorities,it follows that there are in our view severalkey areas in which the capacityof MAFF needs to be strengthened:

Policy Analysis. Up to now MAFF has had very little capacity for policy analysis and advice,which means that importantdecisions, which may have major ramificationsfor the sector,are taken by other ministrieswith less insight into the implicationsfor agriculture. For example, it appears that MAPF played a largelypassive role in the decision to lower import tariffs on chicken,which destroyedthe local broiler industry. It is vital, in our opinion, that MAFF should be able to strongly influence,if not by itself determine,policies for the sector for which it is responsible.

It is equally important,in our view, for MAFF to have an active role in influeri'ingmacro-economic policies, to ensure that the overall policy environLtent is conducive to the development of agriculture. Areas of governmentpolicy in which MAPF should have an input include:

o exchange rate and wage policy; trade policy (includingtariffs on imported foods and agriculturalinputs) and governmentpolicies in regard to import substitutionand export promotion:

* monetary policy (includinginterest rates and measures to promote the financialcredit institutions);

* taxationpolicy and the use of subsidiesas a developmenttool;

* regionalpolicies within Tonga;

e privatizationof agriculturallyrelated industries:

* land and conservation;and

• nutritionand health.

Financialand EconomicAnalysis. After reviewingmany project documents,the answer to "why have many projects failed?' is felt by the mission to be a deficiencyin financialand economic analysisand, in particular,a failure on the part of project designersto use appropriatevalues for the factors of production. The most frequent error is to underestimate.he opportunitycost of labor (as indicated by the returnsto labor availablefrom existing crops and from casual labor). This problem is well analyzed in an SPC project publicatfon which should be more widely used by project analysts.15/

15/ F. V. Sevele A Shadow Pricing System for the Kingdom of Tonga South PacificCommission, Noumea, 1984. - 47 -

It is importantto act quickly, since poor economicand financial analysis can create a vicious circle of poorly designedprojects, high demands on local resources,and poor projectperformance. MAFF is not at present equipped to undertakesuch analysis. The most urgent need is to undertakebasic (i.e. not very sophisticated)analysis of project proposals, (includingthose from donor agencies),particularly for their impact on farm incomes.

ProJect/ProgramPreparation. MAFF now largelydepends upon donors or donor-financedconsultants to draft projectdocuments, since it has little capacity itself. MAFF should take a more proactiverole in preparingits own programs and seekingexternal assistanceto support them. A good approachwould be for planning staff to work with their technicalcolleagues to developbroad sub-sectoralprograms (e.g. for research,livestock or forestry)that would be suitablefor external funding. Interesteddonors could be asked to move "up stream'by assistingthe preparationof such programs (see below).

Farming SystemsApproach. As indicatedin para. 5.15, there is need to fill in some of the present gaps, particularlyin farming systems research and farm economicsanalysis, and to provide the necessary linkagesbetween research,extension and economic analysis. The Ministry is too small to require the settingup of a separateunit for this purpose. The existing researchcommittee can be used to coordinateannual research and extensionactivities, but for longer term strategicchoices a broader decision-makingmachinery is required (see below).

Quarantineand Quality Control. MAFF has indicatedthat it intends to upgrade the QuarantineDivision, which we suggest could also be used to undertakefinal quality assuranceof export couw'dities. An external assistanceproject might be necessaryto suppcrtthe establishmentof the Division.

Marketing. A small unit needs to be establishedin MAFF to manage the post harvestmanagement program and to undertakethe market promotion functionssuggested in para. 5.20. A small, but well qualifiedand enthusiasticstaff is requiredto quickly elevate these functions. Given the importanceof launchinga post-harvestmanagement program, it might be useful to establisha close organic relationshipbetween marketingand other export-relatedfunctions, such as quarantineand quality control/assurance.

Forestry. The ForestryDivision needs to be strengthenedto carry out even a minimum program of resourcemanagement. Both the head office and district offices requireadditional manpower and funding,and there are several areas, such as researchand extension,where external technicalassistance will be needed.

5.26 If these gaps in the present structureof MAFF can be filled, it will then be in a much strongerposition to determineits future development strategy. For example, in assessingthe feasibilityof promotinga new export crop, MAFF needs to have answers to a half-dozenkey questions: - 48 -

(i) Can the crop be successfullygrown under Tongan smallholder conditions,what inputs are required,and what yields can be expected?

(ii) What pests and diseases can be expectedand are there feasible controls available?

(iii) Does the crop fit into and complementthe existing farming system, or are there short term costs to its adoption or a threat to long term sustainability?

(iv) Do we have a market for the crop, locallyor overseas,and what are the requirementsof the market, in terms of volumes required, seasonality,price, quality, quarantineregulations and packaging?

(v) On the basis of the above questions,how much can the .farmerearn from adoptingthe crop, and how does it comparewith alternative enterprises?

(vi) What specific forms of intervention,including extension, are requiredfrom the public sector to encouragethe crop, and what priority should be given to these in comparisonwith other competingclaims on governmentresources?

5.27 MAFF alreadyhas a suitablemechanism, in the form of regularmeetings of its heads of divisions,under the chairmanshipof the Director,to review proposalsand recommendfuture strategiesand policies. However, it needs much greater support from the PlanningDivision, serving as secretariateto the Director,in preparingpolicy papers for discussion,evaluating project proposalsand assistingtechnical divisions in formulatingprojects and programs. The present PlanningDivision in MAFF is unable to perform even the minimum planning functionsof the Ministry,and needs to be strengthenedas a matter of urgency. An elaborateplanning apparatusis not necessary: a few additionalexperienced staff would make all the difference. HAFF has recently secured some overseas staff, and expectsat least one Tongan to return from overseasgraduate training soon.

5.28 Longer-TermIssues for MAFF. We have tried to indicate some of the more urgent tasks that MAFF should gear up for. This is a challengingtask for a small ministry,which will impose strainson its managementand wili need support from donors. In the longer term, new challengeswill arise which MAFF needs to start thinking about now. These includepossible new programs in the environmentarising out of the proposedmonitoring activity, and a more systematicattempt to monitor land tenure and land use issues,particularly the movementsand possible impact of the terms of land leases.

5.29 In forestry there is a need to step up the rate of plantationon 'Eua, to lay more emphasis on agroforestryand to developmethods of managing and regeneratingnatural hardwood forests. This will requirea significant upgradingof the ForestryDivision and its financialresources. In the meantime,with a view to the long term conservationof the resource,some urgent actionsare to set aside more land for plantationson 'Eua, seek externalassistance to undertakean inventoryof hardwood forestsand to mount a pilot project to rehabilitatelogged areas. - 49 -

5.30 MAFF may also have to give greaterattention at some stage to three special groups within the wider farmingcommunitys

Vomen: women's activitiesin farmingshould be monitored and assisted,especially the productionof pandanus and mulberry,as well as crops such as vanilla, and other business activities. There may also have to be specificprograms to identifywomen with access to land and to help them utilize it efficiently.

Outer islandss there may be a need for special programs to monitor incomesand activitiesin the remoter islandsand to identifycrops suitablefor those areas (e.g. kava on volcanic islands).

Lower income grougs: a need may emerge in the future to monitor food suppliesof poor people and to design specialagricultural programs for peri-urbansquatters (especially if causing land degradation).

M4AFFmight seek external assistanceto look into some of these issues,while it concentrateson immediatepriorities.

5.31 Training of MAFF Staff. Trainingneeds of staff at the professionallevel are commonlymet throughoverseas training,either degree or short courses. The value of this overseastraining is extremelyvariable in terms of both its overall qualityand its suitabilityto Tongat typically the trainingcurricula cover neither the crops nor the approach requiredfor analyzingand understandingTongan farming systems. However, this is rapidly changing and farming systemsapproaches are more common in overseas universities.

5.32 Discussionswith MAFF suggest their greatestneed is in the areas of agriculturaleconomics (especiallyproject economics)and planning,marketing and systemsapproaches to agriculture. There are some staff being trained overseas at presentwho will be able to do some of this work, but will be inexperienced.Moreover, for a small organizationlike MAFF, the costs of having people under training overseascan be high because they may be the only people in MAPF with particularskills.

5.33 There is also a growing need for MAFF staff to advise and train farmersnot just on technicalsubjects but on how to manage their farms as a business. This will mean that MAFP staff will have to graduallymove from being productionoriented to being commerciallyoriented, and will need training in the relevantskills. This has been recommendedfor external support.

5.34 The Ton2a CommoditiesBoard. Given the continuinglosses on its operations,there is an urgent need to decide on the future of the TCB. In so far as its agriculturaloperations are concerned,several strategiesmight be considered:

(i) Given the poor prospectsfor its major coconut products,TCB should dispose of its oil mill and dessicatedcoconut factoryand handle only copra, since it would los, less money in this way. - so -

(ii) Governmentshould liberalizethe copra trade, either by complete decontrolor by licensingprivate traders. Some form of registrationand controlwould be desirableto ensure that copra exports meet quality standards--particularlyin regard to aflatoxin levels. If the private sector offers higher prices to the farmer, TCB's copra tradingcould be terminated. A remaining issue would be whether and how to maintain the present subsidyon copra marketing from the outer islands, since these could no longer be cross-subsidizedby TCB from copra bought in Tongatapu. If it is decided to continue paying a subsidizedprice, for social/regionaldevelopment reasons, then either TCB or licensed traderscould undertakethe marketing,on Government'saccount, on a contractbasis.

(iii) TCB might use its assets and experienceof Tongan agricultureas equity for a joint venturewith an overseasmarketing company to developmarkets for Tongan produce,using the marketing expertise and internationalconnections of the joint venture partner.

(iv) There is a great potentialvalue in developingdessicated coconut, coconut cream and other products processeddirectly from the whole coconut, providedthe manufacturingprocess uses less labor than the farm family has to use to produce copra. Governmentmight thereforecontract with TCB to explore alternativelow-cost, small-scaletechnologies, either in conjunctionwith a joint venture partner or with technicalassistance provided through officialchannels.

5.35 Data Collectionand Analysis. As indicatedearlier, data that are essentialfor sound decision-makingare often hard to find in Tonga. But, given the size of the country,this should not be a difficultproblem to put right. In general, there is no need for an elaboratedata system,which Tonga can ill afford. In many cases, simple and inexpensivedata will serve the purpose quite well. It is more importantthat availabledata (such as the excellentdata on farm economicsquoted in this report) should be used for policy and project analysis.

5.36 Some suggestionsfor improvingstatistical collection have been made in the report. These includedata that would emerge from farming systems researchand farm economic surveys. The report also suggests some simple improvementsin the data underlyingthe estimationof the national accounts (see Appendix).

C. IMPLICATIONSFOR DONORS

5.37 It is clear that Tonga does not lack for externalassistance; on the contrary,it has a very high level of assistanceon a per capita basis. The quality and relevanceof externally-assistedprojects has, however, not always been good. Project identification,design and appraisalfor so many small projects to the standardsrequired by many aid agencies is extremely difficult,given the fundingand time availableand the limitedcapacity of local institutions. Delays in gettingprojects approved are also considerable. There are also sometimesconflicts between aid agency staff, - 51 -

who want good projects,and their managements,who want to see the country program implementedon time.

5.38 There are no easy answers to these problems,some of which are common to all small countries. The view put forwardin this report is that the major responsibilityfor Improvingthe effectivenessof ODA must lie with the Governmentand, in the case of agriculture,with MAFF. However the donors could help to improvethe impact of their assistancein a number of wayst

Concentrateon Priorities. Some of the ODA projects reviewedby the mission are of very doubtfulvalue and tend to divert the limited capacity of governmentagencies from more urgent needs. As stated earlier, no externalaid is "free" to a country like Tonga. This report has tried to suggestsome of the principalpriorities for agriculture,and if these are endorsed by the Governmentit would be helpful if donorswere to support them.

Improve ProjectAppraisal. As suggestedearlier, some projects have been designedwith seeminglylittle knowledgeof the sector. Both the Governmentand donors must try to ensure that projects are relevantto Tongan conditionsand fully reflect the economic realitiesof farming.

SupportLarger Projects. MAFF is trying to oversee a large number of small projects (see para. 5.7). Each of these has its own monitoring and reportingrequirements, counterpart staff and sometimeseven specialproject managementarrangements. Tonga simply does not have the capacity to do this. Moreover,many of these projects are supportingisolated activities which bear little relationshipto the broader programs on which MAFF needs to concentrating. We would thereforeurge donorn to concentrateODA on fewer, but more carefully selectedand preparedprojects or programs,which can have more impact and would prove more cost effectivefor donors and Governmentalike.

Move Upstream in the ProjectCycle. In many cases, HAFF does not have the capacity to prepare broadlydefined programs. It would be thereforedesirable to invite a "lead,donor to adopt a whole program (such as the FSRE or post-harvestmanagement programs suggested earlier),and to provide a full servicecovering (i) assistancein the identificationand preparationof the program; (ii) sustainedfinancial assistanceover the full length of the program; (iii) technical assistancein areas where qualifiedTongans are not availableand (iv) support for institutionaldevelopment and training,to ensure that the program is sustainablewhen the foreignpersonnel are withdrawn.

Improve Training. As indicatedearlier, it is importantto ensure that overseas trainingmeets Tonga'sneeds. In some cases, the best approachwould be to provide a specialtraining program, specifically designed for Tonga within Tonga. In other cases it is on-the-job, side-by-sidecoaching that is most required,rather than formal training. For this purpose donors might try to providepersonnel with appropriatetraining experience--not just technicalknowledge--to help develop local skills. - 52 -

Consultants'Fund. Tonga, a very small country, frequentlyneeds specialistsfor various purposes,especially for project preparation (technicaland economic aspects),marketing studies and for preliminary technicalscreening of possibleprojects. It will always face this problem becauseof the narrownessof its labor market. Requests are often made to overseasagencies for assistance,but these involve a lot of administrationand, after a long wait for a decision,may result in refusal. There is a strong case for an externallyfunded consultancy fund which MAFF could use to hire various experts. This would not only provide a quick-responsefacility, but allow MAFF to hire overseas specialistsfor short as well as longer pericds. For example, the preparationof a report on the market prospects for a certain product in Australiamight requirehiring a person in Australia for a week. - 53 - APPENDIX Page 1

AGRICULTURALVALUE-ADDED AND EXPORTS

Introduction

1. This review uncoveredmuch evidence to show that the official estimatesof GDP underestimatethe contributionmade by agricultureto the economy. The estimatesof value-addedin the agriculturesector are complied from separate estimatesof three distinct categoriesof production: productionfor home consumption(sometimes called "subsistence'production), productionfor the domesticmarket, and exports. The mission concludedthat all three categorieshave been underestimatedin the past. If correct decisionsare to be made about Tonga future developmentprospects and resource allocations,it is importantto understandthe real contributionthat the sector makes to Tonga's GDP and foreignexchange earnings.

2. Tonga is now revisingits nationalaccounts estimates,with assistancefrom the Asian DevelopmentBank, but unfortunatelythe reviseddata were not availableat the time of the review. The mission thereforemade its own rough estimatesof value-added,in each of the above categoriesof production. These are not presentedas reliableestimates, but as indicators of likely orders of magnitudepending the productionof reviseddata.

Estimatesof Production

3. Rome Consumption. The estimatesof home consumptionwere derived from data on the number of familiesrecorded in the 1986 census, the estimated proportionof food consumptionthan is produced at home, and the market value of food consumed.The estimatesinclude non-food products,such as housing materialsand tapa, produced by the family,as well as the value of pigs and other food consumedat ceremoniesand feasts. This ceremonialconsumption is quite significant,and helps to boost the mission'sestimates over previously availabledata. The methodologyused in making these estimates is given in Attachment1.

4. DomesticallyMarketed Production. Availablemarket statistics clearly understatedomestic sales, because they only includeproduce sold within market premises,not that which cannot enter the market and is sold outside on the street,which now accounts for as much as half the volume sold on market days. Nor do the statisticsinclude the increasingamount of road- side sales throughoutthe country.

5. While it is clear that the recordeddata understatesales, it is difficultto estimate the size of the market accuratelyin the absence of more detailed field surveys. The mission thereforeroughly estimated domestic food sales at T$15 million in 1988, or somewhat less than the level of imported foods. Tongans interviewedby the mission felt that this was the correct order of magnitude,but this is clearly one componentof productionfor which better data are badly needed, and could be readilyobtained. - 54 _ APPENDIX Page 2

6. Exports. The value of exports is derived from official figures, with the major exceptionof the export of unprocessedagricultural products, particularlytraditional rootcrops, which have been greatlyunder-recorded in the official figures. The mission thereforemade a special study of this trade (see Attachment2) and adjusted the official figures upwards accordingly.

7. Total Production. The combinedestimates for the three categories of productionprovide the followingestimates of total productionand exports:

Table l: ESTIHATEDVALUE OF AGRICULTURALPRODUCTION AND EXPORTS T$ million at 1988 prices

1984 1988

Productionfor home consumption 31.5 32.3 Domestic sales 10.3 15.0 Exportss coconut products 9.0 1.6 bananas 1.4 0.7 vanilla 1.9 1.5 ethnic niche markets 1.5 3.1 other niche markets 2.7 2.5 16.5 9.4

Total Value of Productioa 58.3 56.7

Estimatesof Value-Added

8. Value-addedwas derived from the above estimatesof productionby deductingan estimate of purchasedinputs. While input use obviouslyvaries quite considerably,depending upon the techniquesused by the individual farmer, the main contrast is between the low-inputtraditional technology used for most staple food crops, for both home consumptionand sale, and the relativelyhigh inputs (particularlychemical use) required for commercial crops grown for export.

9. The followingallowances for the cost of inputs were made in convertingthe value of productionto value-added:

Productionfor home consumption 2 1/2Z Domestic sales (allowingfor transport) 102 Exportss Bananas/vanilla 402 Ethnic niche markets 102 Other niche markets 202 - 55 - APPENDIX Page 3

10. Using these conversionfactors, the followingestimates were made of the contributionof agricultureto GDPs

EstimatedValue-Added (T$ millions,in 1988 prices)

1984 1988

Productionfor home consumption 30.6 31.5 Domestic Sales 9.3 13.5 Exports 13.5 7.5

Total 53.4 52.5

As discussedin the main report, these rough estimatessuggest that agriculture'scontribution to the economy is at least 25 percentmore than impliedby the existing publisheddata. - 56 - Attachment 1 Page 4

ESTIMATIONOF TONGANHOUSEHOLD CONSUMPTION

1. A 1986 study surveyed 4 villages on Tongatapuover a 12 month period and collecteddata on consumptionper adult equivalentof Tongan stapfes from householdproduction. The results are set out in Table 1. Table 1: Consumptionof Staples from8 Villge.,1988 (Householdproduction)

Consumption,kg/week Der adulteuivalent Valueof con- Farm sumption pe VIllal __gote t dult-eauivalent Price Staple 1 2 a man TI/Kg TI/weekTI/year

YaM 1.01 1.18 1.68 1.91 0.90 1.72 19.28 Toro 1.04 0.97 1.69 1.28 0.80 0.87 19.24 Kape 1.82 6.69 8.89 8.98 0.27 1.08 66.22 Ca-sva 9.78 2.70 10.57 7.67 0.16 1.16 69.80 Kumals 0.21 0.08 0.02 0.09 0.80 0.08 1.85 Bamane/Plantin 0.58 0.70 8.84 1.69 0.26 0.42 21.97 Breadfruit 2.10 5.04 2.27 8.87 0.16 0.51 20.29 Otherlocal stples 0.07 0.01 - 0.08 0.20 0.01 0.28 T7.21 19.17 28.86 19.92 6.26 278.38

SourcesDelfore .JC. FoodCroD Production in Tong": CharactXersticsand Activity Sudast SouthPacific Smalholder Project: University of MmeEnland 1998. Notes The survey included one village with consumption of all food items approximatelyhalf the others. This data has been omitted.

2. For Tonga, the ratio of adult equivalentsto the total populationis approximately0.75. The total of T$273/yearper adult equivalentis equivalentto T$218 per year. However this figure does not include livestock, (poultry,pigs, chickens),coconuts, fruit, housingmaterials, mulberry, kava, pandanus,dyes and other agriculturalitems normally consumedby the household. Estimatingthis at T$60lhead/yeargives a total consumptionof T$278. Adjusting for a rise in prices of 12 percent to 1988 gives an estimate of T$312 in 1988 prices.

3. However, this is an estimatefor rural dwellers only. Although urban dwellers grow and consumemuch of their own food or get it as gifts from relatives,the average for Tongatapuwould be less. The estimatewas thereforeadjusted back to T$280/head. This amounts to about $1,400 per family per year.

4. Similardata were not availablefor other islands. Some impression of differences in consumption patterns was obtained from interviews and discussions with agricultural workers. For example, for Vava'u it was estimated that household consumption from own production was 30 percent higher overall after allowing,for lower urbanizationand less purchasedfood. However, prices in Vava'u were estimatedat 85 percent of Tongatapuprices, and the unit value of that food was thereforeless. A correctionfactor of - 57 - Attachment1 Page 5

1.11 (1.3 x 0.85) is used to estimatehousehold consumption per head for Vava'u at T$31}.

5. Populationestimates for 1988 were based on 1986 data, with a 0.6 percent growth rate since. Table 2 sets out estimatesfor household consumptionbased on these assumptions.

Table 2: HouseholdConsumption Estimates for Tonga, 1989

Population correctionfactors Production 1988 volume price overall Slhead Total $1,000

Tongatapu 64,400 1.00 1.00 1.00 280 18,000 Vava'u 15,300 1.30 0.85 1.11 311 4,800 Ha'apai 19,100 1.50 0.60 0.90 252 2,300 Eua 4,400 1.30 0.80 1.04 291 1,300 Niuas 2,400 1.70 0.50 0.85 238 600

95,600 27,000

6. These estimateswere checked by asking severalhouseholds on Tongatapu,RI'apai and Vava'u how much it would cost to purchase food normally produced and consumed in the householdfor one week. The sample was small, but answerswere reasonablyconsista'ttt with these estimates.

7. This figure does not i - fish consumption. Fisheriesofficials suggestedT$1.00/week per adult equivalentaverage for the whole countryof marine products consumed by catchers. This would amount to an additional T$4.1 million, giving a total of T$31.1 million for householdconsumption from the agricultural,forest and fisheriessectors.

8. The 1989 AgricultureSurvey should assist in the productionof more reliabledata, but does not include livestock,marine products,housing materials,pandanus, mulberry, dyes and some other commonlyconsumed agriculturalproducts such as fruit. - 58 - Attachment2 Page 6

ESTIMATESOF UNPROCESSEDAGRICULTURAL EXPORTS

1. Table 1 shows the volume and value of unprocessedagricultural exports,according to official sources. The data covers

* vegetablesand fruit for the Pacificoverseas markets whole and green coconuts,taro, cassava,yams, plantain,breadfruit, taro leaves, limes and other fruit and nuts. These are exported to Pac- ific Islanders(mainly Tongans) living overseas in New Zealand, Australia,USA, Hawaii and to other Pacific Island countries; American , Western Samoa and . This trade is dominatedby private exporters.

* fruit and vegetablesfor the generaloverseas markets bananas, melon, passion fruit juice and fruit pulp, pineapplesand pumpkins. Except for pumpkins to Japan, these are mainly exported to New Zealand and Australiaby the CommoditiesBoard; there is a small volume of private exports.

2. The quantity and value of private exports is considerablyunder- stated in the official statisticsfor a number of reasons:

* Farmersunderstate the consignment'svalue since Customs documents are availableto the tax authorities;

* Some exports are classifiedas passengerseffects or as unaccom- panied baggage, oft,n specified'no commercialvalue';

- To get around these problems Customsuse a standard f.o.b. value which is usually considerablybelow actual f.o.b. value;

* Not all produce exports are reportedby Customs to the Statistics Department,and some produce is shippedwithout documents,especial- ly by air.

3. The use of standard fob values is clear from Table 1. Unit values per kilogram tend towards 50C for bananas,25c for cassava,T$1.00 for yam, T$l.OO for kape and plantain and so on. The standardprices do not take ac- count of price changes. In April 1989, a New Zealandbuyer was buying (in large quantities)yams for T$1.50kg (standardvalue T$1.00) and cassava for T$0.80/kg (standardvalue T$0.25). A full containerof mixed root crops has a value of between T$8,000 and T$10,000. - 59 - Attachment2 Page 7

Table 1: UNPROCESSEDAGRICULTURAL EXPORTS 1908 (k IIograma)

New Zealand Australia Amr- Wast- FijiUSA Hawn'lOther Total Volueo lean *rn Prvate Quasi- Private qusil- SamoaSamoa (USA) cent Govt / aovtLo kg To /kg

Bananas 772 1815328 1816095 659862 0.60 Cassava 698862 27604 17 862 720 100 722191 157654 0.22 Coconut, whole 104849 167828 6326 75068 28 185 878777 926o9 0.25 Ya, Sweet potato 187594 61104 8062 1870 66 107 1841 12229 267858 268879 0.98 Taro tarua 72848 85810 10006 76 800 804 120889 99?80 0.88 MixedLe 61661 66 7140 607 1800 6856 8 6602 47626 0.72 Kape 42669 7867 241 128 481 50858 50640 1.00 melons 40 15844 4864 74 447 882 21101 19169 0.91 Fruitjulce 17600 17600 28480 1.68 Coconut,sree 260 1589S 268 16406 8981 0.24 Passionfruit 10120 16120 24060 1.49 Swamptaro 18870 1672 426 SO 60 18 16495 12988 0.88 Plantain 8565 10822 8a 72 20 94 267 14868 14486 0.99 Fruitpuree /d 6600 0650 9760 1.60 Breadfrult 2161 181 446 476 8214 1620 0.60 Taro leaves 1204 11 15 24 1254 1617 1.21 Pineapples 761 761 600 0.66 mangoesA Ilmes 98 20 4 21 748 788 0.99 Ginger 178 a88 511 618 1.01 FruitsI nuts 480 466 218 0.46 TomatoM. 151 12 168 281 1.60 Capslcum i6 76 228 8.00 Carrots 64 54 106 2.00

Total 1216982 1602797 76006 76118 18888 4489 161 6478 21658 201 80222861478960 0.49 Country totals 2616779 158121

/a MainlyTongs Commodities Board ffb Standardfob values are used, see xplanationin text. c Mixedconsignments, minly rootcrops. /d Technicallya processed commodity. Note: does not lncludepumphins, see text.A few very minor exports are omitted. Source: AnnualForeign Trade Report for 1988, Statistics DepartmentGovernment of Tonga

4. From discussionswith exporters,shipping companies and the Bank of Tonga, the mission estimatedan average of 13 containersper month from Ton- gatapu to New Zealand and 7 per month to Australia (based on the period May - October 1989). Exports from Vava'u were estimatedat 8 per month over the same period. This totals 336 containersper year with an estimatedvalue of T$2.7 million to T$3.4 million. Assuming non-containerexports amount to 10? of the total, the estimatesare T$3.0 to T$3.7 millionwhich, when pumpkins are added (whichwere omitted from the figuresand amounted to about T$400,000)brings the estimate to T$3.4 to T$4.1 million comparedwith the official figure of T$1.5 million for 1988 (Table1).

5. Private exports of vegetableshave developedalong a path which demonstratesthe strengthand importanceof the Tongan social system as well as Tongan entrepreneurship.Current exportsare a mixture of all the follow- ing methods, but moving over time from (a) to (d) New Zealand, the major - 60 - Attachment2 Page 8 market, is used as the example;a smallervolume is marketed similarlyto Australia.

(a) Farmers send sacks of vegetablesto relativesin New Zealandby sea and sometimesby air, especiallywith family going to New Zealand. This produce is sold but proceedsare very difficultto get back from relatives;nevertheless, the transactionjoins the complex fab- ric of obligationsand will make it easier for the farmer to get a future church donation,a trip to New Zealandor other assistance from overseas relatives.

(b) A farmer sends a full containerto relatives/friendsin New Zealand, who sell the produce and remit some of the proceedsback to Tonga. Farmers complain about the extreme difficultyof gettingall or even some of this money.

(c) A full containeris consignedby a farmer 1/ who flies to New Zealand to meet it, sell the producemainly on credit then collect the money over 2 or 3 months. The farmer enjoys the additional benefit of a period with relativesin New Zealandbut has to bear the cost of staying in New Zealand This is currentlythe most com- mon method, but some farmersare realizingthe additionalcost of being away from the farm.

(d) A buyer from Tonga or New Zealandbuys the produce and consignsa containerto a New Zealandagent or flies to New Zealand to market the producehimself. Increasingly,this is done by contractingwith one or more growers to provide the produce at an agreed price. In Auckland there is a network of sellerswhich is becoming formalized to the point where there are professionalretail and wholesale Tongan marketers.

6. Financecan be a problem,especially on Tongatapuwhere the Bank of Tonga does not lend for food crop export financeand where local buyers commonly cannot pay cash. Vava'u has a larger number of buyers than Tongatapu and the Bank of Tonga is very active in financingexports. This system, in which a surprisinglylarge volume of growers participate,has developedwith little governmentassistance except for produce inspection.

1/ A group of farmersmay also fill a container. - 61 - Attachment 3 Page 9

PROPOSALSFOR IMPROVINGSTATISTIC DATA COLLECTION

1. The implicationsfor agriculturalpolicy are considerable. The agriculturalsector is clearlycontributing a higher proportionof GDP than official figures suggest,and will continueto do so. There is clearly a need to improve the current system through realisticvaluations and reductionsin undereporting. The mission discussedthis with the Departmentof Statistics (DOS) and MAFF and suggestedthat weights of producebe collectedat the time of produce inspectionby HAFF staff,.which then goes to DOS for collation. MAFF then providesan f.o.b. price estimate for each commodityfrom prices paid by exporters,adjusted to an f.o.b. price. This would requireno addi- tional resourcesand be more reliable than the current system.

2. Some suggestionsfor improvingstatistical collection are:

For Exports:

* MAFF should make monthly estimatesof f.o.b. values of all major unprocessedexport crops through discussionswith exportersand farmers. (This should be a functionof the planned Marketing Sectionof HAFF);

* MAFF produce inspectorsenter the weight of produce they inspect for export by sea or air on a separatepart of their inspectionrecord form which is sent to the StatisticsDepartment;

* the StatisticsDepartment collates the volume data on a monthly basis asd applies the monthly f.o.b. values suppliedby MAEF to get an estimateof total monthly exports;

* seek the cooperationof shippingagents to provide regular informationon number and destinationof containersof agricultural exports from Nuku'alofaand Vava'u.

For domesticmarkets:

* The presentmarket survey method be redesignedand a rigorous samplingmethod used to estimateprices and volumes flowing through the markets at Nuku'alofaand Vava'u;

* from time to time small surveys should be made of roadsideand other sellers to assess volumes and trends of prices and sales. Estimates should be made of this volume.

These changeswould delink the export statisticsfrom the income tax system and provide sufficientlyaccurate informationat little extra cost. - 62 -

BIBLIOGRAPHY

Abbot, D. (1986),Development Planning in Tonga: the 5th DevelopmentPlan and RegionalPlanning Seminar 14 - 18 July 1986. CentralPlanning Department, Tonga.

ACIAR, (1988), PotentialAustralian Market for Coconut and Coconut Products, Working Paper No. 22.

Ahlburg, Dennis A. (1986),Population and Economic Developmentin the Island Nationsof the South Pacific,Islands/Australia Working Paper, No. 86/7, National Centre for DevelopmentStudies, The AustralianNational University, Canberra.

Ahlburg,D. and Leven, M. J. (1990),The North East Passage:A Study of Pacific Island Migrationto and the United States. National Centre for DevelopmentStudies, Australian National University, Pacific Research Monograph No. 23.

Asian Development Bank (1979), Pacific Island Choices: Rural Limits and Opportunities.South PacificAgricultural Survey, February - June 1979, Manila, Philippines.

Browne, C. (1989), Economic Development in Seven Pacific Island Countries InternationalMonetary Fund.

Byron, R. N. (1986), Policies and Options for the Forestry Sector of South Pacific Island Economies,Islsnd/Australia Working Paper, No. 9;6/4,NCDS.

Byron, R. N. (1988),'Forestry and Fisheriesin the Asian-PacificRegion: Issues in Natural ResourceManagement', Asian-Pacific Economic Literature,Vol. 2, No. 1: pp 46 - 80.

Connell,J. (1987),Migration, Employment and Development'n the South Pacific, South PacificCommission, Noumea, New Caledonia.

Deacon Ritterbush,S. (1986),Entrepreneurship and BusinessVenture Development in the Kingdom of Tonga. PIDP (PacificIslands Development Program), East-West Center,Honolulu, Hawaii.

Dean, M. (1988),N.Z.M.F.A. Coconut ReplantingScheme Review,Kingdom of Tonga. AGRICO, N. Z.

Dean, M. D. K. and Felemi, I. M. N. (1981),Vanilla DevelopmentPlan for All Island Groups Outside Vava'u, Ministry of Agriculture,Fisheries and Forests (MAFF),Nuku'alofa.

Delforce, J. C. (1987), Food Crop Production in Tonga: Characteristicsand ActivityBudgets, Research Note 4, South PacificSmallholder Project, University of New England,Armidale, Australia.

EconomistIntelligence Unit (EIU) (1988),EIU CountryReport, Pacific Islands: ,Fiii, SolomonIslands, Western Samoa, Vanuatu, Tonga, No. 4; The EconomistIntelligence Unit 1988 - 89, London,United Kingdom. - 63 -

Engleberger,K. (1986),Vegetable Production in Tonga, Ministry of Agriculture Fisheriesand Forests, TechnicalBulletin 6, Nuku'alofa.

F. V. Sevele (1984),A Shadow Pricing System for the Kin8dom on Tonga. South PacificCommission.

FAO (1982), Study of the Coconut Industry Tonga: Mission Findings and Recommendations.Report prepared for the Governmentof Tonga, Rome.

FAO (1987),Formulation Mission - CoconutReplanting, Tonga.

FAO (1986), The World Banana Economy 1970 - 1984, FAO Economic and Social DevelopmentPaper 57, 1986.

Fa'anunu,B. 0. (1981), The Economicsof Export Vanilla Productionin Tonga, MAFF, Nuku'alofa.

Fakalata, O., Tu'ipulotu,T. Manisela, M. (1988),A Study of the effects of differentcrop combinationson the cabbagecaterpillar (Plutella xylostella and Crocidolomiabinotalis) during the warm months,Ministry of AgricultureFisheries and Forests, Nukulalofa.

Falvey, R. E. (1986), Export Instabilityand the Pacific Island Economies, Island/AustraliaWorking Paper No. 8611, NationalCentre for DevelopmentStudies, The AustralianNational University,Canberra.

Fia, S. (1989), WatermelonFertiliser Trial," in Gyles, A., Speijer,P. R., and Kami, V. (EDS), Proceedingsfor a Workshop on On-Farm Trials in Tonga, South Pacific SmallholderProject, University of New England,Armidale.

Fleming,E. M. (1986),Agricultural Market Developmentin the South Pacific,PnD Thesis, Universityof New England,Armidale.

Fleming,E. M. (1986),A Study of AgriculturalMarket Developmentin the South Pacific Region: Structure,Conduct and Performance,Occasional Paper 3, South Pacific SmallholderProject, University of New England,Armidale, Australia.

Fleming, E. M. (1986),Analysis of AgriculturalMarketing Performance in Five South PacificCountries, Occasional Paper 6, South Pacific SmallholderProject, Universityof New England, Armidale,Australia.

Fleming, E. M. (1988),Agricultural Marketing and Price Policy: Choosing the Correct Incentives. Paper presented at Workshop in SmallholderAgricultural Developmentin Tonga.

Fleming,E. M. and Hardaker,B. (1986),Agricultural Supply Response in the South Pacific Region, Island/AustraliaWorking Paper No. 86/16, National Centre for DevelopmentStudies, Australian National University, Canberra.

Forests Act, The Law of Tonga. Volume II, 1967, 1961.

Forsyth, P. (1986),Economic Problems of InternationalTransport for the South Pacific Island Economics,Island/Australian Working Paper No, 86/10, National Centre for DevelopmentStudies, Australian National University,Canberra. - 64 -

Fortech (1982),Fuelvood Study - Tonga. Prepared for the South PacificBureau of Economic Co-operation(SPEC), Suva, Fiji.

Gannicott,K. and Throsby, C. D. (1988),The Quality of Education in the South Pacific:Some PreliminaryHypotheses, National Centre for DevelopmentStudies, AustralianNational University, Canberra.

Gyles, A., Petelo and Petelo, Ha'unga (1988),Kingdom of Tonga: Proceedingsof the Workshop on Farming Systems Research and Extension, 9 - 10 May 1988, Universityof New England,Armidale, Australia.

Gyles, A., Hardaker,J. B., Felemi,M., and Verspay,H. M. H., Farm Management Handbook for Tonga, Planning Division,Ministry of Agriculture,Fisheries and Forests, TechnicalBulletin No. 8, July 1989.

Gyles, A., Sefania, S., Fleming, E. M., and Hardaker, J. B. (1988), Farm ManagementHandbook for Tonga, Ministry of Agriculture,Fisheries and Forests, South Pacific Smallholder Project, University of New England, Armidale, Australia.

Hakohako,T. (1988),Livestock Feed Utilizationin Tonga, Paper presentedat the CTA RegionalLivestock Workshop on EffectiveUse of Local Feed Resourcesin the Feedingof Pigs and Poultry in the PacificRegion, College of Agriculture,Fiji, September5 - 9.

Hardaker, B. and Fleming, E. (1986), Policy Issues in AgriculturalMarket Developmentin the South Pacific Region, Island/AustraliaWorking Papar No. 86117, NationalCentre for DevelopmentStudies, Australian National University, Canberra.

Hardaker,B., Delforce,J. C., Fleming,M., and Sefanaia,S. (1988),Smallholder Agriculturein Tonga. Report of the Couth PacificSmallholder Project in Tonga, 1984 - 85, 2nd Edition, Universityof New England, Armidale,Australia.

Hardaker,J. B. and Fleming,E. M. (1986),The SouthPacific Smallholder Project: Outline,Occasional Paper 2, South PacificSmallholder Project, University of New England,Armindale, Australia.

Hardaker,J. B. Second CoconutUtilization Survey, 1973.

Hardaker, J. B., Delforce, J. C., Fleming, E. M., and Sefanaia, S. (1987), SmallholderAgriculture in Tonga:Report of the SouthPacific Smal-holder Project in Tonga 1964 - 85, South PacificSmallholder Project, University cf New England, Armidale (Revisededition 1988).

Hassall& Associates(1983), Review of the CocoaRehabilitation Project - Western Samoa. Report prepared for the Governmentof Western Samoa and the Australian DevelopmentAssistance Bureau, Canberra.

Hassall & Associates(1988), Western Samoa Cocoa Rehabilitationand Development Project: First Report on Cocoa Processingand Marketing,Canbarra.

Holo, T., Latu, P., and Vi, P. (1979), Watermelon Production in TonRs, unpublishedreport, Ministry of Agriculture,Fisheries and Forests,Nuku'alofa. - 65 -

Isnar (1983),The Planningand Managementof AgriculturalResources in the South Pacific. Report of the Workshop of 5 - 16 October1987, IRETA, Alafua,Western Samoa, .

Kingdom of Tonga (1987),Fifth Five-YearDevelopment Plan 1986 - 1990, Central PlanningDepartment, Nuku'alofa.

Kingdom of Tonga, The Law of Tonga 1976 Acts, 1977.

Lamont, J. (1982),Development Bank Credit in Rural Areas of the Pacific; the Implicationsfor Rural Development. Proceedingsof a Workshop held in Tonga, July 1982. Universityof the South Pacific,Rural DevelopmentCentre, Tonga.

Larsen,A. and Upcott, A. (1982),A Study of the Forest Resourcesof 'eau Island and Their Potential for Future Management, Chandler, Frowzier and Larsen, Rotorua, New Zealand.

Latu'ila, A. (1989), Irish Potato Variety Trial, in A. Gyles, P. R. Speijerand V. Kami (EDS), Proceedings of a Workshop on On-Farm Trials in Tonga, South Pacific SmallholderProject, University of New England, Armidale.

Lisiate 'A 'Akolo (1989),The CommoditiesBoard - A RestructuringProposal.

MAFF (1982),A Review of the CoconutReplanting Scheme and Aspects of the Coconut IndustryRelated to Coconut Production,Planning Unit TechnicalPublication No. 1/82, Nuku'alofa.

MAPF (1988),Export Banana Productionin Tonga, TechnicalBulletin 7, Ministry of Agriculture,Fisheries and Forest,Nuku'alofa.

MAFF (1988),Administrative Structure Review.

Ministry of Agriculture,Fisheries and Forests and German Plant Protection Project Root Crop Productionin Tonga, (1989).

TechnicalBulletin No. 10 June. Menz, K. H. and Fleming,E. H. (1989),Economic Prospectsfor Vanilla in the SouthPacific, Technical Reports No. 11, Australian Centre for InternationalAgricultural Research, Canberra.

Menz, K. W. (1988),Smallholder Agricultural Development in Tonga: Proceedings of a Workshop,Institute for Rural Development,University of the South Pacific, Tonga, 12 - 13 May 1988. ACIAR ProceedingsNo. 24.

Ministry of Agricultureand Lands (1985),Coconut Development Project Phase II - ProJect PreparationStudy.

Ministry of Agriculture,Forestry and Fisheries (1982),A Review of the Coconut ReplantingScheme and Aspects of the IndustryRelated to Coconut Productionin the Kingdom of Tonga, TechnicalPublication No. 1182.

Opio, F. and Sui, S. (1988), Feasibility and Organization of Commercial - 66 -

Productionof Pias and PoultryProduction in the Kingdomof Tonga,Mission Report to the Ministry of Agriculture'sLivestock Sector, FAOIUNDP Regional Livestock DevelopmentProject Number RAS1861039,Tonga.

Perry, L. J. (1986),Labour Markets in the South Pacific:An overview of Fiji, SolomonIslands, Tonga, Vanuatu and WesternSamoa, Island/Australia Working Paper No. 861ll, National Centre for Development Studies, Australian National University,Canberra.

Persley, G. J. and Perrar,P. (1987),South PacificAgriculturet Challenges and Opportunitiesfor ACIAR and its Research Partners,ACIAR TechnicalReports 5, ACIAR, Canberra.

Pone, S. P. and Pearson, M. N. (n.d.), Virus Diseases of Vanilla in Tonga, Ministry of Agriculture,Fisheries and Forests,Nuku'alofa, n.d.

Poole, A. L. (1970), Forest Potential and Policy in Tonga. The New Zealand Forest Service,Wellington, N. Z., 1970.

Potter, L. M. (1986), Tongan Soils: Site Characteristicsand Management Practices,Occasional Paper 7, SouthPacific SmallholderProject, University of New England, Armidale,Australia.

Rathey, R. (1984), Final Report of the AgriculturalEconomist, July 1981 - January 1984, Nuku'alofa,Government Printer.

Revington,B. (1989), 'EuaForest Farm ManagementPlan 1989 - 1994, Ministry of Forestry,Wellington, N. Z.

Ritterbush, S. Deacon (1985),Business Venture Development in the Kingdom of Tonga. Pacific IslandsDevelopment Program.

Rondinelli,D. A. (1987),Development Administration and U.S. ForeignAid Policy, Lyme Rienner, Boulder.

Rondinelli,D. A. (1983),Development Projects as PolicyExperiments: an Adaptive Approach to DevelopmentAdministration, Methuen, New York.

Shaw, B. (1983), Food and Nutrition Policies for South Pacific Countries: Determinantsof GovernmentPlanning, In Thaman and Clarke (1983'.

Shaw, B. (1982),"Smallness, Islandness, Remoteness and Resources:An Analytical Framework,'Regional Development Dialogue, Special Issue 1982,Nagoya, Japan, pp 95 - 109.

Speijer, P. and Smit, N. (1989),Report on Pumpkin Production,Monitoring and Trials in Tonga,Ministry of Agriculture,Fisheries and Forests/GTZ,Nuku'alofa.

Speijer,P. R., Holo, T. F., and Takau, I. (1989),Pumpkin Production for Export in Tonga, 1989, TechnicalBulletin No. 9, Ministryof Agriculture,Fisheries and Forests and Tongan German Plant ProtectionProject, Nuku'alofa.

StatisticsDepartment (1988), Tonga AgriculturalCensus 1985, Nuku'alofa. - 67 -

Sykes,W. R. (1977),The Pteridophytesof 'Eua, SouthernTonga, The Royal Society of New Zealand, Bulletin17.

Taufatofua,P., Pole, F. S., and Smit, N. (1986), Screeningand Breeding for Sweet Potato Scab (ElsinoeBatatas) Resistance in Tonga, Paper prepared for the UNDP/FAO/GTZ/IRETARegional Workshop, Apia, Western Samoa, 8 - 11 September.

Thaman, R. R. (1976),The Tongan AgriculturalSystem; vith SpecialEmphasis on Plant Assemblages. Thesis,University of the South Pacific, Suva, Fiji.

Thaman, R. R. and Clarke, W. C. (1983),Food and National Developmentin the South Pacific, Universityof the South Pacific, Suva, Fiji, pp 25 - 48.

Thaman,R. R. Agrodeforestationand AgriculturalDevelopment; the Role of Modern Agricultural Development in Deforestation and the Neglect of Trees,' in InternationalConference on AgriculturalDevelopment in the Pacific Islands in the 90's. The Universityof the South Pacific, Suva, 1989.

Thompson,A. P. (1976),Forestry Liaison Visit to the Kingdom of Tonga. New Zealand Forest Service,Wellington.

Thorpe,T. and Cox, 0. (1988),Report on a ForestryLiaison Visit to the Kingdom of Tonga. Ministry of Forestry,New Zealand.

Thorpe, T. (1989),Report on a ForestryLiaison Visit to the Kingdom of Tonga. Ministry of Forestry, New Zealand.

Tonga (1985),Ministerial Statement on Government Policies for Investment in Tonga, Nuku'alofa.

Tonga,Central Planning Department (1987), Fifth Five-YearDevelopment Plan 1986 - 1990, Nukulalofa.

Tonga, Central Planning Department (1988), Ha'apai 1988195 Regional Plan, Nuku'alofa.

Tonga, Department of Statistics (1985), Tonga Agricultural Census 1985, Nuku'alofa.

Tonga,National Food and NutritionCommittee (1987), The 1986 NationalNutrition Survey of the Kingdom of Tonga. SummaryReport prepared for the National Food and NutritionCommittee, Nuku'alofa.

Tu'itupou,F. (1988),Broiler Trial, in Gyles, A., Speijer,P. R., and Kami, V. (EDS), Proceedingsof a Workshop on On-Farm Trials in Ton4a, South Pacific SmallholderProject, University of New England,Armidale.

UNDP/IWorldBank (1985),Tonga: Issues and Options in the Energy Sector, Report No. 5498 - Tonga.

Vinning,G. (1989),Growth, Production and Distributionof Species,Working Paper No. 27, AustralianCentre for InternationalAgricultural Research, Canberra.

Wall, D. (1986), Coconut Processing in the Pacific Islands, Island/Australia - 68 -

Working Paper No. 86/18. National Centre for DevelopmentStudies, Australian NationalUniversity, Canberra.

Ward, G. R. et. al. (1979),South PacificAgriculture: Choices and Constraints, South Pacific Agricultu'al*urvey.

Waugh, G. (1986),The Developmentof fisheriesin the South PacificRegion with Reference to Fiji, Solomon Islands, Vanuatu, Western Samoa and Tonga, Island/AustraliaWorking Paper, No. 86/2, National Centre for Development Studies,Australian National University, Canberra.

Wijmeersch,P. van (1986),Root Crops Productionin Tonga, Regional Projectby the Governments of the South Pacific Island Countries, Field Document 13, FAOISPC, Suva.

Goering, T. James (1979),Tropical Root Crops and Rural Development,World Bank Working Paper No. 324.

World Bank (1986),The Kingdomof Tonga: An IntroductoryEconomic Report, Report No. 6379 - Tonga. TONGA

- bin Roods

ec-~~~~~~~d

1 B z t75I0 f 5,

Vf we &meA

GROUP TONGATAPU

TONGATAPU Ewa GROUP o

05 10 "IN

0 10 20 Kda~~~~flmesers 175-0'O A

,tUVaws' Mc,4 f* HAPAI

-6 VAuoh ~GROUP N__

o, s lopmg" Ma r_ _n_ d o00~'j~g~ /lb 20 IG_n 176 1 1@ 1 1@~~~~~~~~~~~~~~~~7

-15U VAGROW W. UPOAA

-21° 21 Now Cl Fr.

00 250 500 W t K_c IL N.Z 30°

g00 m , TONGATAPU ½. "aNGA GROUP NEWZEALAND

1760 17,57K.5 IF 1650 l800