Public Disclosure Authorized

DISCUSSIONPAPER

SouthAfrican Agriculture:

Public Disclosure Authorized Structure,Performance and Options for the Future Public Disclosure Authorized Public Disclosure Authorized THE WORLDBANK INFORMALDISCUSSION PAPERS SOUTHERNAFRICA ON ASPECTSOF THE DEPARTMENT ECONOMYOF SOUTHAFRICA PREVIOUSWORLD BANK PAPERS ON SOUTHAFRICA

Previouslypublished in the World Bankseries of informaldiscussion papers on SouthAfrica:

Levy,B. January1992. "HowCan SouthAfrican Manufacturing Efficiently Create Employment?An Analysisof the Impactof Tradeand IndustrialPolicy".

Kahn,B., Abdel,S. and Walton,M. May 1992."South Africa: Macroeconomic Issues for the Transition".

Fallon,P. October1992. "An Analysisof Employmentand WageBehavior in SouthAfrica".

SouthemAfrica Department,May 1993. "An EconomicPerspective of SouthAfrica".

Belli, R, Finger,F., BallivianA., August1993. "SouthAfrica: A Reviewof Trade Policies".

Riley,T., November1993. "Characteristicsof and ConstraintsFacing Black Businesses in SouthAfrica: Survey Results'.

SouthemAfrica Department,November 1993. "SouthAfrica: Paths to EconomicGrowth" (Digest).

In addition,a numberof technicaland seminarpapers prepared by WorldBank staff and SouthAfrican counterparts in key sectorshave been discussedin the country. OREWORD lbis discussionpaper is the sixth is a series of Informalstudies on differentaspects of the SouthAfrican economyprepared by staff membersof the World Bank and SouthAfrican associates. The studieshave been discussed with a wide range of South Africans and have benefited from South African ideas and inputs. The Bank's study program aims to support the ongoing effort by South Africans across the institutionalspectrum to understandthe weaknessesand potentialof their economy, and identifyoptions that may be availablein a post-apartheidera. These options focus on creatingequitable and sustainable economicgrowth and improvingthe distributionof assets and income.

The purpose of this report is to analyzethe performanceof SouthAfrica's agriculturesector and to outline the strategicoptions for improvingthe overallstructure and performanceof the sector, especiallyin terms of incomegrowth, job creation and increasedequity. Ibis report concentratesmainly on the large farm sector, because of its dominancein terms of area and output and because of the lack of informationon agriculture in the homelands.

The conclusionsthat emergefrom the analysisin this report depictthe agriculturalsector in SouthAfrica as one that is dominatedby large-scaleproducers and has been able to producelarge volumesof output- achievingnational food self-sufficiencyfor most commodities.ITis achievementhas, however, come at a considerablefinancial and efficiency cost. The sector has been, and to some extent continues to be, characterizedby significantpolicy distortionsthat have resulted in less-than-optimallevels of efficiency in many parts of the sector, e.g., averagefarm size is too large, ownershipof rural assets, most notably land, are highly skewed, manyfarmers tend to use too much capital equipmentand too little labor. This performance,combined with the active suppressionof black farming, means that the agriculturalsector has underperformedin terms of its contributionto national income,exports, and employmentcreation.

The combinationof problemswithin the agriculural sector and the need for a redistributionof land and support services in the name of greater social justice, make a compelling argument in favor of a restructuring of the rural economy. Such a restructuring would encouragegreater efficiency in the agricultra sector as a whole and would actively support the market-basedredistribution of land and necessarysupport services to those injured under apartheid.

SouthernAfrica Department The World Bank February 1994

copynigt c 1994 The World Bank 1818 'H" Street,N.W. Washington,D.C. 20433, U.S.A.

The vwi epressed us dt paper are hose of the authorsand should not be atribated to dw World Bank, to afiiaed orgd anions, or to membersof is Boardof ExecumiveDirectors or the countriesthy represent. ACKNOWLEDGEMENTS

This report was prepared by a team led by Robert E. Christansen Cask Manager)and Rogier van den Brink. Especially extensive and valuable cotribudons were made by J. van Zyl, N. Vink, M. Mbongwa, H. Binswanger.Numerous other peoplemade contributions, including:C. Blignautand the staff at the Departmentof Agricultre, D. Cooper, M. de Klerk, G. Donovan, P. Fallon, B. Njobe, K. Treu, J. van Rooyen,L. Sibisi and the staff at the DBSA,to name only a few. Particularly usefulbackground reports were prepared by M. Roth,H. Dolny and K. Wiebe, and the Oxford Food Studies Group. Without the help and guidanceof these people and many others, this report would not have been completed. Despite the ontibutions of others, responsibility for the contents of this report and any remainingerrors is the team's alone. ABBREVIATIONSAND CO ld

AMS Aggregate Measureof Support BTr Board on Tariffs and Trade DFB DeciduousFruit Board EEC European EconomicCommission GDP Gross DomesticProduct LSU LivestockUnit SAAU South African AgriculturalUnion SACGA SouthernAfrican Cane GrowersAssociadon SADC SouthernAfrican DevelopmentCommunity SAM Social AccountingMatrix SARB South Africa Reserve Bank SASMAL South Afican Sugar Millers AssociationLimited SECOSAF Secretariatof the EconomicCommunity of SouthermAfrica TFP Total Factor Productivity TABLE OP CONTENTS

Eee ilveSunuary ...... i

Chapterl: RoleofAgricultureIntheNaaroeconomy ...... 1 Introduction ...... 1 MaceoonomicPerformance ...... 2 Agriculture'sConbution to the Economy ...... 9

Chapter 2: Sucture, Resource Endowment, and Performance oftheAgriadulaSector ...... 21 Stucture ...... 22 RsourceEndowment ...... 28 Performance ...... 35

Chapter 3: Evolution of the Contemporary Agrian Structre ...... 44 hitroduction ...... 44 Faming in 19th Cent South Africa ...... 45 The Siuaion at the Tumnof the Cen yfbr theAfrican Farmer ...... 46 AgrarianDevelopment1910-1947...... 47 1910-1947:ProtectionofWhizeAgriculture ...... 51 1948:Apartheid andthe RuralAfrican Household ...... 53 Conclusions...... 60

Chapter4: MarkedngandPridng ...... 61 ntroduction .61 Overviewof Mijor MMrketrkgSchms . 62 Role of Farner Cooperaives ...... 65 Pefbrmance of the Vadous Marketing Schemes ...... 65 Concentrationin the Marketingand Premig Stmcre ...... 80 MarketingStuctures in the Homelands ...... 81 MarketingMargins ...... 85 Self-Sufficiencyand Exports ...... 86 Conclusions ...... 89

ChapterS: Eficiency ofAgricultural Production ...... 91 Introduction ...... 91 PriceEfficiency ...... 94 ScaleEfficiency ...... 112 TechnicalEfficiency ...... 122 Efficiency ssues in the Homelands ...... 129 Chapter6: Famm ...... P ri a biatt 132 Declining Farm Revenues.133 Increasing Farm Costs .135 Cost-priceSqueee .136 Financial Indicatorsof Profitability .138 Debt Burden and Farm Lending Policies .142 Financial Assistanceto Farmers.146

Chapter7: Direction for theFuture .150 Overview.150 Efficiencyand Equity Issues in Agriculture. 153 Optionsfor Rura Restructuring .154 Conclusion.157

Referen s .. 158 Anne 1.171 Annex2 .176 Annex3 .188 TABLES.FIGURES. AND TEXT BOX

1.1 GrowthRates in Real GDPand otherReal Indicators ...... 4 1.2 TotalFactor Productivity Growth Before and After 1973: SouthAfrica andOther Countries .. 5 5... 1.3 SelectedSectoral Multipliers (direct and indirecteffect) According to the ProductionStructures for the Years1978, 1981 and 1985 . .11 1.4 Compositionof GDPsince 1911,by sector(%) ...... 12 1.5 Employment,Economically Active Population .. 14 1.6 The Growthin BudgetedExpenditure, 1979-1989 ...... 18 2.1 LandUse and Population...... 24 2.2 Crop ProductionPotential and Present Cultivationof White-OwnedFarmland ... 30 2.3 ArableLand and PopulationDensity by Region ...... 32 2.4 WaterResources ...... 33 2.5 Real Growthin Agriculture- 1950to 1990 ...... 35 2.6 Compositionof AgriculturalGross Value ...... 37 2.7 GrossValue of AgriculturalProduction in the Homelandsof SouthAfrica.. 41 2.8 AbsoluteValue of Commercialand SubsistenceAgricultural Production in the Homelandsand its Contributionas a Percentageof GDP, 1985 ... 42 2.9 tAigrant and CommuterEarnings Comparedwith AgriculturalEarnings, 1985 ... 43 3.1 AfricanPopulations on VariousClasses of Lands, 1916 ...... 48 3.2 LandAreas by LandTenure Systems in the Unionof SouthAfrica ...... 49 3.3 Land QuotasDelimited by the 1936Act ...... 49 3.4 Allotmentof AgriculturalHoldings During 1916 .. 53 3.5 TargetAfrican Areas (ha) Designated by the Commission...... 55 3.6 Quotaand Non-QuotaLand (ha), 1 January1976 ...... 56 3.7 Non-QuotaLand ...... 57 3.8 QuotaLand as of 31 December1990 ...... 57 3.9 BlackOut-Migration from the WhiteRural Areas, 1980-1985...... 60 4.1 Marketingschemes: Percentage composition of the grossvalue of agriculturalproducts - 1990 ...... 64 4.2 WelfareEffects of PriceDistortions for SouthAfrican Maize Production ... 66 4.3 DomesticPrices and ExportRealization for Whiteand YellowMaize, 1986to 1990 ...... 67 4.4 TheAggregate Measure of Support(AMS) for Maizein SouthAfrica ...... 68 4.5 GrossWheat Producer Prices and MeanWorld Prices ...... 71 4.6 Price StabilizationMeasures in ControlledAreas ...... 74 4.7 Concentrationin the Livestockand MeatIndustry ...... 75 4.8 Sucroseand CanePrices: 1982to 1991 ...... 78 4.9 Total Cane/SugarProduction: 1981 to 1990 ...... 78 4.10 MarketInvolvement of RuralHouseholds ...... 82 4.11 Producers'Share of ConsumerValue .. 85 4.12 AverageProduction and Consumptionof SelectedAgricultural Commoditiesin SouthAfrica, 1985-1990 ...... 87 4.13 Differencebetween South African Producer Prices and the World (Reference)Price for MajorCrops ...... 88 5.1 RealLand Prices,Real Capital Formation, Long-Term Debt, and Arrears... 99 5.2 AnnualEarnings in the Farmand Non-FarmSectors at 1985Constant Prices ... 101 5.3 ComparatveSectoral Increases in RealWages, 1972-1983 ...... 102 5.4 AgriculturaWage Income of BlackPeople on WhiteFams, 1970-1987...... 103 S.S LaborEmployment, Wages and Productivity...... 104 5.6 WageInereases and Variationsin ContractLabor by Region ...... 106 5.7 Distributionof HarvestingMethods in the WesternTransvaal, 1968-1981 ..... 108 S.8 AverageFarm Sizes in the World, 1970 ...... IS 5.9 FactorIntensities in Agriculture,1988 ...... 117 5.10 Land MarketTransactions and Prices,Republic of SouthAfrica .119 5.11 Output, Input, TFP, Labor and Land ProductivityIndices ...... 123 S.12 Anmn Average GrowthRates by Period, 1947-1991 ...... 124 5.13 Average Shares in Revenueand Costs and Annual Growth Rates, 1947-1991 ... 127 6.1 GrossRevenues in AgriculturalSubsectors .. 133 6.2 Price and VolumeIndices of SelectedAgricultural Product Categories .. 13S 6.3 Priceand Volume Indices of FarmingRequisites ...... 136 6.4 AverageAnnual Growth Rates in CertainPrices and Quandties,1960-1990 .... 137 6.5 FertilizerPices: PercentRise on PreviousYear .137 6.6 Incomeand Expenditure,Assets and Liabilitiesof Agriculture,1970-1991 .... 139 6.7 AgriculturdFinancialndicators .140 6.8 Real LandPrices, Real Capital Formation, Long-Term Debt, and Arrears.141 6.9 CommercialFarm Debt in RSA by Creditor: 1970-1990.144 6.10 Farmers'Debts at the 33 CooperativesProducing 90% of all Grain in the Republicof SouthAfrica-Before and Afterthe 1991/92Drought ... 148

F-gut

1.1 GDP Growth(1960-1991) ...... 3 1.2 TotalFactor Productivity - AgriculturalSub-Sectors (1947-1990) ...... 13 5.1 Primary Inputs(1947-1991) ...... 95 5.2 Capital Inputs(1947-1991) ...... 96 5.3 IntermediateInputs (1947-1991) ...... 97 5.4 Total Factor Productivity,Output and Input (1947-1991)...... 125 5.5 TFP, Laborand LandProducivity (1947-1991) ...... 126 5.6 AgricultureOutput (1947-1991)...... 128 5.7 AgricultureInputs (1947-1991) ...... 12Z 6.1 Real Net FarmIncome, Total Fa tor Productivityand Agricultural Terms of Trade (1947-1991)...... 138

exexBox1

What is the Relation between Farm Size and Efriciency? ...... 113 Souh Afri Ar?ubre: Stncture. Pednnanam& Optionsfor )hesuture

EXECLM1V SllMARY

Intrducdion

1. AgricultureIn South Africa is widely regarded as a highly sophisticatedand successfil sector. Among the evidence cited In support of this view is the fact that South Africa is self-sufficientwith respect to most of its major agriculturalcommodity requirements. At the same time, the sector's small and declining share of GDP (4.7% in 1990 comparedto 12% in 1960) is interpreted as indicatinga pattern of secular decline of agrcultural productionthat is consistentwith a normal pattern of economic growth anddevelopment. Consequently,it is often assumedthat agriculture'sfuture contributionto GDP- -in terms of employmentand income-will continueto diminish. It is in this contextthat agriculture'srole in a growth stzategyfor a new South Africa is typicallyregarded by policy makers-both in and out of government-asdiminishing and overshadowedby expectedgrowth in the industrialand servicessectors.

2. The domipant form of agriculturalproduction in South Africa is the whiteowned, large-scale farm-accountingfor 90% of the value added and owning 86% of the agriculturalland. The large-farm sector is commercially-oriented,capital-intensive, and generally produces a surplus. In contrast, the frming sector in the homelandsis unableto meet the subsistenceneeds of the homelandpopulation, with the result that the homelandsare net food importers.

3. The current dominanceof the large farm modelof production is largely the result of more than a century of policy induceddistortions. Initially,the focus of agriculturalpolicy was to reduce and then eliirnate competitionfor white-ownedfarms from black farmers. Having achievedthis during the first half of this centuryby confning black agricultureto the overpopulatedhomelands, agricultural policy has more recently concentaed on providing direct financial support to the large-scalefarming sector.

4. The central question that needs to be answered-and the one that defines the purpose of this report-is whetheror not SouthAfrica's past agriculturalstrategy has been successfulin terms of its ability to produce agricultura goods efficientlyand equitably. The results of this analysis will have a clear Impacton the developmentoptions for the rural economyin a new South Africa. If the past straWgyis found to be effectivein terms of either of these objectives(efficiency and equity), then it only needs to be modifiedto meet the needs of a newly enfranchisedelectorate, e.g., if large farms are efficient,then only the ownershipof these farmswould be an issue. However, if the present strategy is neither efficient nor equitable, then It wil be necessary to reforulate the entire strategy with the objective of restrucuring the rural economy.

5. The hypothesisthat emergesfrom this report indicatesthat agriculture'spresent low share of GDP is W the result of a normal developmentpattern, but instead is a consequenceof wideGpreadpolicy distortionsitroduced by decadesof governmentintervention combined with the dominanthistorical role of the miningsector in the economy. These interventionswere guided-in large measure-by the general political and economicphilosophy of white domination-apartheid. Some of these distortions are not peculiar to agriculture, but characterizethe entire economy, e.g., high capital-intensityof productionin the presenceof widespreadunemployment. Nonetheless, it is true that agriculturehas been and remains

Eiawe $"=my Page i South African Agricaufre: Strucure. PerformanceA Optons for' the Future subject to policy distortions that have been especiallyfar reaching. As a consequenceof the incentive structure created by policy disWortions,agriculture's structure, characteristics,and most notably, Its performan(;eare in sharp contrast with what would be expectedgiven Its resource endowment.

6. The purpose of this report Is to analyzethe performanceof SouthAfrica's agriculturesector and to outline the strategic options for Improving the overall structure and performance of the sector, especiallyin terms of incomegrowth, job creationand increasedequity. This report concentratesmainly on the large farm sector, because of its dominancein terms of area and output and because of the lack of informationon agriculturein the homelands. This summaryis intendedto provie the reader with the highlightsof the chapters along with the Implicationsof the analysisfor strategic optionsfor agriculture and the rural economy.

The Role of Agriculturein the Macroeconomy

7. In 1989, SouthAfrica's GDP was US$80.4biUlion, with 13% of GDP derived fromservices, 45% from industry, 26% from manufacturing,11% from mining, and 5% from agriculture. For the same year, the total value of SouthAfrican agricultural production was approximatelyUS$7.2 billion or about half again as large as that of the entire SADC region.' Employment in agriculture in 1990 was approximatelyequivalent to 761,000 full-time workers and accounts for about 10% of total wage employment. In addition, about 17 million blacks-40% of the black populationof SouthAfrica-reside in rural areas, with more than 13 million of this number concented in the homelands.

8. The themes that characterizeagriculture's role in the broader economy can be summarizedas follows. Agriculture'spresent limitedrole in the economy can be explainedby the dominanceof the miningsector, the skeweddistribution of natural resourcesin favor of a white minority, the adoptionof highly capital-intensiveagricultural technology, the reductionof employmentaccompanying the adoption of this technology,and the objectiveof achievingself-suffic'ency in major agriculturalcommodities.

9. The salient features of the macroeconomicenvironment that have affected agriculture can be summarizedin four main points.

L. Declininggvrowth and productivity. In most sectors, policiesfavored a capital-intensive production pattern. The increased investment in capital intensity, however, did not generatesufficient growth with the consequencethat the total productivityof the economy declinedduring the last two decades. This contributedto the slow growth of net GDP in recent years, e.g., 1.03% per year during the 1986-91 period. As a result, the macroeconomicenviroment facing agriculturewas an especiallyunfavorable one.

In 1989, SouthAfrica producd more than the combind total of the SADC counties for a numberof baic agicual prducs includingmaize, wheat, swnlowe seed, and sugr. Ihe value of SouthAfiaes's agieultural cxpos is broadlyequal to the value of total SADC countryagriltu ports (in 1988USS1.40 billion for the SADCcountries and US$1.46billion for SouthAfrica). Nonethuess,it is estmated that the total area underarable and peranent croplandin SouthAfrica is only sty more than half that of the SADC countriescombined. Unlike mostSADC countries,South Africa is self-sufficienin ereas producdonand, beween 1980 and 1989, producoedsurpluses of such basic food cops as maize,wheat, rye, dry beams, rain sorghum,growdnuts, soybeans, oats, and barley.

Ewecive Smnary Page a SogthAfi*an A&rdgure:Srwe. Peformce& Orionsfhr the Fture ii. Increasingunemployment and unequa1income distributinn Investmentin capital intensitycombined with policiesthat createdhomelands as labor reserveswith the objectiveof keepinglabor costsrelatively low contributedto very highunemployment levelsand political tensions. Labor-segregation policies resulted in a relativelyunsklled labor force and an extremelyunequal distribution of income,services, and demand. Theseforces bave been instrumena in causingagriculture to moveaway from reliance on the relativelyplentiful factor-labor. Ill. Aninward-lookingIigg ftam=rk. Thegovernment has created an incentive stucture that encouragesproduction for the domesticsector, rather than for internationalmarkets. 'Publicsector policy strongly supported domestic producers through tariffs, quotas and otherbarriers. As a result,many producers-including in agriculture-arerestricted to comparativelyhigh-cost domestic sources of supply that reduces international competitiveness. iv. Publicsector investment programs. Publicsector investment was used as a meansof compensatingfor declinesin private sector investment. These investmentshave, however,often been undertaken with little regardto the socialrate of return. 10. The single most importantfactor in explainingthe decline in growth is the lack of total productivitygrowth in the economy.The absenceof significantgrowth in productivity,in turn, stems from the capital-intensive,inward-looling growth path chosenfor the economy. ComparingSouth Africa's total facor productivitygrowth wih thOatof a numberof comparabledeveloping counties, Chapter1 demonstratesthat SouthAfrica consistently under-performs. Over the period1950-73, South Africa'stotal factor productivity shows near-zero growth, i.e., 0.2%per year. Over the period1973- 1984, totalfactor productivity becomes negatiWve, i.e., -0.5%per year. Overthe period 1981-1988, total facto productivitydeclined by 1.1%per year. 11. The low productivitygrowth that chacterized the macroeconomywas also evidentin the agriculturesector as a whole,although not to the sameetent. Between1947 and 1991,the totalfactor productivityof SouthAfrican agiculture increased by 1.26%per year-a rate thatis relativelymodest by Internationalstandards. This overall pattn, however,mbsks the substanialdifferences in productivity growthbetween the mainagricultural sub-sectors. The export-orientedand inherentiylabor-intensive horticulturalsub-sector grew at 2.42% annualyover the 1947-1991period, whereas the livestockand fieldcropsub-sectors both recorded productivity growth of only0.77% per yearover the sameperiod. 12. Agriculturalpolicies in SouthAfrica were closely intertwined with general economic, social and politicalpolicies. The policyenvironment that accountsfor muchof the observedperformance can be dividedinto three part Thebeginning of the firstphase was marked by the LandAct of 1913. During this phase-whichlasteci unW 1940-the basic institutional framework of a dualisticagrarian structure was set up. The secondphase-consisting of the postwarera prior to 1980-sawthe transformationof the sectorinto a highly mechnizedand capital-intensivefarm structure. The third phase constitutedthe policyreform and structuraadjustment process embarked upon since the early 1980s.

Eecal S&Muay Page lii Sosth Aftum Ariukxe: Srucwrg. P&nnanwe t OPws fbr th Futur

St=c=e, ReSprce Endowment.nd Performanceof thie AgriculturalSector

13. South African agrcultural production dwarfs that of the rest of Southern Africa. Despite a relativelylimited naural resource base, its agriculturaleconomy has reachedhigh productionlevels in a wide range of agriculturalcommodities compared to other countriesin the region. In this regard, the sector's performance is consistent with one of its objectives-self-sufficiency. This impressive performanceis based on a resource endowmentthat-while not generous-is highly varied allowingfor considerablevariation in commodityproduction. The ownershipof this resourceendowment, however, is extremelyskewed-the legacy of racially-basedpolicies.

14. Agricultureoccupies an importantposition in the SouthAfrican economy in terms of production levels and as a supplier of inputs to the agro-processingsector. Agriculture's share of GDP, however, has declinedfrom around20% in the 1920sto 4.7% in 1991and its growth in the past 40 years has been uneven, averaging2.5% in real terms since 1950, with rapid growth in the 1970s bracketedby slower growth in the 1960s and 1980s.

15. Some95% of the value of agriculturalproduction in South Africa originatesin the large-scale, commercialsub-sector. Over the past 30 years, the compositionof agricultureby gross value has moved in favor of horticultural products and, to a lesser extent, livestock. Horticultural commoditieshave generallybeen more buoyant than the field crops, with a rise in the total contributionto sectoralvalue of productionfrom 14.4% in 1980/81to 20.9% in 1990/91. Stronggrowth in the horticulturalsector has been shown in pardcular by commoditiesdestined for export, notablyfruits, but wine and vegetableshave also risen sharply. The share of field crops peaked at 48.5% in 1980/81 and fell to 34.2% in the following decade.

16. Land distributionin SouthAfrica is amongthe most highly skewedin the world. Four centuries of conquest, occupation, annexation,land acquisitionsand transfers, land purwhases,and consolidation have resultedin a land distributionwhich is comparablein its inequalityto thoseof manyLatin American counties that often experienceda similarhistory of Europeanconquest and settement. What sets South Afica apart from the rest of the world, however,is the relative emptinessof much of its rural landscape- exceptfor some high-potentialareas along the eastern seaboard. In most of the country, ruralvillages, settement, scattered firms and homesteadsof farm laborers such as those found in the rural areas of Europe, Latin America, Asia, and elsewherein Africa are rare.

17. Approximately86% of agricultaur land is held by 67,000 mostly white, large-scale farmers, supportinga rural populationof 5.3 million. Consequently,the large-scalefarming areas-the platteland- have a very low populion density (about 16.2 hectares of agricultural land per rural resident). In contrast, a rural popudationof about 13.1 million resided in the homelandsin 1988, on an area of 17.1 million hectares. Thus, 29% of South Africa's rural population (mostly farm workers and their dependents)lived on nearly 86% of the agriculturalland, while 71 % of its rural populationlived on the remaining 14%.

18. Governmentaid to the two sectorshas been unequaland variable. The ratio of total government expenditurefor white and black agriculturebefore the World War II was about 179 to I and around 14 to 1 in the 1950s. By the 1980s,this ratio increasedto around4 to 1. In 1987, the 14 Departmentsof

&Ecawfr siuwnaay Page iv SosahAftr= A.rllwu,a Sirmure. P& =mAncgod for ha Futr Agriculturehad a combinedbudget of R2.1 billion, of which R1.7 bilion was allocatedto white agrculture. Moreover,large-scale, commercil farmers obtain 96.7% of all transferpayments in the budget. 19. Ihehomelnds todayare over-crowded,poverty-stricken and lackinfrastructure comparable to the white farmingsector, despitesome attempts at development.Iheir economicsignificance as a reservoirfor surpluslabor has intenfied over time. In 1951less an 5 millionblacks lived in the homelands.This numberpeaked at 14 millionin 1985and declinedto 13.1 millionIn 1988. These changesresulted from the reselement policiespursued by the goverment, in pardcularbetween 1960 and 1983,when miilions of blackswere relocated to the homelands.Well over half of householdincome in the homelandsis today derived from repatated migranteanings and pensions. Agricultural productionis estimatedto meetonly 16%of the population'sfood requirements and constitutesat most 10 to 20% of householdincome. 20. Thefirst impressionta emergesfrom an overviewof the agrtculre sectorIn SouthAfrica-as it standstoday-is of a set of strucures that has been very successfulin producinglarge volumesof commodities.Tbe volume of productionrelative to neighboringcountries and the highlevel of foodself- sufficienc all confirmthis impression.What is less apparent,however, are the highlyartificial means hat have been employedto establishand maintainthese agriculturalstructures. Understandingthe disIminory mecms and cotivances that havebuen used to createthe dominantelements of the sectoris esset to und ing the optionsavaiable to the sectorfor the future.

EvolutkAmof the Agdan 21. Giventhe domice of the large-farmmodel in SouthAfrica, it is someimesassumed that this modelis moreefficient than the alternatives-mostnotably small-scale, labor itensive production.Not onlydoes intnaonal evidenceargue against this presumption,but the historicalexperience in South Africadoes as well. In fact, oneof the mosttelling lessons to emergefrom the SouthAfrican experience is the successof small-cale,commerci blackfarmers durng the midto late 19thcenury. The most convinig evidenceof this successcan be foundin the unusualmeasures taken to discrminateagainst thesefarmers. 22. Afican familyfaming in the latter half of the 19thcentury was viableand ableto meet the agriculturalneeds of ura and industri marketsthat developed around the major mining centers. Given relativelyudistr maket forland, labor, iputs, andoutputs, African farmers progressively adopted am farmingtechnologies, expanded into new farmingindustries, and competedwith seter farmingin domesticand in certainexport crops markets. Criticalto this successwas the inabilityof a weakstate to intervenein the factormarkes andthe implicitsupport for Africanfarming from land companies and big landownerswho eaned rents fromAfrican tenant farmers. Theseconditions encouraged African farmersto accum capital,wealth, and farmingskils. 23. The formationof the Unionof Souti Africain 1910led to the creation-overtime-of a policy environmentcharacterized by the suppressionof Africanfrming and its eventua isolationfrom 20th cettury mainstreamagriculure. With its main objectivebeing to tmn Africanfarmers to laborersand

eucutin UWY Page ' Sou&Afia Aupkcbur: Struru, PeAWna B OrCionsfor the Fturs- channelthem to whitefarms, mining ad manufacturing-theapparats of the Unionof SouthAfrica was strongerand richerthan state fiomationsof the earlierperiod. In addition,the landcompanies and big landownerswho had previouslyfrustrated earlier policies to cuil Africanfarming, were now either includedas part of the new policyenvironment, or their leveragewas reducedrelative to that of the miningindustry and setter farmers. Whatwas damaging African farmng was not onlythe legislative and administrativemeasures that werelevelled against it but, more importandy,the systematicclosing of all avenues-whetherthese were land, input and outputsmarkets, or credit-exceptthe labormarket to Africanfarmers to accumulatecapital, wealth and farmingskills.

Agrcultral Marketingansd Pring 24. Ihe suppressionof small-scaleblack farmers was complemented by the extensivesupport systems providedto the large-scalefarming enterprises. Perhaps the mosteffective form of supportto the large- scalefarms has beenthe pricingand marketing systems. In essence,the combinadonof controlledinput andoutput prices along with singlechannel marketing systems for the majorityof agriculturalcommodities have restrictedcompetition and providedlarge-scale farmers with an incomeguarantee. Duringthe 1980s,efforts to liberalizethe pdcing and marketingsystem have introducedmore competition-and contributedto the rise in bankruptciesamong large-scale farmers-but also must be regardedas incomplete. 25. Thebulk of agriculturalproducts in SouthAfrica are marketedunder schemes set up in termsof the MarketingAct of 1968. The originalpromulgation of this Act in 1937was closely related to other legislation,including the Co-operativeSocieties Acts, the LandActs and the Landand Agricultural Bank Act. As of the early 1990s,about two-thirdsof the valueof farm producewas marketedunder the MarketingAct. 26. The Markeint Act is enablinglegislation that appliesto all productslisted in its schedule.It providesfor the promulgationof subordinatelegislation called -Schemes", where a Schemeis instituted for a productor group of products,and a ControlBoard is set up to administerit. Each Scheme empowersthe relevantBoard to carry out the normaladministrative functions and also empowersit to carryout specificfunctions that influencesthe marketingof the product. TheMarketing Act provides for four maintypes of schemes,tailored to the specificcharacteistics of the commodityin question.These include: shishanel fixed pn schemes Producersare legallyobliged to markettheir products throughthe Boardor its appointedagents and pricesare fixed for each season. The major domesticcrops such as maizeand wintergraiDs (wheat, barley and oats) fall intothis category.

singlchamel OLschM@a. Producersmarket their productsthrough a poolconducted by the Boardswhich pays on receiptof the product. Supplementalpayments are madewhen the final realizationof the pool,after deduction of poolepenses, is known.Crops facing relatively elastic demand,e.g., exportcrops, fallin this category.

u sceme (alsocalled price support schemes). Producerssell their produceon an

EweaaveSaUOY Page w SouthAft*an Ar4iukure Sur Prfan & Opn for ** Funtre

open market, although they are often subject to stringent quality controls that could restrict domesticcompetition. The Board interveneswhen prices drop below a fixed minimumprice by purchasingsurplus suppliesfor distributionand resale at a later date. Schemesinclude those for red meat.

Uiw scheame. The Boardacts in a supervisorycapacity and as a mediatorin the arranging of price and purchase contractsbetween producersand buyers. Products include canningfruit and cotton.

27. The opportunity costs of the current marketing system include, among other things, the development of a relatively concentrated sector aimed at specific market segments. By catering exclusivelyto large-scale commercialfarmers, this system prevented both the developmentof black farming and the developmentof informalmarketing systems which could have created substantialblack employment.Also, by focussingon marketingof high quality,first world agricultura produce,jobs were lost in marketingof lower qualityproduce, both domesticallyand for export.

28. Current trends in marketing are going from a controlled, cost-pluspricing system to a more market-orientedsystem (removalof domestic controls, reductionof protectivemeasures, and reduction of tariffs). Given the prominenceof high levels of concentrationin the agricultureand food processing sector, the establishmentof clear policy guidelines setting out Government's objectives and policy instrumentsis a high priority. Three alternativepolicy approachescan be distinguished:

laissezfaire: under this policy the Government would adopt a passive attitude towards concentration,hoping that the possiblebenefits from economiesof scale and competitivenesson intrnational markets would outweigh the potential costs from monopolisticpricing. In recent decades policy has been rather close to laissez-faire.

active anti trust aravunder c: this policy the mandateand powers of the CompetitionBoard would be substantilly increased with a view to actively reducing the existing levels of concenration through forced divestment of holdings by the major firms. In practice plementationof this policy approachis likely to prove difficult becausethe Governmentwil continueto lack sufficientinformation to pursue a legal approachto anti-trustand becauseforced salesof subsidiariesare likelyto lead to purchasesof divestedfirms by the existingconglomerates rather than by new entrant to the marketand the unemploymentimplications of possibleclosures inducedby an active antitrust policy would be politicallydifficult to accept.

the indirect Uoach throughdereeulating the economicenviromnent: under this third approach Governmentwould not actively attemptto break up the dominantfirms; rather it would seek to promote changesin the economicenvironment which over time will remove the economicbasis for concentrationand reduce barriers to entry for potenal competitors. Adoption of this approach requires recognitionthat the problem of concentrationcannot be resolved in a short period, but that over a longer period the presenceof actual competitionor the threat of potential competition can be the most effective instument in combatting the adverse aspects of concentration.

&ecutfvs wmaiy page vii SorthAfrka Aic L Strcture.rure: Performance & Options for th Future

Efficiencyof AgriculturalProduction

29. Agriculture in South Africa has consistentlyproduced surpluses, and as a result is typically regarded as an efficient sector. In recent years, however, the financialproblems experiencedby many farmers-largely as a result of drought and policy changes-have caused some important sub-sectors in agricultureto be seen as inefficient The issue of large farm efficiencyin SouthAfrica is critical to the developmentof a strategy for the rural economy. If, as some argue, the large-farmmodel is a valid one- characterizedby economiesof scale-then a strategy that consists largely of removing existingpolicy distortions would increase the efficiency of the sector as a whole and further changes would be unnecessaryin terms of increasingefficiency. If, however,there are no significanteconomies of scale associatedwith large-farmproduction models, simply reformingpolicy will only yield a portion of the potentialefficiency gain. The balance of the efficiencygain could only be achieved by restructuring- through downsizingexisting farm units-the present farming sector.

30. Unfortunatelyin South Africa, a direct comparisonof efficiencybetween the large-scale and small-scalefarm sectors is difficult if not impossible,because black owned, small-scalefarms have all but disappeared. Amplehistorical and internationalevidence suggests, however, that when comparedto small-scalefarms many large-scalefarming enterprises in South Africa, as elsewhere in the world, are inefficientwith respect to profits, employmentand output generatedper unit of capital invested.

31. The move towards larger farms in South Africa was stronglyassociated with farmers' adoption of m on, whichwas facilitatedand strengthenedby subsidizedinterest rates, tax breaks, andthe support of output prices above border parity. Yet in spite of massive additionalfinancial assistanceto farmers, the profitabilityof agriculture remainedlow. By 1984, an estimated22,700 farmers (33% of the total) had debt burdens beyond the critical level. Most of these farmers were concenuated in the Transvaal and Orange Free State-the regions where mechanizationof monocroppedcereals had been pushed to economicallysub-optimal levels and exposedfarmers to high risk levels.

32. The policy objectives that justified these interventionswere twofold-food self-sufficiencyand maintainingincome levels for large-scalecommercial farmers. Agriculturalproduction in South Africa increasedat a rate of over 3% annuallyover the last four decades, while the populationincreased at an averagerate of 2.8%. Moreover,the social objective of maintainingincomes for large-scalefarmers at levels comparableto the incomesin the urban sectors seems also to have been obtainedfor at least some farmers. From 1970/71to 1991, total net farm incomeincreased in real terms from R2,728 million to R3,469 million, but with significant fluctuationsthroughout the period and a considerably skewed distribution.

33. The findings of this report with regard to the three major types of efficiencyexamined are as follows:

Price effici Thne agriculturalsector experienceda policy-inducedshift toward a large-scale, capital-intensivefarm technology, heavy reliance on intermediate inputs, and toward less dependenceon labor. This reflects farmers' rational reaction to the prevailing structure of incentives, in particular, taxation and financial assistance programs. In other words, most individualfarms were efficientlyrun given the prevailing structure of incentives. Althoughthe

Em&tI Stnry Page Vi Sou_hAfan Airiculkure:&rucgure. Performance & OtIons fr the FOure

price of the input bundleassociated with the prevailingtechnology is one elementin the structure of incentivesfaced by farmers, it has not alwaysbeen the dominantone. Prior to 1973, large- scale farmers display considerableprice efficiency. However, starting in the mid-1970sfarmers demonstratedless responsivenessto changesin relativeinput prices. This seemsto have been the result of a combination of financial assistance programs, tax incentives, and technical improvementsthat outweighedinput price changes. In effect, farmers respondedto a distorted set of incentives that induced a different production technologythan that which would have prevailed in a more market orientedregime.

Scal efficiency. The same set of policy distortionsthat affected price efficiency also created incentivesfor the emergenceof large-scalefarms. As a result, prior to the end of the 1983s, there was an increasinglyconcentrated agrarian structure in SouthAfrica. In the grain sector for instance,the resultsof this report indicatethat once the effects of the distorted incentivestructure are removed, a wide range of farm sizes (above approximately100 hectares) are efficient with respect to scale-in other words, a larger farm is not necessarilymore efficientthan a smaller farm. The crucial factor in determiningthe greater efficiencyof a large farm over a smallfarm is managerialexpertise. For SouthAfrican farms coveredby official statisticsand smallerthan approximately100 hectares, there is empirical evidenceto suggest that small-scalefarms use resources more intensivelythan large farms. However, whereas little evidence exists on the efficiencyof black, small-scalefarming in South Africa, internationalexperience indicatesthat in an undistortedpolicy environmentsmall farmsare often at least as efficientas their large-scale counterpartsprovided the working capital constraintis removed.

Technicalefficicy. Trends in technicalefficiency can be grouped into three distinct periods. First, improvementsin the technicalefficiency of SouthAfiican agriculturewere virtuallyabsent during the 1950sand early 1960s. During this period, agriculturalproduction increased through an expansionin cultivatedarea, supportedby the introductionof tractors. Second, modest,but significant,gains in technical efficiency appeared in the late 1960s and continuedduring the 1970s. These gains are associatedwith the intensificationof agriculture during this period including the introductionof high yielding varieties. Also during this period the relationship between labor and machinery changed from a complementaryone during the period of area expansionto one of substitution,especially as a result of the mechtion of the harvesting process in cereal farming. Third, substantialgrowth in technicalefficiency was recorded after 1983, as farmers adjustedcapital inputsto changesin the structure of investmentincentives. For the post-warperiod as a whole, however, the trends in technicalefficiency differed substantially by sub-sectorwith livestockand field crops performingwell belowthe average,while the export- oriented and labor-intensive horticulture sub-sector showed dramatic increases in technical efficiency.

34. Hence, the picture of South African agriculturethat emerges from this analysis of efficiencyis a varied one, but one for which several important sub-sectorscan be characterizedby: (i) a gradual deteriorationof price efficiencywhich in effectshows that farmerswere compensatedfor the impliedloss of efficiencythrough other channels; (in)scale efficienciesthat are the result of a combinationof a distortedpolicy environmentand managerialability; and (iii) unevenimprovements in technicalefficiency that are modest by internationalstandards and differ markedlyby sub-sector.

E&euve Summwy Page ex South AfrcfwaAtricubure: Mructure.Performanwe & Optionsfor tho Future

35. The policy distortionsthat have given rise to the social inefficienciesthat characterizelarge parts of the agriculturalsector (e.g., subsidizedinterest rates, favorablecredit regimes and tax coneessionson agriculturalinvestnment) are not particularlydifficult to reverse. Over the past decade, a wide range of policy changes in the general direction of greater market orientation have been introduced and have affectedboth agriculturalinputs and outputs. Tle impact of these changeshas varied considerablyby subsector: debt-free producers of some export commoditieshave benefited, while profits for others, notably the grain producers, have fallen. For the sector as e whole the capital-intensityof production declined throughout the 1980s and employment increased through 1987. More recently, however, employmenthas again declined. Althoughthe cause for this recent decline is unclear, it may be due to political uncertainty over the future of white-ownedfarms, extensificationof agriculture in marginal cropping areas, and the financial deteriorationof the sector.

Farm Profitability

36. The ultimate test of efficiencyand financial sustainabilityof any enterprise, however, is the responseof large-scalefarmers to the withdrawalof the preferentialpolicy treatment. If as the structure of incentives(e.g., prices of inputs and outputs, taxes, and/or marketingsystems) changes,farmers are unable to respond adequately and ensure their profitability,this constitutesa reliable indicator of their inefficiencyin an undistortedpolicy environment.

37. Since the early 1970s, agriculturein South Africa has been exposedto a cost-pricesqueeze that reduced farm profitability. This was compounded during the 1980s by: (i) unfavorable weather conditions;(ii) reductionof farm subsidies;(ifi) liberalization of marketingpolicies that often resulted in lower real producerprices; (iv) and higher capital costs due mainlyto higher interest costs. The failure of many firms to respond to the changing price environment(recall the finding above of only modest price efficiency)meant that many farms continuedto rely on the established,capital-intensive production methods. Given the variable agro-climaticconditions to which South African farming is exposed, the continuedreliance in importantsubsectors on mechanized,monocropped farming also entailed a higher exposureto climatic risk and increasedthe volatility of farm profits.

38. As a result, the debt burden carried by many farmers escalated to unsustainablelevels, in particularthose in the summerrainfall areas (e.g., wheat and maize growers). Althoughreal farm debt reachedits highest level in 1985 and has declined since then, the financialposition of many farmers has continuedto deteriorate,as illustratedby a continuingincrease in the debt-assetratio. Moreover,the ratio of short-term debt to total debt increasedconsiderably. Higher capital costs, tighter credit standards, fewerfarm subsidies,and the growing unsustainabilityof the debt burden shouldhave positiveeffects on labor employment in the sector and force less efficient farmers into foreclosure. However, the governmentcontinued its attemptsto aid the sector financiallyand when the 1992 drought occurred, substantialamounts of debt which had been accumulatedduring the 1980s were effectivelywritten off through a massiverecapitalization effort targeted to grain producersand totaling R3.4 billion.

39. In summary,the declinein farm profitabilitysince the eady 1970sand the evidenceregarding the efficiencyof large-scalefarms suggests that, from the perspectiveof society, manyparts of the agriculture sector have operated inefficiently. It is important to stress that this does not necessarilymean that

E&ecudveSwmay Page x Southg ian AgricuAure:Structure. Performance A Opions for the Future

individual farmers are poor managersor not responsiveto the prevailing structure of incentives, but instead it means that the policy environmentfaced by farmers has resulted in inefficientperformance of the sector as a whole. The realizationof the adverseimpact of the policy environmenton efficiencyand the economycontributed to the agriculturalpolicy reformsthat were begun in the early 1980sand which have led to some Improvementin efficiency.

40. As noted earlier, a critical elementof an evaluationof the efficiencyof the sector is the role that the scale of operationsplays in farm efficiency. The evidencepresented in this report suggeststhat scale economieshave existed in South African agriculture, but they stem from the distortions introduced through the policy environment. Hence, if the bias toward large-scalefarms was eliminatedfrom the policy environment, the cost advantageenjoyed by large-scale faums would likely disappear. This conclusionmust, however,be treated with caution,because there is tremendousvariability of production models across the agriculture s'-or. Therefore, it would be incorrect to argue that a smatler scale of operations would be appropriate for all subsectors. Nonetheless, it is probably the case that cost structures are either neutral with respect to scale or that diseconomiesof scale would exist if policy distortionsare removed. Consequently,it is possibleto assertthat a combinationof intenational evidence with regard to farm size and recent South African experiencesuggest that a smaller scale of operations in many agriculturalsub-secto would be at least as efficientas the larger scale farms once the policy distortionsare eliminated.

Optionsfor the Future

41. This report argues that while the agriculturalstrategy pursued in South Africa has achievedits two main objectives-food self-sufficiencyand acceptableincome levels for white farmers-it has done so by distortingthe policy environmentand causingsociety to pay a considerablefinancial and social cost. Quite apart from the racial consequencesof this policy environent, society has been deprived of a greater contributionto GDP that a more efficientand dynamicagricultural sector would provide. At this point in time, the agriculturesector remains in a state of uncertaity. Nearly a decadeago, the financial cost of the agricultural stegy became sufficienty burdensome, that a limited liberalization was introduced. This policy change, along with adverse weather conditions, had the effect of exposing farmers to a more exacting and market-orientedenvironment. As a result, profits declined and debt increasedin many parts of the sector. Ihis result is a telling one, becauseit suggeststhat many elements of the large-farmsector are not efficientunder the new policy environment.

42. The problem of inefficiencyin many parts of the large-farmsector, however, must be examined in the contextof two salientfeatures of contemporarySouth Africa. First, the present highly inequitable distributionof resources-most notablyland-is not sustaible. Internationalexperience demonstrates that economieswith a land distributionsimilar to SouthAfrica's are prone to a pattern of civi disorder and violence. In Zimbabwea fairly substantial,but poorly designedland reform programhas made only limitedcontributions to output and employmentgrowth, and left the land issue as an unresolvedpolitical and economic liability. Most other dualistic countries have fared much worse. Chile, Colombia,El Salvador, Guatemala,Nicaragua, Mozambique, Peru and Angolahave sufferedprolonged periods of up to forty years of unrest and civil war. Strongpeasant support for revolutionaryforces indicatesthat this was closelyrelated to these countries' delay or inabilityto effectivelyredistribute land to a large number

Exec&iWveSwnmaiy Pagexi SOwhAfrican AfriculSure r. PerbimaMc & Otis for the Fuur of beneficiaries. These prolongedperiods of conflicthave resulted in capital flight, economicstgnation or decline, social and politicaldisintegration, and untoldhuman suffering. Second,majority rule in South Africa is imminent.

43. The combinationof wealnesses in the old strategy, inequitiesin the distributionof and access to resources, and politicalchange means that quite apart from issuesof inefficiencyin the large-farmsector, the inequity of the present land distributionwill necessitatea major rural restructuringincluding land reform. In light of internationalexperience and the present circumstancesin rural South Africa, the conclusionof this report is that there are few developmentoptions for agricultureand the rural economy availableto a new Governmentin South Africa.

44. One option is to continueliberalizing the present policyenvironment, while leaving the present structure of agricultureproduction units largelyunchanged. The process of ongoingliberalization would include: (i) further reform of the input and output marketing system; (ii) efforts to reduce the concentrationin the agro-processingsector; (tii) revision of land sub-divisionguidelines; and (iv) restructmuingof the present agriculturalcredit system. Continuedliberalization along these lines will likely result in more bankruptcies among large-scale farms, the expansion of small-scale farming (especily near urban areas), expansionof the horticulturalsector, and contractionof cereal and livestock production. Althoughthe appeal of a more efficientagricultural sector through continuedliberalization is clear, the drawbackto this option is that changes in the operationand ownership of the large farm sector are likely to be very slow. The chief reasons for this are: (I) in an environmentof uncertainty over land ownership,few large-scalefarmer wil be wfllingto make the investmentsnecessary to respond to the new policy environent; and Oi) few of those who were disfranchisedunder apartheid wil have the resources neededto gain access-either throughlease or purchase-to land currentlyheld in the large- scale farm sector. Hence, while this option offers some increasedefficiency, it is only a parti gain because such an option essentiallyaccepts a large-farmstructure which-as this report demonstrates-is Inefficientin many critical subsectors. Further, the pace of change is likely to be unacceptablyslow.

45. A second option would be to expand the first focus of the first option to include a rural developmentstrategy for the homelandsthat wouldfeature upgradingof agriculturalsupport services and investingin improvedphysical and social infrastructure. This has the advantageof continuingthe process of policy liberalizationand of concentating publicsector resourceson some of the most obviousvictims of aparteid. Such an option would be likely to contribute to a significant expansionof agricultural production in the homelands, but one that is limited to those few areas in the homelands with good agriculturalpotential. An obviousdrawback to this approachis that-given the fragile conditionof many of these areas-any effortto intensifyagricultura productionin the homelandsis likelyto have undesirable environmentalconsequences. A further problem with this option is that wbile it seeks to address the poverty in the homelands,the efficiencygains are incompleteand the problem of distributionalinequity remains largely unaddressed.

46. A third option builds on elementsof the two options described above-continuedpolicy reform and investmentin the homelands-andin additionsupports a redistributionof agriculturalland in the large- farm sector. A redistributionof agriculturalland would achieve three critical objectives: (i) reduce uncerainty experiencedby current ownersthereby encouragingthose who continuefarming to invest; (n) address the present inequitable distribution of land use; and (iii) encourage, if the redistribution

Ewcutve S&nary Page xi. Sogg*Afic A& W S&nwmure.Perfonauce & Otlons for ih Future. mechanismIs properly designed,those with the greatestinterest in lnd use to gain accessto and use land efficienly. This latter objectivewoud lead to a more dynamicrural economyand to greater employment and incomecreation among low-incomegroups than would either of the first two strategicoptions.

47. The specific objectives of a land redistributionprogm would be to address the inequitable distribution of land and to encouragethe efficient use of agriculturalland. In order to achieve these objectives, it is essentialthat there be agreementon the precise objectivesand methodologyof the land redistributionprogram. Internationalexperience indicatesthat the salient elements that characterizea successfulland reform program are that:

i. market-assisted land redistribution programs tend to perform better than those administeredand operatedby the public sector;

ii. the role of the public sector in a land redistributionprogram centerson ensuringadequate suppliesof land in the market and monitoringthe overall operationof the program;

iii. crteria for participationare necessaryand must be discussedand agreed in advanced;

iv. welfare objectivescan be met by includinga grant componentin the program;

V. a matchinggrant schemethat forces participantsto use some of their own resources in order to gain access to land will help to assist in self-selectionof participants and encouragethe productiveuse of land;

vi. the grant elementsof the program are essentialin order to accomplisha redistributionof assets and to ensure that beneficiariesemerge from the program with a net increase in their asset position and low debtfasset ratio as a means of ensuring viability and sustainabilityof their enterprises;

vii. in additionto Adressing the fundamentalissue of social Justice,these options are likely to significandyincrease n rural employmentand ensurethat the cost of the program is very reasonable;

viii. a redistributionprogram will not be able to provide land for everyoneand the program will need to be complementedby a rural safety net and by programs for urban groups, and

ix. althoughthe discussionconcentrates on agricultre and small farmers these are not the limits of the program. It is envisionedthat the progrm will be a vehicle for supporting a wide range of land use activities, including trading activities, and small-scale enterprses.

48. Finally,it is importantto consideranother set of consequencesof the present agriculturalstrategy- -its envmea costs. The present overcrowded homelands are endowed with a generally poor resource base and are under considerablepopulation pressure that has led to significant environmental

uEmcUiveSawy Pagexii iJll

i~ E

] %:b

o 6tEa

.- a SouthAfA Aguture Stucture.Pfgmance A Ovtionsr the Futare-

CHAPTER 1: ]&OLEOF AGRICULTUREIN THE MACROECONOMY

Introduction

1.1 The importance of the macroeconomicenvironment to agricultural perfonmanceis well- recognizedand the experienceof SouthAfrica is no exceptionto this rule. What is striking about South Africa, however, Is the extent to which macroeconomicand sectoralstrategies have encouragedlarge- scale, capital-intensivefarm models,thereby defying the economy'scomparative advantage, i.e., plentiful suppliesof inexpensivelabor. This choiceof a strategy is, of course, a consequenceof forcingeconomic policy to conformto an especiallypernicious social policy. This chapterargues that agriculture's current role in the SouthAfrican economy reflects not only the dominantrole of the miningsector in the economy and the series of distortions that typified both sectoral and macroeconomicpolicy, but also the subordinationof agriculturalpolicies to social policy-apartheid.

1.2 The salientfeatures of the macroeconomicenvironment that have affectedagriculture can be summarizedin four main points.

a. Declinine growth and productivity. In most sectors, policies have favored a capital- intensiveproduction pattern. The increasedinvestment in capital intensity,however, did not generatesufficient growth withthe consequencethat total productivityof the economy declinedduring the last two decades. This has contributedto the slow growth of net GDP in recent years, e.g., 1.03% during 1986-91. As a result, the macroeconomic environmentfacing agriculturehas been an especiallyunfavorable one-chiefly because slow growth in the economytranslated into low demand for agriculturalproducts.

b. Increasing unemploymentand unequal income distribution. Investment in capital intensity combined with policies that created homelands as labor reserves with the objective of keeping labor costs relatively low has contributedto high unemployment levels and political tensions. These labor-segregationpolicies resulted in a relatively unskilled labor force and an extremely unequal distribution of income, services and demand. These forces have been instrumentalin causingagriculture to move awayfrom reliance on a relativelyplentiful factor-labor.

c. An inward-lookingpolicy framework. The governmenthas createdan incentivestructure that encouragesproduction for the domesticsector, rather than for internationalmarkets. Publicsector policy has stronglysupported domestic producers through tariffs, quotasand other barriers. As a result, many producers-includingthose in agriculture-have been insulated from international competition, thereby reducing incentives to improve efficiencyand distortingmeasures of comparativeadvantage.

d. Public sector investmentprograms. Public sector investmenthas been used as a means of compensatingfor declines in private sector investment. These investmentshave, however, often been undertakenwith little regard to the social rate of return.

Chapur 1: Rolk of AgrkukureIn the Macroeconomy Page 1 &Sh fric Asdgu uSruc. , PteformancW& Optionsfor the F_ture 1.3 Theimpact of socialand macroeconomic policy choices on agriculturehas been profound and has resulted in comparativelylarge, capital intensive,and-in many cases-inefficientagricultural enterprises.Although the agriculturesector has been able to achieveself-sufficiency for most commodities,this has only been possible as a resultof an interventionistand distortedpolicy environment. Consequently,agriculture has not been a sourceof growth in either employmentor incomefor the economyas a whole. 1.4 This chapter is divided into two main parts. lThe first examinesSouth Africa's macroeconomicperformance by analyzingtrends in incomegrowth, productivity, and unemployment, growth strategiesand public sector investmentexpenditures and the secondexanines agriculture's contributionto the macroeconomy.

MacroeconomicPerformance 1.5 SouthAfrica has a per-capitaincome of aroundUSD 2,500per year-the highestin sub- SaharanAfrica. Per-capitaincome for whitesis aboutUSD 9,250-similar to countriessuch as Ireland, Israel,and Spain-whileper-capita income for blacksis onlyabout USD 975-similar to levelsfound in middle-incomecountries in Africasuch as Cameroonand COted'Ivoire. Moreover,economic growth has declinedover the last three decades,while unemployment has increasedto unsustainablelevels. Declininggrowth seems not to havebeen the resultof a declinein investment,instead, it wasthe result of the low productivityof the investments.As economicdecline became apparent, public investment attemptedto compensatefor the lowproductivity of privateinvestment, but withan increasinglylimited effecton productivityand growth. DecliningIncome Growth and Productivity 1.6 Startingin the 1970s,growth of the SouthAfrican economy was characterizedby two interrelatedstructural trends. First, aftera periodof substantialgrowth in GDP followingthe Second WorldWar and continuinginto the late 1960s,an increasinglyvolatile downward trend in growthrates is evident(Figure 1.1). Second,the overallcapital-intensity of the economyincreased, while labor- absorptiondeclined. The substitution of capitalfor laborbrought little gain in productivityIn the main productivesectors of the economy,and total productivity of the economytherefore stagnated in the 1980s.

ChapterJ: Roleof Agriculturein the Macroeconomy Page 2 SouthAfrtc Agttwure: SIuture. Perfonmnee & Optlonsfor the Future

Figure.1.1: GPrwt

ANNUALAND 5-YAR AVERAGEGROWTH RATES (PECENTAGES) 7.5

5.0- 1-

2.5 -

04-~~~~~~~~~~~~~~~~~~~~~~~~~~~~*

so ea 64 86 1&117072 4'78 78 00 02 0i4 Oil 08 90

-i- AN AVAGE ----- PllVOVS 5 Rn AVBRAt:I

Source: World Bank, 1993.

1.7 Trends-in GDPQ and Productiviy. Between 1946 and 1991 GDP growth at factr cost averaged 3.9% per year. But around this trend there is a cycle of accelerating growth rates between 1946 andl971, followetdby rapid deceleration and increasing volatilitybetween 1971 andl1991. In real GDP, grwthratesdecline from 3.49% (for the period 1971-75) to 1.03%S(for theperiod 1986-92) (able 1. 1). in 199 and 1991, real GDP fell by 0.6% and by an estimated 2.1 % in 1992. Further, with population growth around 2.6%tper year, both GDP and GDI per capita have fallen throughout the 1980s.

2 Some modificato of the figu on GDP and anployme are neededto Wmkacmuwt of the cmm of an in&rmal economy WAldkthat is gennal atam thti it has grown in rncentyemr, thae is consitablc re inq about Psie One focust ee;nft wuggs tht in 1990, it amunlted to 8% of GDP, providedanployment for about thx milion people, and nwamed onenchum _ for incomedistnidion to Iowerincome groups (Roukens de lange, 1990).

Chizper 1: Role of Agriallure in the Macrowonomy Page 3 South Afran Arulure: Structure. PerformanceA Ooions for the Future

Table 1.1: Growth Rates in Real GDP and other Real Indicators (percent per year) 1961-65 1966-70 1971-75 1976-80 1981-85 1986-92

GDP 5.94 5.15 3.49 3.13 1.36 1.03 GDI 4.62 5.81 6.39 4.02 0.03 0.27 Private Consumption 4.62 6.00 5.08 2.64 2.65 1.97

Employment 3.18 2.53 2.41 1.42 0.07 -0.23

Notes: 1) Gross Domestic Income (GDI) - rea GDP adjusted for movements in the extemal tenms of tade. 2) Employment- wage and salaried employment including domestic servants.

Sources: South Afnican Reserve Bank (SARB) National Accounts Data Base. SARB Quarterly Bulletin, September 1992. South African Labor Statistics 1992. Central Statistical Service.

1.8 What factors explain this steady decline in income? External shocks such as the faWlin the gold price and internationalsanctions cannot fully accountfor it. Since 1971, the overall terms of trade have moved in favor of SouthAfrica. The oil-price shock was matchedby a bigger increasein gold prices'. However, non-gold terms of trade have declined by 60% since 1960, with most of the latter decline having occurred in manufacturing.

1.9 Neither can the underperformanceof the economybe explainedby structural rigidities In the formal economy. Basic economicrelationships show that the key aggregates,such as exports, imports, and investments,are highly sensitiveto relative prices (Kahn, Senhadji,and Walton, 1992). Moreover, even the most narrow view of the labor market-i.e., taking apartheid's extreme labor market segmentationas given-also indicatesthat employmentand wage behavior in the formal, white sector is quite consistentwith relative price movements(Fallon, 1992). Further, savingshave exceededinvestment for mostof the period. Fiscal deficitsdo not seemto be crowdingout private investment,and the private corporatesector has had a net savingssurplus over and above investmentsince 1972.

1.10 The single most importantexplanation of the decline in macroeconomicgrowth is the lack of significantproductivity growth, stemmingfrom the choiceof a capital-intensive,inward-looking growth strategy. ComparingSouth Africa's total factor productivitygrowth 4 with that of several comparable developingcountries Crable 1.2) showsSouth Africa a consistentunderperformer (Moll, 1991). Over the period 1950-73South Africa's total factor productivityshows near-zerogrowth (0.2% per year). Over

3 As the world's leading producer of gold, South Africa besefitted substantially from the large irses in the gold price dUing the 1970s. From 1971 to 1975 alone, the dollar price of gold ireased by 470%.

' Total factor productivity (IT!') is a mewasureof output grmwth which cannot be ascribed to growth in inputs such as lbor and capital. It is usualy asocit with technologicalchange. Changes in output are compared with a weited average of chang= in inputs used. The ins weights are usually calculatedas the shares of total incomes received by labor, capital, sad ather inps.

Chapter ): Role of Agriculture in the Macroeconony Page 4 SouthApan ArcuueSrurg. Perfonnoce A OAdons fir the Futare

the period 1973-1984factor productivitybecomes negative (-0.5% per year), and over the period 1981- 1988, it declinesby 1.1% per year (Moll, 1990quoted in Kahn et at., 1992).

Table 1.2: Total Factor ProductivityGrowth Befotc and After 193: South Aftica and Other Countries

Growth Rates (% per annum) CDP Employment Labor Quality Capital Stock TFP 1950 - 73 Korea 7.S 3.1 1.6 5.9 2.8 Taiwan 9.3 4.3 1.6 7.7 3.5 Japan 9.3 1.6 0.6 8.0 5.5 Argentina 3.8 1.4 0.8 3.5 1.4 Brazil 6.8 2.8 1.3 6.2 2.1 Chile 3.7 1.1 0.6 3.1 1.6 Mexico 6.4 2.5 1.5 6.1 1.9 South Africa 4.6 2.4 1.4 5.5 0.2 1973-84 Korea 7.4 2.5 2.4 10.8 1.4 Taiwan 7.6 3.4 2.5 10.0 1.2 Japan 3.7 0.6 0.5 3.4 2.0 Argentina 0.7 0.9 1.3 3.2 -1.6 Brazil 4.3 3.9 1.8 8.4 -2.0 Chile 1.2 1.7 1.0 1.7 -0.9 Mexico 4.6 3.5 1.4 6.8 -0.6 South Africa 3.3 1.5 1.6 4.9 -0.5

Sources (i) South Africa: modifiedfrom Terence Moll, *Outputand ProducdvityTrends in SouthAfria: Apastheid and EconomicGrowth' (Phdthesis, Cambridge,1990). pp. 110-114;(ii) other countries: AngusMaddison, 7he World Econy in dt 20th Centwy (Paris, 1989)pp. 81, 91.

1.11 Increasing Capital Intensity. Public investment,i.e., governmentinvestment in the 1960s and parastatal investmentin the 1970sand first half of the 1980s, accountsfor most of th-eincrease in capital intensityof the economy. However,the increasesin investmentare associatedwith decliningproductivity of capital and appear as a long-termsecular increasein the capital-outputratio of the economysince the 1960s. Gross capital formation as a percentage of GDP has consistentlybeen at very high levels comparedto internationalstandards. Gross fixed investment(public and private)as a proportionof GDP stayed above 20% for most of the period after 1971, and peaked in 1976 at 28%. After 1976it declined precipitouslyto a low of 18% in 1987, punctuatedonly by a short upturn in 1980-1982. During those two years there was an increase in the capital-output ratios in the private sector from around 1.6 to 2.2 and in the public sector from 4.0 to 6.0. ITe initial difference in absolute levels can be explained by the inherently larger infrastructure component of public investment. However, the sudden increase of the public capital output ratio in the 1970s, concurrent with a decelerating growth in GDP, points to

Chapter 1: Rol ofAgrkuliure in the Macroeconomy Page 5 South AfricanAitricubure: SutrCsure. Perfonnance a Otdons for the Future considerableoverinvestment of the public sector in excesscapital-intensive, infastructural capacity, e.g., in energy, transport, communications,and water. The justificationfor such excess capacitywas found in the perceived strategic importanceof these investments(e.g., the massive investmentsmade in the manufacturingof chemicalsand in a coal-to-oilsynthetic fuels programin order to reduce dependenceon internationalmarkets 5) and to provide a rangeof servicesfor the exclusivebenefit of the white population.

1.12 The increase in capital intensityin the private sector reflects a public sector policy objective- keeping the cost of capital low comparedto the cost of labor. The cost of capital was kept low through negative real interest rates until 1984, the extension of a long-term-foreign-exchange-ratecover to enterprises and the overvaluationof the exchangerate, which lowered the domestic cost of machinery imports. At the same time, the real cost of labor increasedsubstantially. Eir, real wages of white, skilledlabor increasedsubstantially during the 1960swhen rapid GDP growth was associatedwith skilled (white)labor shortagesand a sharp increasein the wagesof white workerswhich outstripped any increase in labor productivityand coincidedwith growing unemploymentamong unskilledblack laborers. This outcomereflected substantial labor marketimperfections, primarily the segmentationof labor marketsinto white and black labor, and the resulting induced scarcity of white skilled labor. Second, while black formal unemploymentnearly doubledbetween 1970 and 1991 (from 27% to 53%), the interracialwage gap at comparableskill levels narrowedfrom about 90% to 14% (largelyas a result of increasedactivity of labor unions); and black wages increasedconsiderably in real terms. During the 1980s, however, significantreal increases in black wages in the formal sector came to a halt.

1.13 Since 19&5,the increase in the capital-outputratio has slowed because of a large decline in investment,rather than an increasein output. The weakeningof investmentsince 1985 resulted from a substantialdrop in parastatal investment,an increase in political uncertainty,and, more fundamentally, a lack of aggregate demand in the economy after a long period of basically unproductive public investment. The lack of aggregatedemand basically reflectsthe fact that the bulk of purchasingpower is concentratedin the hands of a smallminority, whose incomeshave been growing very slowly because of the slow growth of the economyand the increasingtax burden on households.'

IncreasingUnemployment and DistributionalInequalities

1.14 The capacityof the economyto absorb labor has been decliningsince the 1960's, and growth in total employmentdeteriorated since the 1970s. Black unemploymentin the formal sector is currently estimated at 53%. Against the background of decline in black employment,the extremely unequal distributionof incomeand social servicesbetween the black and white populationhas created a highly dualistic society.

5 he governmt is (or was)direaly involvedin theproduction of goodsand sevices thatwere doened of strategicimportance through, interalia, the followingpublic corporations: ISCOR (Iron and Steel Indusrial Corporation-privatizedin 1989),SASOL (South African Coal, Oil, and Gas Corporation),FOSCOR (PhosphateDevelopment Corporation), ARMSCOR (Armaments Corporationof SouthAfrica), SATS (South African Transport Services), and ESCOM(lectricity SupplyCommission).

s Taxeson individuals(income tax, taxeson transfers,interests and dividends,and GSTNVATtaxes) have increasedfrom 3.3% of GDP in 1966/67to 9.9% in 1980181to 15.1% in 1990/91. Taxeson corporationshave stayedapproximately consant around 5-6% of GDP.

Chapter1: Role of Agricuturein the Macroeconomy Page 6 South African Aur_ufrg. Structure. Perfonnance& Ornionsfnr the Future

1.15 Empoym . With the slowdownin economicgrowth came a decline in employment,masked, however, by differencesamong the sectors. Total employmernthas declinedsince the late 1980s;formal sector employmentis now estimated to be less than half of the labor force. However, goverment employmentincreased through 1992, while manufacturingsector employmentdeclined from 1990.

1.16 There is considerablecontroversy about employment statistics. Thetimeseries presentedin Table 1.1 in Annex 1 (registeredunemployment) is closelylinked with formal sectoremployment. As apartheid has acted to segment populationsinto formal and informal employmentstreams, the registered series (A-C) is a reasonablyaccurate indicatorof unemploymentamong whites, but is less so for coloreds and Asians, and much less so for the black population. Nonetheless,the existingdata may be indicativeof a numberof broad trends.

1.17 Formal unemploymentover the period 1975-88, based only on persons reporting joblessness, increasedfrom 0.2 to 1.1% for whites, from 0.6 to 2.0% for coloreds,and from 0.9 to 3.2% for Asians (columnsA-C, Table 1.1 in Annex 1). Unemploymentfigures based on populationsurveys-which are also to question-however, (columnsD-F) show that it tended to rise throughoutthe 1980s, peaking in 1986-87at 15 and 27% for black males and females, 11 and 16% for Asian males and females, and 13 and 16% for colored males and females. Women experiencesharply higher rates of unemploymentthan men. Unemploymentis particularlyserious among youngerage groups. Malesless than 25 years of age represented44% (62%, females62%) of the total unemployedamong coloreds, 68% (57%) for Asians, and 36% (34%) for blacks.

1.18 Distributionof incomeand social services. There are serious incomeinequalities between and withinpopulation groups in SouthAfrica. Althoughinterracial income inequalities have declinedin recent years, declining labor absorptioncapacity and high labor force growth rates have dampenedper capita incomegains. A recent estimateputs 44.8% of the overall populationbelow the poverty line, while less than 50% of the labor force have formal sector jobs (Bureau of Market Research, unpublished information). In 1987, per capita incomes of whites were on average 9.5 times higher than those of blacks. Whites' share of property incomes was around 75%. Gini coefficientsare among the highest found anywherein the world-0.69 for aggregateincome in 1980, 0.82 for farm-landownership in 1989, and a manufacturingsector Gini coefficientof 0.82 in 1982.

1.19 Poverty is worst in the rural areas, with no apparentdifference between the incidenceof poverty in the rural areas of the Republicof SouthAfrica and the homelands. In both areas, the percentageof blacks living in poverty is approximately68%. In the urban areas, the percentageof blacks living in poverty is about 32%, with a low of 25% in the metropolitanareas (Urban Foundation, 1991; Central StatisticalService, unpublishedfamily surveys).

1.20 The unequaldistribution of incomeis mirrored in the discriminatoryprovision of social services. For instance,the white urban areas rank amongthe top four or five countriesin the world in terms of per- capita spending on infrastructure;but taken as a whole, the urban areas have worse facilities than in countries with similar per-capita income. In the main urban area (the Pretoria-Witwatersrand- Vereeniging,or PWV area), nearly 70% of black householdsdo not have direct access to water, and about 50% live in structuresbuilt of impermanentmaterials. In education,public expenditureper pupil was over four times higher for whites than for blacks, and lower for blacks in the homelandsthan for

Chapter1: Rok of Agriculturein the Macroeconomy Page 7 SouthAfrican 4lrculture: Structure. Perfonnance & Optionsfor the Future those in the white rural areas. Electricity is provided to virtually all whitehouseholds, but only to about 20% of black households.

Industrializationvia Import-Substitution

1.21 At the end of the SecondWorld War, SouthAfrica's starting point in terms of humanresources, infrastructure, technology, administrativecapacity, exploitablephysical resources (gold, coal, etc.), financial and monetary services, and supply of low-wagelabor comparedfavorably to that of other developingcountries in purely economicterms; but the country embarkedon a costly, capital-intensive, import-substitutinggrowth path of industrialization.Thus, SouthAfrica failedto exploitits comparative advantage,which would have pointed to the developmentof labor-intensiveexports of manufactured products; instead, through import-substitutionof capital and intermediategoods it followed a capital- intensiverather than a labor-intensiveproduction pattern.

1.22 In large measure because of this choice, South Africa's post-war export performancehas been very poor. Its share of world manufacturedexports fell by more than half between 1955 and 1985,from 0.8% to 0.3%. During that period the value of world manufacturingexports grew by 11.3% per year, while the correspondinggrowth rate for South Africa was below 8%. Its share of total developing country manufacturingexports fell even more steeply-from 12.6% in 1955 to under 2% in 1985 (Moll, 1991).

1.23 For the fortyyears up to the 1980s,South Africa's strategyof import-substitutingindustrialization reflecteda devotionto protectionism,whose instruments included quotas, tariffs, exchangecontrols, and subsidies (van Zyl and Groenewald, 1988). Internationalsanctions provided further justificationfor maintainingprotection, and makingso-called strategic investments. After 1972, quotaswere increasingly replacedby tariffs, and the process acceleratedafter 1983. Despitesubstantial liberalization of trade in the last decade, the system is still subject to frequent change, is overly complexwith a very wide range of tariff levels, and displays a considerableanti-export bias (Fallon et at., 1993).

1.24 To a considerableextent, this policy can be ascribed to the response of governmentand markets to the dominanceof the buoyantforeign exchangeearnings from miningover the years. SouthAfrica was not immuneto the Dutch Diseasethat has affectedother economieswith strong primary export sectors; a comparativelyhigh valuation of the currency was countered by intensifiedprotection of domestic industry against low-ost imports. Large parts of domestic industry received prices that were either directly administeredaccording to a cost-plusformula (e.g., fertilizers)or kept above world marketlevels throughtariffs andquotas which restricted imports (e.g., diesel engines).Mining continues to predominate in SouthAfrica' s exports(although with considerableannual variation) with gold alonetypically providing 30-50%of total merchandiseexports. Agriculturalproducts provided 10% of non-goldexports (R4 billion in 1990).

Public Investmentand Expenditures

1.25 The decline of the South African economy in the 1980s could have happened earlier, but was effectivelypreempted by the counter-cyclicalinvestment behavior of the public sector duringthe 1970s. This stabilized the economyby generatingshort-run demand, but, given the nature of the investments

Chapter1: Role of Agrculturein the Macroeconomy Page 8 South Afrcan Aericubure:Structure. PerformanceA Optionsfor the Future made, did not improve the productivityof the economyin the long run.

1.26 Until the late 1970s the growth path relied heavily on investments, in particular public investments,i.e., by the governmentand the parastatals. Between 1960 and 1976 the share of public investmentin GDP increasedfrom 8 to 15%. After 1976 it started a long declineto 7% in 1990. Until that year nearly half of gross fixed investmentwas undertakenby the public sector, whichaccounted for less than a quarter of GDP`. However, there is almostno relationshipbetween investment to GDP ratios and growth rates over longer periods for the economyas a whole and in its key sub-sectors, i.e., mining, agriculture, energy, gas, and water, etc. As alreadyshown, total factor productivityhas been historically low and has declined in recent years. In manufacturing,total factor productivitydeclined for more than a decadeafter 1972 and recorded only very modest gains after 1983 (World Bank, 1993).

1.27 As a percentageof GDP, total public expenditures(government consumption and investment,and investmentby public corporationsand economicenterprises) increased from 31.3% of GDP in 1978-82 to 37% in 1992. Public consumptionexpenditure has increasedeven faster-from 70.3% of total public expenditureto 89% in 1992. This increasedid not lead to a significantexpansion of public servicesto the non-whitemajority of the population,but rather improvedthe quantityand qualityof servicesto the white minority.

Agriculture's Contributionto the Economy

1.28 Agricultureis widely regarded in SouthAfrica as a highly sophisticatedand efficientsector, and its smallshare of GDP (4.7% in 1991) is seen as evidenceof a steady decline of agriculturalproduction consistentwith a normal pattern of economicgrowth. This section argues, however, that agriculture's current role in the SouthAfrican economyreflects not only the dominantrole of the miningsector in the economyand the series of distortionsthat typifiedboth sectoraland macroeconomicpolicy, but also the subordinationof agriculturalpolicy to social policies- apartheid. Someof the distortionsare not peculiar to agriculture,but characterizethe entire economy,e.g., high capital-intensityof productionaccompanied by widespreadunemployment. At the same time, agriculture's growth in productivity-whilemodest by internationalstandards-has been better than that characterizingthe economy as a whole. A dramatic consequenceof the combinationof increased capital intensity and relatively favorable productivity increases has been a very low contributionby agricultureto overall employmentin the economy.

1.29 The impactof the strategic decisionto cast agriculturein the role of support sector has been far- reaching. Agriculture is generally characterizedby constant returns to scale and an inverse relation between farm size and productivity. In South Africa, however, a minority of white farmers were supposed to own and use more than 87% of the agriculturalland, make the economyself-sufficient in basic agriculturalcommodities, and earn incomelevels comparableto the urban sectors.

1.30 As a result, the agriculturalsector followeda particularlydistorted growth path, at least through the mid-1980's, and the present profile of the sector is very similar to that of the economyas a whole.

7 In manufacturing,about 30% of investmenthas been undertakenby the publc sector. In 1986,before the denationalization of somepublic corporations,about 60% of the total capitalstock was ownedby the pubEcsector. Evae of the non-govenment capital stock,more than 25% belongsto publiccorporations, a ratio that was as low as 8% in 1960.

Chapter1: Role of Agricukurein the Macroeconomy Page 9 Soth Aficon Aerkfuture: Structure.Performance & Oodonsfor the Future

The sector as a whole is characterizedby:

a. a tendencyto assessagriculture's importanceto the economyin terms of its contribution to other sectors rather than directly to GDP;

b. decliningshare of GDP and low productivityof non-exportoriented subsectors;

C. increasingcapital-intensity and declining employment;

d. domesticorientation of the agriculturalsector (exceptfor horticulture) - i.e., towards assuring national self-sufficiencyin basic agriculturalcommodities;

e. public expenditureson agriculturethat displayeda willingnesson the part of the public sector to provide subsidiesto large-scaleagriculture to ensure its financial success;

f. agriculturalpolicies designed to eliminatecompetition from small-scaleblack farmers and to insure factor and input supplies,output marketingchannels, and a price structurethat guaranteesthe profitabilityof the farm.

Tle moortanceof SectoralLinkages

1.31 Agriculture'scurrent and future importanceto the economylies to a large extent in the sectoral linkages that exist, rather than in the direct contributionof agriculturalproduction to GDP. In terms of forward linkages, agriculture has supplied raw materials for an expanding number of agricultura processingindustries that create output and jobs (381,000 in 1976 accordingto van Zyl et at., 1987).

1.32 Agriculturealso creates a derived demand for goods and services through backward linkages. From 1987 to 89, farmers spent R365million for packing materials, R862million for fuel, R938million for fertilizers, and R701 million for dips and sprays. They further invested R926 million in tractors, machinery, and implements, and another R671 million in fixed improvements(fencing, buildings) (Abstractof AgriculturalStatistics, 1992).

1.33 Van Zyl et al., (1987), using input-outputmatrices for the years 1978, 1981 and 1985, have attemptedto quantifythe impactof changes, direct and indirect, in agriculturalproduction and prices on other sectors in the economyproduced by both forwardand backwardlinkages among sectors (Table 1.3). Agriculture has consistently among the highest employment/productionand employment/capital multipliers,and a low income/productionmultiplier (not shown). More generally,the sectoralmultipliers, presented in Table 1.3, demonstratethe decline in employmentlinkages in all sectors that has resulted from increasedcapital intensity.

1.34 For a given increase in agriculturaloutput, the agricultural sector would thus experience the greatest gain in labor employment,but the rise in incomewould be comparativelylow. This largely reflects historicallydepressed agricultural wages. The parametersare usefulfor short-termforecast and for understandinginter-sectoral linkages. The model suggeststhat a 10% increasein agriculturaloutput

Chapter ): Role ofAgricaktre in the Macroeconomy Page 10 sth Afrian Aglu,br Stwfae. PeforMnce A Oprtonsh_r the Future becauseof rainall or technicalchange would affect total GDP by 1.23%.8

Table1.3: Selced Scoral MultNLiers(direct and indirecteffect) Acoordingpg dt Producdon Strutures for the Years 1978. 1981 and 1985

Ae#DIOn Bmolovment A Produedon A Capital Satr (labors per R million) (laborers per R million) 1228 anj 1985 1978 1981 1985 Agricuklur 244.1 136.8 100.9 101.7 65.1 36.0 Gold fmiAng 189.6 88.3 51.1 111.5 58.9 28.4 Fertilzr & PesScAieIndusty 133.1 64.6 43.2 63.4 35.9 19.6 AgriculturalMacinery 127.9 58.8 43.9 98.5 58.8 33.8 lecoticty, stearn& gas 94.2 56.1 29.2 18.5 10.0 4.5 Construcon 263.1 140.6 82.8 154.8 93.7 51.8 Trade 140.1 78.2 46.5 82.4 48.9 27.4 Transport 122.7 73.7 45.2 30.7 19.4 10.3 Services 27.7 16.3 6.4 12.0 7.1 2.9

3Oure: van Zyl et sL, 1987.

Decliningshare of GDP andlow productivityof non-exoortoriented subsectors 1.35 Overdie past cenurySouth Africa's economy has slowlychanged from one dependenton mining and agricultureto a broadlydiversified manufacturing and serviceseconomy. Agriculture, forestry, and fishing, which represented21.0% of GDP in 1911, declinedto 4.7% in 1991, growingat rates consistenlybelow the averagegrowth of the economyCable 1.4). Mimingand qurying fell from 27.7%to 10.4%over the sameperiod. Manufacturing,which represented only 4.0% of GrossDomestic Product(GDP) in 1911,rose to 24.9%of all economicactivity in 1991. In othercategories of economic activity,the finance,insurance and real estatesector grew rapidly in the 1960s,but has sincetended to movewith the generaleconomy. 1.36 However,the abovepattemn of a decliningshare of agriculturecannot be interpretedas a totaly

Intial resuls of a recet anysi based on a 1988Socal AcoountngMatrix (SAM) paint a somewhatdiffrnat pcture. In the SAM anaysi-which considersadditional rounds of speadng, whereasan input-outputanalyis only consders intation tte,itu producton exdibit someof the loest employmentmulns of the entireeconomy. Howe , imptat agribusinessproducts such as food, laterand textiesrank very high in terms of employmen generated (EBet,intenul DSA memorandur, 1993). An analysisthat used a 1985input-outpt matrx and focussedon backwardand forwardlinkae found that thebackwad linkages of agr u conmparedvery fivorably to bakward inagesof non-agiultura avi in tms of inoomeand empoymen genaed. Forwardlinkages of agrulure comparedless fvorably to othersecto However, agribunssas a sod-forward and backward liaes of agricuail productin combined-madesubantia contbuns to theconomr: it contributed15% to total valueadded and 26% to total employment(van Svaeter, Faux, and van Zyi, 1992).

Chapter1: Rok of Agrcukarwin the Macroeconomy Page11 SouthAfican Agrrckuture:Structure. Performance & Oxionsfor the Future normal pattern of economicgrowth. First, in contrastto the normal patternof developmentunder which industrializationproceeds by extracting agriculturalsurplus, South African agriculture has been a net recipientof t.hesurpluses generated by the mines (Brand, 1969;Merle Lipton, 1985);throughout the 20th century, white farmershave benefittedfrom substantialfinancial resource transfers in order to maintain their viability and to ensure farm income levels comparable to urban levels. Second, international comparisonssuggest that South Africa's agriculturalshare of GDP and, in particular, employmentis rather low even comparedto other economiesin roughlythe same categoryof economicdevelopment- i.e., the class of middle-incomeeconomies (Lipton and Lipton, 1993). Whereas the low share of agriculture in GDP is pardy the result of the dominance of the mining sector, its low share in totl employmentis mainly the result of the existence of the very skewed land-ownershipdistribution and production activitiesdominated by ranching and large-scale, mechanizedfarming.9

Table 1.4: Comoositionof GDP since 1911. by sector (%I

Agriculture, Mining& Manufacturing Commerce, Other* Total Foresty, Quarries Catering, Fishing Services 1911 21.0 27.7 4.0 47.3 100 1920 22.2 18.3 7.3 15.6 36.6 1o0 1930 14.2 15.6 9.4 15.1 45.7 100 1940 12.7 18.7 12.4 13.9 42.3 100 1950 17.7 13.5 16.4 14.0 38.4 100 1960 12.4 13.7 20.5 14.2 39.2 100 1970 8.1 10.0 23.2 15.1 43.6 100 1980 7.0 22.0 21.7 11.7 37.6 100 1990 5.0 10.7 25.6 13.5 45.2 100 1991 4.7 10.4 24.9 13.5 46.5 100

Note: Namibiaexcluded from 1984. * Includes: electiciy, gas, water, construction, transport, storage, communication,finance, r estate, business services, community,social and personal services. Sourc Abstract of AgriculturalStaises, p. 81; RSA 1992 (a); Table 75.

1.37 The low productivitygrowth that characterizedthe agriculturaleconomy was also evident in the

' South Africa's doubly compresed agrcultual seceorcompares with those in countriessuch as Jordan, Uuguay, and Ven=la-countries with smilarly sewed land-ownershipdistributions and highly capital-intive (Jordan)or land.xtei agdiulturl seors (Uruguayand Venezuda). Jordan's agricultureprovides around 8-9% of GDP and employment. Its natual resourcebase is weak: 91.5% of the land arecareeives less than200 mm of rainfall. Howevaer,agriculAtm is highy capita- itnsive, land onersip is highlyskewed, and manyhfma are non-owneropaed and rey heavilyon agriultul guesiworks (World Bank, 1990). Uruguayanagriature provides11-12% of GDP and employment.TMe sector is dominad by large-scale, exve livestockproducion and arabi land is subsantitay underutilized.Shareroppi is a significantfeature of Xt cropped areas, while an inreasigy marginaid rural populationpresents growing economic and socialproblems (World Bank, 1988. Venezuelanagriulm provides5-6% of GDP and employment.Land ownerhip reman very concentraeddue to an ineffective land reformprogram. Over 80% of the land is ownedby 7% of the landholderswho mainlyuse the land for grazingpurposes (WorldBank, 1992).

Chapter1: Rok of Agriculturein the Macroeconony Page 12 ,ushAfiA Atbu: aC% . PaEM= & for thu Fmaure man agricur sub-sectors. From 1947 to 1991, total fctor productivity (UP) of South Afican agdcultue increasedby 1.08% per year-a relativelylow rate by internationalstandards. Ibis aeate, however, masks substaa differencesamng the sub-sectors. Over the same period, TFP in the epo- oriented and inheaey labor-iesive hoticultural sector grew at 2.38% anually, whereasthe more inward-looking and capital-Intensivelivestock and fieldcrop secors both recorded very low annual productvity growth of only 0.70 and 0.72% respectivelyover the same period. From 1985 to 1991 agega TFP growth of agricdtre acceleratedto 2.85% per year as a reWlt of policy reforms th increased the cost of capital relative to labor and considerablyslowed down non-labor input growth (Thirtle, Sartoriusvon Bach and van Zyl, 1993).

Figure 1.2: Tot Factor Product-Agriral Sub-Sectors(1947-190

280 260 . 240 &220 ,!200

160 140/.7 120 j- / 100 80 vt ~ e_ 4<

47 49 51 53 55 57 59 61 63 65 67 69 71 7375 77 79 81 83 85 87 89 Year

Source:Ihirtie et al., 1993.

IncreasinizCapItal-intensIty and DecliningEmployment

1.38 The increasedcapital hieity of agriculturewas accompaniedby a declinein employmen in dte last two decades. Agriculral employmentincreased over the period 1951 to 1970 by almost65%, then declinedover the period 1970 to 1991 by 60% (Table 1.5). Thus while agriculturecontinues to make an importantcontribution to total employment,its share is declining. In 1951, it provided employmentfr nearly 33% of the economicallyactive population, and 31% in 1970. By 1991, however, agriculture

Chaupr 1: Rolt qf4A r# In *h Macroconamy Page 13 SouthAfica Agrkyiure: Structure.Performnce At ODrlOnSfor the Futur employedless than 14% of the economicallyactive population. However, these data excludethe TBVC areas and the unrecorded employmenton white-ownedsmallholdings t0 .

Table 1.5: Enmlovmont.Economicallv Active Pooulation ('000) Year Agriculture, Mining ManufacturingCommerce Services Other Not Employedor Total Ecoomic Forestr, & & & Fnance Not Classifiable Populstion Fishing Quanying 1951 1,509 510 502 328 1,074 468 202 4,593 1960 1,687 615 644 643 1.137 509 486 5.721 1970 2,482 680 1,026 898 1,596 861 571 8.114 1980 1,306 836 1,465 1,298 2,005 969 810 8,689 1985 1,180 743 1,380 1,281 1,965 1,067 1,076 8,692 1991 1,005 680 1,260 1,663 2,339 985 2,282 10,215

Note: Table excludesTBVC ama. South African Statistics,1990, p. 7.5. South AfricanStatistis, 1992, p. 7.7.

1.39 Since 1970, mechanizationhas also causeda reductionin employment,while rapid wagegrowth in the non-farmsector has pulled workers off farms. Althoughfarm employmentincreased from 1983 to 1987 because of higher real costs for machinery,high debt loads, and reduced farm subsidies, it has resumed its decline, possiblyas a result of increasingpolitical uncertainty.

1.40 Capital OUtputratios in agriculture are twice as high as in manufacturing,with little apparent economic justification. The increased capital intensity in agriculture, concurrent with a substantal displacementof labor after 1970, started with the introductionof tractors after the SecondWorld War. Since the late 1960s, a number of other stimuli, includingadministered producer prices for inputs and outputs, further increasedagriculture's capital-intensity,resulting for instance in heavy Investmentsin combineharvesters in the main grain producingareas. Negativereal interestrates for agriculturalloans, coupled with double digit inflation, appearedin the early 1970sand largely explain the over-investment in agriculturalcapital throughout the 1970sand early 1980s. This patternof agriculturaldevelopment ran counter to the overall factor endowmentof the economyand appears to have been costly in economic terms (e.g., de Klerk, 1983). Furthermore, in some of the main agricutural sub-sectors,the pattn of over-capitalizationwithout efficiencygains has contributedto the insolvencyof a substantial nber of fanns under the weight of debt service.

1.41 Favorable taxation of agriculture constitutedan importantdistortion in the policy environment affecting agriculture. lTe absenceof a land tax or capital gains tax attracted investmentfrom the non- farm sector. High inflationsince 1970 also made investmentin rural real estate attractivefor investors,

I lhe latterare esaed to number53,000 and cove 0.8 millionha. Around30,000 altholding an usodfor reddal purpose.

Chapter1: Role of AgrkuAlurein the Macroeconoy Page 14 SouhAfran Agridture: Structure.Perfonnance & Ontionsfor the Futgre and tax Incentivesassociated with acceleratedcapital write-off rules provided further incentives. The Introductionin 1970 of a 100% capital write-off provision in the year of purchase was an important modvator in the rapid spread of combineharvesters (de Klerk, 1983). Different tax arrangementsfor differentsubsectors within agriculturecreated intrasectoraldistortions, e.g., betweenlivestock and game farming, and between forestry and sugar production. The effectivering-fencing" of parts of agriculture in the recent past bas again changed the tax incentivestructure.

1.42 Negative real interest rates constituted one of the most important mechanismsto stimulate agriculturalinvestment. Adoptionof preferential interest rates for the agriculturalsector as an active instrumentof monetary policy has had important implicationsfor agriculture, principally through the direct impact on the cost of capital and, but also indirectlythrough the costs of manufacturedinputs. Moreover,starting as far back as 1970, long-terminterest rates for landhave been substandallysubsidized by the Government12 .

Domestic Orientaton

1.43 The import-substitutionstrategy had a profound impacton SouthAfrican agriculture. One of the .ainobjectives of the sector, to reduceexternal dependence in basicagricultural commodities, was largely achievedby the 1980s. Agriculturalproduction increasedat a rate of 3.1% annuallybetween 1955 and 1986, while populationincreased by an average rate of 2.8% (van Zyl et al., 1987). Self-sufficiencyin aggregateavailability of the major domesticallyconsumed agricultural commodities (i.e., maize, wheat, dairy products, vegetables,potatoes, and sugar) was attained, while considerabledomestic surpluses-in spite of periodic droughts-in such crops as wheat, maize, and sugar were availablefor export. However, protectionfrom internationalcompetition lowered the efficiencyof importantagricultural sub-sectors, and the domestic orientation of large parts of the sector amplified the negative impact of declining GDP growth in the 1970s and 1980sthrough its impacton demand.The growth of aggregatedemand for farm productswas also considerablymoderated by decliningper capita incomes". Moreover, in spite of the

It Rngfncing rers to tx pvos ha do n tax payes to rdheir taxablek m in oneacdvity by deduing m _ntrs de inmnothe activity in which more favorabk tax raes apply. Th absenc of such promn often made agriul imvesmensatrcti to inveso.

12 Three iret re time series fiom 1965to 1990are preseted in Table 1.4 in Annex 1. BAChseries is givenin noMinal and real terms (nominal ae lss the inflationrate). The lendingrte (D) is the cost of short- to mediumtermfinancing to the privatesector. The govanmaetbond yield (E) is the cost of long-termgovernment borrowing. The Land anklong4rm interest rate (F) is the rate of inerest on long-termloans to fannms,prindpally for the purchaseof farmland. Presumably,the Land Bank shouldnot be a better guarator of credit thn the govemrnmentitself. Hence,one would expectthe land Bankrate to exceed,or at best equal,the Bondmte. Sinoce1970, however, long-terminterest tes for land have recived fairly sbantial subsidizeon by the Govemment,as refleced in a comparisonof columnse and 0. Table 1.5 in Annex1 showsthe trendsin real interes rates chargd to farmersby three of the main sourcesof furmcredit- the Land Bank, the agctura cooperativesand the commercialbanks-in the periodsince 1955.While these rates were positive up to 1972,they remainedbelow the inflationrate for the nextdecade, i.e., whenthe inflationrate movedinto double digit figures. Bdween 1985and 1987 real rates wore again negative,and have been positivesince that time. The effect of these inest rate policy changescan also bz seea in the movementsin landprices in this period.

" However,the decreasinginaia wag gp mayhave offset this trend somewhat,given the higherincome easicity of demandfor foodof the poorer, i.e., black, segmentsof the gnal popubtion.

Chapter 1: Role of Agricuktre Anthe Macroecononmy Page 15 SouthAfrican Agriculture: Structure. Performance & Options for the Fuiure aggregate food self-sufficiency achieved in production, there is little evidence of increased food security for poor, black households.

1.44 Agricultural exports have traditionally contributed between 5 and 10% of South Africa's total export earnings 4-not to be discounted, but a relatively small sbare. Moreover, exports of domestic surpluses of maize, wheat, and sugar, often incur losses, given prevailing world market prices. The main domestic deficits are in red meat and oilseeds, which have to be imported on a regular basis, along with a significant proportion of agricultural inputs (especially chemicals and machinery).

1.45 Tight protection against foreign competition, in conjunction with other government interventions such as the controlled and largely monopolistic system of marketing boards and subsidies on inputs and interest rates, has also contributed to distortions in the sector. In agriculture, protection of the domestic sector was generally through quota restrictions (now replaced by tariffs) administered under the large number of monopoly marketing systems. The reduction of some of these distortions during the 1980s has exposed the financial and economic weakness of parts of the sector.

1.46 The exception to the domestic orientation of South African agricultural is the horticulture sub- sector, which holds an international comparative advantage in production. Its inherent labor-intensity may have prevented an exaggerated replacement of labor by capital. In addition, the horticultural sub-sector benefitted greatly from the currency depreciation of the 1980s. In particular, when international sanctions caused a stagnation in the quantity of horticultural exports their value increased substantially because of concurrent sharp devaluations of the Rand.

1.47 The direct impact of exchange rate variation on agriculture depends on the degree to which (a) inputs used are imported, or, if locally manufactured, have an import content, and O) output is tradable or affected by the prices of tradable goods. It is important, in this respect, to recognize that there are substantial variations among the various agricultural subsectors, based on prevailing production technologies, marketing arrangements, and crops grown. In the context of South African agriculture, real currency depreciation will increase the relative profitability of producers of tradable goods (e.g., horticulture), but tend to exacerbate the cost/price squeeze for producers of those commodities which are not generally tradable or are only traded when there is a domestic surplus. Moreover, depreciation is likely to be an effective instrument to reduce the foreign exchange content of agricultural production systems, especially if strengthened by changes in import protection for farm inputs.

Public Expenditures

1.48 Before the Second World War, significant public investment in the agricultural sector was often

" Timeseries data reportingtrends in agriculturesshare in foreigntrade, are not yet regularlypublished in officialbulletins. Data on iuports of productioninputs (fuel, chemicals,machinery, and equipment)are also not readilyavailable in time seriesform. It is thereforedifficult to analyzewhether agriculture is an efficientuser of foreignexchange. In the 1960s,the agriculturalsector in South Africa appearedto be a relativelyefficient user of foreign exchange,compared with other sectors. One esdmate (apparnntlynot more recentlyupdated) found that agricultureused RO.13 worth of importsto produceRl.00 of output,compared with RO.73for textiles,and RO.64for machinerysectors (8rand and Tomlinson,1966). However, it is doubtfulwhether such relativeefficiency of foreignexchange use has continueduntil now.

Chpter : Roleof Agriculturein theMaoroeconony Poge16 &h Afic re: Strcture. Performance& Optionsfor the Future justified by what was referred to as the 'poor white problem'. Substantialfinancial assistance was provided to white farmers in order to keep their incomes in line with incomes in other sectors of the economy. After the SecondWorld War, the general import-substitutionstrategy provided an additional rationale for public assistance to agriculture. The objective of attaining self-sufficiencyin basic agricultural comnmoditiesassumed even stronger strategic importance when threats of boycotts and sanctionsemerged in the 1970s and 1980s.

1.49 In the 1980s, real public expenditureson agricultureamounted to about 1.2 billionRands per year (1985 terms). Table 1.6 shows trends in budgetaryexpenditure on agriculturefrom 1979to 1989. The totals representthe combinedbudgets of the various departmentsof agriculture, includingthe homeland departnents, but reflect a narrow definitionof agriculturalspending. In real terms, public expenditures declinedin recent years becauseof tighter budget constraintsbeing experienced,the high budgetarycost of exporting surplus grains at a loss, and the relative decline of the political power of white farmers. However,the downwardtrend mustbe interpretedwith caution,because a sometimesunknown proportion of expenditurewas moved off-budgetin this period. In addition, certain budgetary expendituresoccur at irregular intervalsand are difficultto assignto one particularyear. For instance,the 1992government budget included about R2.85 billion in guarantees and loan write-offs for the commercial sector. Moreover,subsidies granted for certain agriculturalcommodities are difficultto apportion to producers alone, since consumersalso benefitfrom subsidies. Such subsidiesaveraged R370 millionper year from 1985186to 1991192(Abstract of AgriculturalStatistics, 1992, p.1 12). While budgetaryexpenditure on black agriculture in the homelan.s increased during the period under consideration, a very small proportion represented transfers to farmers; more than 95% of these funds were absorbed as administrativecosts (Vink and Kassier, 1991).

Chapter1: Rok of.4grlcadureIn the Macroeconomy Page 17 SouthAfrkan Ag_ricu_ure:Structure. Performance A Optionsfor the Future

Table 1.6: 'Me Growth in Budketed Expenditure. 1979-1989

Year Total Agriculture Total White Agriculture Total Black Agriculture (real) (real) (real)

1979 785,130,197 636,558,114 148,572,083 1980 1,00S,186,491 830,514,451 174,672,040

1981 983,493,870 795,944,649 187,549,221

1982 998,875,232 790,514,140 208,361,092 1983 1,326,134,968 1,095,133,593 231,001,375 1984 1,538,051,591 1,253,082,558 279,657,405

198S 1,478,765,538 1,200,347,000 266,073,538 1986 1,116,495,578 833,789,207 262,799,119 1987 1,548.339,178 1,234,704,430 293,939,033 1988 1,225,377,126 840,874,517 359,010,974 1989 981.802,742 613,672,085 347,650,836

Average Annual 2.3% -0.4% 8.9% Increare

Source de Klerk, 1991. Agrcultural Policies

1.50 This discussionof agriculturein the economyshows that the sector's limited role is mainlythe result of the dominanceof mining, the skewed availabilityof natural-resourcesin favor of the white minority,the high capital-intensityof the agriculturaltechnology used, the reductionof employmentthat accompaniedthe adoptionof this technology,and the domesticorientation of the sector that resultedfrom the objective of achieving self-sufficiencyin major agricultural commoditiesat the expense of other economicconsiderations. These features are, in turn, the results of the implementationduring the past century of several highly interventionistagricultural policies that were closely intertwined with the country's general approachto the economy, society, and politics. (Chapter 3 discusses this history in some detail.)

1.51 In the 20th century, agriculturalpolicies movedthrough three distinct phases.

1. From the Land Act of 1913 to 1948. During this phase the basic institutional frameworkof a dualistic agrarianstructure was established.

2. From 1948 to the 1980s. This phase-with the introductionof the tractor and other machinery-saw the transformationof the sector into a highly mechanized and capital-intensivefarm structure.

Chapter1: Rok of Agricuhurein the Macroeconomy Page 18 SouihAfrcan Airicuture: SMructure.P-EIrmacw & Onions for theFuttre

3. From the structural changesof the 1980sto the present. This phase was marked by the movetoward deregulationof financialand agriculturalmarkets and toward a more open political economy, e.g., in 1991, the Land Act and related legislationwere scrapped.

The third and current period has been characterizedby the followingdevelopments:

a. a general move to more market-relatedinterest and exchangerates;

b. an administereddecline in the real producer price' of key commoditiessuch as maize and wheat;

c. an extensivederegulation of control under the MarketingAct, leading, e.g., to the first voluntary dissolutionof a Control Board in 1993s5;

d. a reductionof public financial assistanceto commercialfarmers;

e. changes in tax policies, includingphasing tax write-offsover three years instead of one, thus reducing the implicitsubsidy on capital investment;

f. increasingpressure to apply farmworkersthe same labor legislationas in non- farm sectors, resulting in the applicationof the Basic Conditionsof Employment Act to farm workers in May 1993;

g. the scrappingof the Land Acts and other related legislationand regulationsin 1991, followingthe publicationof the government'sWhite Paper on Land Reform and the resultingpublic debate;

h. changes, throughout this period, in the policy approaches to agricultural developmentin the homelandsaway from a top-down, estate-farmbias; and

i. initiationof the processof amalgamatingthe 14 existingministries of agriculture into a single department.

1.52 Thus since the early 1980s there have been many macroeconomicand sector-level policy reversals, including scrapping of the major racially-definedland laws, reduction of subsidies on agricultural interest rates and public financial assistance, decline of administeredproducer prices and deregulationof marketingenvironment. These have set in motion a series of structural changesin the sector, whose aggregateoutcome is not yet clear. The impacton farmers has varied accordingto region and commodity.While significantgrowth in agriculturalproductivity has accompaniedthe structural changesin the past decade, the effectsof policy changeson employmentare ambiguous. Moreover,the highly-distortedownership structure in agriculturehas remainedin place.

' TheBanana Board, later folowed by theEgg Board, ChicoryBoard and Rooibos Tea Board, and since then, potatoesand dry beans.

Chapter1: Roe of Agrcukure in the Macroeconomy Page19 South Arican Agriculure: Structure. PerformanceA Oprios fr the FEtgrg

1.53 The economicreforms have made the agriculturalsector more flexible, but lack of progress in related sectors may negate many of its benefits. This disjunctionfrom policy in other parts of the economy is a distinguishingfeature of the deregulation of agriculture in South Africa. For instance, marketingpolicy adjustmentsin industrieswhich either provide inputs to farmersor process and distribute farm products have only recently become part of the agenda of reform. Financial policy reforms In agriculturestarted the process of adjustmentin the sector; but while interestrates are now largely set by market forces, exchangerate reform came to an end in 1985 when the financialRand was reintroduced. It can be argued, therefore,that fiscal policy during the 1980s-includingmoving substantialagricultural expendituresoff-budget and the managementof the fiscal deficit-hurt the economy as a whole, and the farm sector in particular.

1.54 The lack of a comprehensiveand coordinatedstrategy of economy-widereforms has thus created or left in place a number of distortionsin the agriculturaleconomy. First deregulationof the marketing system has resulted in a rapid growth in informal marketing,exposing the inefficienciesof the formal marketingsystem. Althoughthe extent of these markets is not yet known, official data alreadyshow a sustaineddecline in the absolute volume of officially recorded sales of food. Secon the continuation of regulation in input supply, processing, and distribution has created further distortions in the farm sector. Becauseof the still overly rigid marketingenvironment, the sharp drop in real producer prices for some key commoditieshas not been matchedby similar movementsin retail prices. The sustained increases in retail prices had an adverse impact on efforts to squeeze inflationout of the economy and have hurt consumers. Similarly,the lack of trade reform and competitionin the input supply industries has squeezed farmers' profit margins. Even the highly lucrative fruit export industries will find their margins under threat if exchangerate movementsand deregulationof input suppliers is not attendedto. Third, inflexible resource and product markets in agriculture make rural restructuring all the more difficult. Despite the efforts at reform, the legacyof past policiesis such that the sector remains highly capital intensiveand has not been able to fulfill its potential in terms of employmentcreation.

Chapter1: Rok of Agrculturein the Macroeconom Page 20 SouthAfrican Agricuture: Structure, Perfonnance& Optionsfor the Future

CHAPTER2: STRUCTURE.RESOURCE ENDOWMENT. AND PERFORMANCEOF THE AGRICULTURALSECTOR

2.1 South African agriculturalproduction dwarfs that of the rest of Southern Africa. Despite a relativelylimited natural resource base, its agriculturaleconomy has reachedhigh production levels in a wide range of agriculturalcommodities compared to that of other countries in the region. Such high performancecan be directly attributedto its sectoralendowment: a widevariety of natural resourcesand agro-climates;the quality of farm management,extension, and research; and a century of investmentin agricultureper se, as well as in rural and agriculturalinfrastructure.

2.2 However, the ownershipof this resourceendowment is extremelyskewed--the legacy of centuries of racially-basedpolicies. The white, commercialfarmers own and farm virtually all of South Africa's agricultural lands, have subsidizedaccess to water resources, receive a range of direct and indirect subsidiesand tax concessionsand massive financial assistance, benefit from a highly-developedrural infrastructure, and are supported by an elaborate service sector which includes banks, co-operatives, marketingboards, research institutes, and extension services. On the other hand, the homelands into which the rural black populationhas been driven are often so overcrowdedthat commercialagriculture is frequentlyinfeasible. However, cattle-historicallyan importantcapital assetfor blacks-continuesto be of considerableimportance: one third of South Africa's herd can be found in the homelands. A small minorityof the inhabitantsof the homelandsfarm smallareas of land, mostly at below-subsistencelevels and with only usufruct rights. Many of these farmers are women or elderly persons who have only limitedinstitutional support and are poorlyserved by agricultural,transport, communications, health, and water supplyservices. This seemsto be the intendedeffect of a century of policies designedto provide cheap labor for both the industrialand the agriculturalsectors.

2.3 SouthAfrica prides itself on havingdeveloped exceptional human resources-in terms of the skills of its farmers and research, extension, and cooperativestructures--to compensate for what is commonly assumedto be a relatively fragile and poor natural resource endowment. Closer analysis, however, reveals a somewhatdifferent picture. Three points need to be made with regard to SouthAfrica's natural resource base:

a. AlthoughSouth Africa's resource endowmentis limited, it is also highly varied, a fact that can be regarded as an importantresource for the sector;

b. Internationalcomparisons show that agriculture'slow contributionto the overall economy of SouthAfrica cannot be explainedby the presumedlimitations of its natural resource base; and

C. Current patterns of natural resource use have resulted from the particular structure of incentives-i.e., property rights, prices, and institutionalarrangements. The incentive structure likely to prevail in a post-apartheidSouth Africa will show that many current natural resource uses are highly inefficient.

Chapter2: Structure,Resource Endowment, and Performanceof theAgricultural Sector Page 21 SouthAfrican Aerkcubure: Structure, Performawce A Options for theFuure

2.4 This chapterdocuments the endowmentand impressiveproduction performance of agriculturein South Africa. It is important to acknowledgenot only the volume of agriculturaloutput but also the impressiveinstitutional support mechanismsthat have been availableto large-scalewhite farmers. Among the central challengesfacing a new governmentwill be to design a rural restructuringprogram that can address equity issues in the ownershipand accessto resourcesin the rural economy, while ensuringthat the production potential of the sector is maintainedand productivity, income and employment are increased. At the same time, a strategy for reorienting the present organizationalinstitutions to the requirementsof a restructuredrural economy must also be examined.

Sttucture

2.5 Land distributionin SouthAfrica is amongthe mosthighly skewedin the world. Four centuries of conquest, occupation, annexation, Trust land acquisitions and transfers, land purchases, and consolidationhave resulted in a land distributionwhich is comparablein its inequalityto those of many Latin Americancountries that often experienceda similarhistory of European conquestand settlement. What sets SouthAfrica apart from the rest of the world, however, is the relative emptinessof much of its rural landscape-except for some high-potentialareas along the eastern seaboard. In most of the country, rural villages, settlements,scattered farms and homesteadsof farm laborers such as those found in the rural areas of Europe, Latin America,Asia, and elsewherein Africa are rare. Approximately86% of agriculturalland (85.8 million hectares) is held by 67,000 mostly white farmers, supportinga rural populationof 5.3 million. Consequently,the whiterural areas-the platteland-havea very low population density(about 16.2 hectaresof agriculturalland per rural resident). Conversely,the level of urbanization in SouthAfrica's formerlywhite areas is relativelyhigh-certainly by African standards. Around74.5% of tihepopulation of the Cape, Natal, Transvaaland the OrangeFree State live in the urban areas (Fable 2.1).

2.6 By contrast, a rural populationof about 13.1 millionresided in the homelandsin 1988,on an area of 17.1 million hectares. Thus 29% of South Africa's rural population(mostly farm workers and their dependents)lived on nearly 86% of the agriculturalland, while 71% of its rural population lived in the remaining 14%.

2.7 Agriculturein South Africa is completelydominated by the largely white-owned,commercial sector, whichproduces 89.8% of the value addedon its 86% of the agriculturalland. The whitefarming sector is commercially-oriented,capital-intensive and generally produces a surplus. In contrast, the farming sector in the homelandsis unable to meetthe subsistenceneeds of the homelandpopulation, with the result that the homelands are net food importers from the white sector. Income from agricultural production in the homelands constitutesno more than 20% of household income. (For example, in Kwazulu, remittances,pensions, and other incometransfers account for more than 80% of homeland householdincomes (Bekker,Cross, and Bromberger, 1993)). Thus, the homelandsshould be viewedas labor reserves-not even as the subsistence sector of a highly dualistic agricultural system. Note, however, that as a share of value added in the private sector of the homelands, agriculture plays a substantialrole.

2.8 A range of services provided by specializedprivate sector service institutions,the Land Bank,

Cmaper2: Strre, ResourceEndowmen, and Perfonnance of the Agricuural Sector Page 22 SosahIAfia Anriuhure:Struture. Perfonnance & Optionsfor the Future

AgriculturalMarketing Boards, Cooperatives,and research institutions,aided by a strong politicallobby, supports commercialagriculture. Subsidies,tax concessions,and tariff protectionhave alsolong benefited the sector. Smallholderfarming, on the other hand, operates in an environmentmainly devoid of these services and subsidies. These farmers have restricted access to support services and opportunities. Decisons on labor allocation, production programs, livestockfarming, and choice of technologyare largely governed by householdeconomics and risk-minimizingstrategies involving part-time farming.

2.9 Governmentaid to the two sectors has been unequal and variable. Accordingto Lipton (1977), the ratio of total governmentexpenditure for white and black agriculturebefore the SecondWorld War was about 179 to 1, around 14 to I in the 1950s, and 2 to I in the 1970s. However, accordingto Vink and Kassier (1991), this ratio again worsened to around 4 to I during the 1980s. In 1987, the 14 Departmentsof Agriculture1' had a combinedbudget of R2.1 billion, of whichR1.7 billion was allocated to white agriculture". Moreover,white farmersobtained 96.7% of all transfer paymentsin the budget.

2.10 Despitethis massivepublic assistance,substantial numbers of commercialfarmers face financl difficultiesbecase of recurrent droughts, over-capitalization,high debt burdens, surplus production, a cost-pricesqueeze, and over-intensiveor Inappropriatecropping systems. At the same time, the black small farmer faces problems related to extreme populationpressure on very limited land, lack of access to markets, insecure and restrictive land rights, an overstockedand damagedland base, and the lack of infastructure, water supplies, transportationnetworks, financial support, and research and extension services (Lipton, 1977; van Zyl and van Rooyen, 1990).

2.11 One indicatorof the severity of the land shortagesin the reserves is given by the ratio of arable land to rural population. Only in the case of two homelands did the ratio exceed .20 arable hectares/resident-KaNgwane(.25) and Bophuthatswama(.27). Five of the homelands registered .10 arable hectares/residentor less. By way of contrast, the ratio in the white areas ranged from 1.37 hectares/residentin Natal to 2.87 hectares/residentin the Cape. The extreme densities recorded in the homelandsseverely constrain agricultural development potential in these areas, and also accentuatestheir role as basicallylabor reserves.

'As of Aprl 1993,South Aficnm fams ae servedby 11 Departmeatof Agriculture,down from 14.

"Th. gaoS valueof arculua producdonwas R13.4 bilion in thesame yar.

Chrapt 2. Srucnwe, ResourceEndoe, and Pefomuanceof the AgriculturalSector Page 23 Table2. 1: LAWUse anl.ghi

1988ToWa RFanfed TOl Irrigaed Towa Total ToWa 1988Urban Rwul Rura pop ArabecLAnd Per Surface Ai"rac Rainfod Alab"c nulgaed ArMac Arabic Populaton Poqulaion aa S of Rur ReaidewA i Ae Om) Land (U Ar ((W2) Land (5 Ana Land (S) Lan (000 tota (wmiaa) of $0o1 ofM (kobkna) : popbtion CAP 642,467 4.52 2910.0 .10 7.067 S.62 36,107 4,815 S 749.3 2.57 NIWzt 55407 1S.85 8,782 3.S IM 19.00 10,37 1.771 78S 69.3 1.37 Taanumnul 223,44 20.3S 45480 3.14 7,012 23.10 6204 7,757 222100 4.8 2.39 O.F.S. 127,29 28-80 36, 0.79 1,006 29.9 340666 1249 1,093 93.3 3.45

KaN c 3.9171,048,6 12.30 1194962 1.63 176079 13.07 137,041 67,S92 S462 boa 74.7 2.26 I 2.1u36,0763 10 4 10.00 3.607 15104 3.630 23.3 0.10 Guawu 7.484 6.40 479 1.70 127 8.10 606 29 640 4.3 0.08 t Kwandebele ~2,208 - 1S.41 340 36 327 9.9 0.10 KaNgwone 3,917 12.30 482 17.10 670 29.40 1,IS2 67 462 12.7 0.2S L dbf 22,137 0Q60 1313 IS.I0 3,343 ISM6 3,453 1S4 2,310 6.3 O.IS QwMM 1,040 19.00 18^ ) 34 228 13.0 0.08 Sublgl 72,860 - 12.U3 9.346 1,424 7.597 15.8 0.12 tBephum 40.011 - - - - 10.81 4.325 294 1.530 IS.7 0.27 Cidlei 8.100 - - - 8.308 672 278 S12 35.2 0.13 Tranakei 43,654 - - * 4.67 2,039 168 2.877 5.5 0.07 Venda 6.807 0.62 42 10.3S 2,873 10.97 747 18 490 3.5 0.15

&572 * * - 7.90 7.783 758 oGnjd gflj 1,220G88 5.4S9 12.2 0.14 -- 12.64 IS4.170 17,774 18.400 49.1 0.84

Sou: a. DBSA. 1990. A Regioal Profle of lbe Sou_rn African Population. pp 2548. b. Data on totl wurfec ae ad rabb lad an taken from DSA. MI991,An lr-Regioa Profie p. 34. SouthAfrian Aerikulture:Struture. Peffonnance & Options for theFuture

AFrarian Structure in the White CommercialSector

2.12 Beforedescribing the agrarianstructure in the white commercialsector, an importantcaveat needs to be made. Reliabledata on farm size exist only for 67,000 large-scale,predominantly white-owned farm units. Such units are situated outside municipalboundaries and may actually consistof more than one farn, e.g., two farms in the same district. In the following, most of the descriptionsof the white conunercial farm sector will focus on these farm units, since only limited informationexists on the estimated 56,000 smallholdingswithin municipalboundaries. These holdingscomprise a total area of 800,000 hectares with an averageholding size of 14 hectares.

2.13 In 1988,a total of 67,000 farmersin the white commercialsector controlled85.8 millionhectares of land and employed 1,144,700 casual and permanent workers. These farms, along with rural businesses, supporteda total rural populationof 5.3 million people.

2.14 Within the white commercialsector, a considerableconcentration of landholdingsexists, which has been increasingover time. In 1988,a total of 22,190owners with holdingsexceeding 1,000 hectares and representing33% of the commercialfarming population held 84% of the land area in the commercial seCor, earned 50% of the gross income and 64% of the farm profit, made 44% of the capital expenditures,and owed 46% of the total debt. Such concentrationis partly due to the Sub-divisionof Aericultural Land Act of 1970p8and the AgriculturalCredit Act of 1966, whichhave provisionsfor the consolidationof what are defined as non-viablesmall farming units iato viable units. The definitionof a viable unit is a farm size that would permit the owner to obtain a full-time,farm incomelevel that is comparableto income levels in the urban sectors. While such legislationhas built a floor under farm sizes and has been supportedby state programsto consolidateland into such viable holdings, little has been done to reduce the possessionof the largest operators. Notably, there is an absence of any formof rural land taxation in all but one rrovince.

2.15 Number of farms. The umber of farms in the mainly white commercialsector increasedfrom 81,432 in 1921 to 119,556 in 1952,before decliningto 59,960 by 1983. From 1983to 1988,the total number of farms increasedslightly to 67,010. Accordingto Hattingh(1986, translationby Viljoen), the shrinkage in number of farming units acceleratedbetween 1950 and the early 1980s;beginning in 1950 when the number of farmingunits declinedat an averageannual rate of 0.9%. The rate was 1.4% in the 1960s, 2.4% in the 1970s, and 2.9% in the first half of the 1980s. All provincesexperienced substantial consolidationof farmland into fewer and fewer farming units: between 1955 and 1988, the number of farming units in the OrangeFree State fell by 53%, in the Cape province by 419%,in Natal by 41%, and in Transvaalby 39%.

2.16 Total farm ar. The amountof land held (ownedor leased) by the commercialsector has risen and fallen in line with the trend in the number of farming units, but at a more moderaterate-from 82.8 million hectares in 1930to 91.4 millionhectares in 1964, graduallydeclining thereafter to 85.8 million hectaresby 1988(Annex 2, Table 2.1). The increasefrom 1930to 1950 is attributedto the openingof

' Sub-divisionof a holdingis only pemitted with ministerialapproval.

Chaer 2: Struaure, ResourceEndownent, and Performanceof the AgriculturalSector Page 25 SotdhAfican djricuhMure: Structure. Performance& Ofijonsfor the Future

new, unoccupied"agricultural lands for farmingpurposes through State resettlementprograms (loubert and Groenewald, 1974). Causes for the contractionof the sector since the 1960s are less clear, although contributingfactors includeTrust land acquisitions,State acquisitionfor parks and reserves, and losses to industry, public works, and urban expansion.

2.17 Farm size. As the reductionin the numbersof farming units outpacedthe decline in the size of the commercial sector, the size of farming units has tended to increase over the past 30 years. Nationwide,farm sizes increasedfrom 738 hectares/farmin 1953 to 1,280hectares/farm in 1988 (Annex 2, Table 2.1). Over the same period, farm sizes increasedfrom 1,284 to 2,194 hectares/farmin Cape province, 471 to 998 hectares/farmin the OrangeFree State, 403 to 629 hectares/farmin Transvaal, and 390 to 609 hectares/farmin Natal.

2.18 With consolidationhas come a greater concentrationof land and wealth in the hands of fewer people. The data in Annex 2, Tables 2.2 and 2.3, showingthe number and area distributionof farms, are not directly comparablebetween periods prior to 1962 and after 1978, since the former farm-size classes were originally in morgen and the latter are in hectares. In 1950, farms greater than 2,000 morgen (1713 hectares)represented 9.7% of the total number, and accountedfor 55.8% of the total land area. By 1988, farms greater than 2,000 hectares in size (18.8%) held 69.2% of the total land area in the commercialsector. At the oppositeend of the spectrum, farmsof less than 100 morgen (86 hectares) in 1950 represented29.7% of the farms but occupied0.9% of total land area, while in 1988, farms of less than 100 hectaresrepresented 22.0% of the total numberand only 0.8% of the land area. The middle 60.6% of farms (100 to 2000 mnorgen,i.e., 86 to 1713 hectares, which constitutethe middle farming class) held 43.3% of the land in 1950. By 1988, this middlegroup (100 to 2000 hectares)held about the same share of the total number of farms (59.2%) but held less of the total land area (30.1%).

Ararian Structure in the Homelands

2.19 The homelandstoday are over-crowded,poverty-stricken and lack infrastructurecomparable to the white faumingsector, despite some attemptsat development. Their economicsignificance as a labor reservoir and dumping ground for surplus labor has intensifiedover time. In 1951 less than 5 million people (30%) of South Africanblacks lived in the homelands;this numberpeaked at 14 million in 1985 (almost 57%-Halbach 1988) and levelledoff at 13.1 million in 1988. These changes resulted from the resettlementpolicies pursued by the government,in particularbetween 1960and 1983, when millionsof blacks were relocated to the homelands(Marcus, 1989). Some analysts argue that the effecthas been functional urbanization (Corbett, 1987), i.e., households dependent on migrant earnings living in dormitory settlementsor border towns. Well over half of householdincome in the homelandsis today derivedfrom repatriatedmigrant earnings and pensions. Agriculturalproduction is estimatedto meetonly 16% of the population'sfood requirements(Huntley et at., 1989:54)and constituteson averageless than 20% of household income,and at best 40 % (Bembridge,1986a:78; Cobbett, 1987; Bekker, Cross and Bromberger, 1992).

2.20 Only a minority of householdsproduce for commercialpurposes. Most depend on non-farm incomesfor their livelihoods. Coetzee(1988) found that nearly 69% of farmers in KaNgwanewere net buyers of maize. Over 95% of the farmers in the Ocumusaward of KwaZuluwere net maizebuyers (in Vink and Kassier 1990). In KaNgwane,20% of the farmers produced nearly 89% of the maize crop.

Capter 2 Structure, ResourceEndowment, and Performanceof theAgricultural Sector Page 26 Sowh AfricanAgricuhture: Structure. Performance& Optionsfor the future

In Ocumisaward of KwaZulu, 20% of the farms produced nearly 98% of maize.

2.21 In 1985,8.4 millionblacks in the homelandsor around 60% of all blacks in the homelandslived below the poverty line, as defined by the MinimumLiving Level of the Bureau of Market Researchof the University of South Africa (Bekker, Cross and Bromberger, 1993). Slightlyhigher incidencesof black rural poverty-in the homelands,but also in the rural areas of the Republicof South Africa-are reported by the (unpublished)Family Surveys of the Central StatisticalService. Simkins(1991) puts this proportion even higher-at 84%.

2.22 The agrarianstructure of thehomelandsseems to be highlyinegalitarian and inflexible,19 although no reliablefigures exist. A substantialproportion of householdsare landless,and cattle ownershipseems to be very unequallydistributed; but only case study materialis availableon resourcedistribution in the homelands. For example, Stewart and Lyne (1988) provide informationon the distribution of land holdings for two sites in GcumisaWard of KwaZulu; for 132 farmers within the Natal Mistbelt, which is well suited to agriculturalproduction; and for a strata of 61 farmers in the hot, dry bushveld where expectedcrop yields are low. In both areas more than 80% of the householdsin their survey had arable land allocationsof less than 1.5 arable hectares, and less than 20% of the householdshad more than 1.5 hectares. Unfortunately,no data are presented on the amountof land area in the > 1.5 hectares category as a percentageof the total land area, to permit an assessmentof the extent of land concentration. In Cobbett's sample of KwaZulufarms, around 90% of the farms had 5 hectares or less of cultivatedland in two sites near white farming areas, althoughagain it is difficultto say what proportion of total area was representedby farms in excessof 8 hectares (Annex2, Table 2.4).

2.23 Rural householdsin the homelandsfall into four basic categoriesin terms of resource access and commercialorientation (Nicholsonand Bembridge, 1991). These are:

a. Resource-poorhouseholds, comprising about 562,000families (31 %) who have no arable land or grazing rights;

b. Smallholders,comprising about 1.03 millionhouseholds (56%), who operateat andbelow subsistencelevels, and who usually do not sell produce;

c. Progressivesmall-scale farmers, comprisingabout 238,100 (13%) households, who adopt some technologyand who sell produce and/or livestocksome of the time; and

d. Market oriented commercialfarmers, comprisingabout 3,100 households(0.2%), who make a living from farming.

Thc so-calledcommunal tenure, which, in the homelandsof South Africa,is in realitya state-designedsysten whichhas lostthe flexibilityof tmly communalsystems found throughout most of sub-SaharanAfrica. In the homelands,individual residential and ambleareas are allocatedby tribal, i.e., governmmeofficials, while intra-communal exchange of plots is severely constrained. In someareas, however,some form of traditionaltenure persists;and individualholdings are passeddown through male inheritance, but pasturageis communal.

Chapter2: Structure,Resource Endowment,and Performanceof the AgriculturalSector Page 27 SouthAfrkcan Ja*_Icuure,, Stru_ctu_re. Perormance A Otions for the Futur

ResourceEndowment

2.24 Currendy, forests, national parks and reserves sometimesoccupy arable land of the highest potential. In the Eastern Transvaal,for example,forests occupyabout 20% of high-potentialarable land. Moreover,In several regions, cultivationcould be significantlyexpanded on currentlyunused arableland or arable land that is used for grazing (e.g., in Natal and the Eastern Transvaal). At the same dme, however, in other regionsprice supportsand subsidieshave expandedenvironmentally and economically unsustainableproduction patterns onto lands more suitablefor a mixed crop-livestockfarming system.

Topography.Rainfall, and Clima_

2.25 South Africa lies within the southern temperatezone between22° and 35°S and enjoys a wide variety of agro-climates,potentially enabling it to grow a wide range of crops. Althoughlocal exposure to the variability of rainfall is high, the variety of agro-climatesand the extent to which correlation between agro-c}imatesis also variable insures the country as a whole rather well against major national crop failures caused by droughts or floods. To be sure, even South Africa's relatively wide variety of agro-climatescannot insure against severe regional droughts--suchas the one that affected the entire Southern Africa region in 1992.

2.26 About 31% of South Africa's surface area is level to gently undulating, while 19% is steep, rocky, and mountainous. The high Drakensbergescarpment rises steeply from the east and south-east, slopes graduallyto the west coast, and encirclesa vast interior plateau. Exceptfor a narrow coastalbelt at a mean altitude of less than 300 m, a narrow table land in the south at an elevation of some 450 m, and the interior of the Cape province ('Great Karoo") with an altitudeof 600 to 900 m, the major part of the country has an elevationranging from 1,200to 1,800 m.

2.27 lTree importantrainftl areas influencecropping patterns in South Africa:

a. a summer (Novemberto February) rainfall area which covers the interior (86% of the country);

b. a winter rainfall area with a mediterraneanclimate in the south-westernpart of the country (11% of the country); and

c. an all-year-roundrainfall area along the south coast (3% of the country).

2.28 Rainfalldecreases from east to west, from about 1,000mm east of the Drakensbergescarpment to less than 100 mm on the arid west coast. Averagemean rainfall is 511 mm per annum. Over 61% of the area receives less than this mean, and 21% receivesless than 200 mm-the lower limit for rainfed crop cultivation. Only 10% of the area receivesin excessof 750 mm. Precipitationvaries widely from year to year. Droughts are experiencedregularly, and floodsare not uncommon. Poor distributionof rainfall among years and regions makes crop productionrisky in the less-humidregions.

Chaper 2: Structure,Resource Endowmenr, and Performanceof the Agrkultural Sector Page 28 South AfricanAncuhare: StrUctUr. Perfoc A Ron-sfor the Future Nature Conservation

2.29 Nature conservationareas, both public and private, roughly equal the area of the homelands. Total area under conservationis estimated at around 16 million hectares, or 13% of the total land area of South Africa. About 11 million hectaresof these areas are privately owned and used for ecotourism and game farming. South Africa's 229 nationallyprotected areas (state forests, parks, and reserves) occupiedaround 5 million hectaresor 4% of the total land area in 1991. The relative size of the nature conservationarea under public ownership is not uncommonwhen comparedto other countries in the world. However, in terms of the absolute number of protected areas, South Africa's number is much higher than for other middleincome economies. Of all the low-incomecountries, only China and India- with a considerablelarger landmass-havea higher absolutenumber (World Bank, World Development Report, 1992, Table 33:282-83).

2.30 SouthAfrica is well-knowninternationally for its high levels of biodiversityand the high quality of the managementof the numerousand large conservationareas that currently exist. However, these naturalresources are managedfor the benefitof the white minority,with localblack communitieslargely excluded from either managementor use. O.' even greater concern is the emerging tendency to incorporatehigh-potential farm land into nature conservationareas (publicand private), possiblyin order to "protect"them from any future land reform programs.

Arable Land in the White CommercialSector

2.31 Lack of completeand reliable data on land use make it difficultto assess the availabilityof idle land. However,the conventionalwisdom is that arableland is to varying degreesgenerally fully utilized, albeit with varying degrees of efficiency. Nonetheless,there is considerablescope exists for horizontal expansionofagriculture on high-qualityarable land. Schoemanand Scotney(1987) assess the agricultural potential2sof the white areas and compare1986 cropping areas with the surfacecomputed for potentially suitable lands. They estimatedthat the land suitable for crop production (in the predominantlywhite areas) representsonly 13.5% of the surface area of South Africa. While 14.3 million hectares of land are theoretically suitable for crop production, only 12.9 million were cultivatedin 1986. There also appears to be ample room for such expansionin the Transvaal (3.1 millionhectares available versus 2.2 million cultivated). Such uncultivatedarable land, of course, is often either under forest or pasture; and expansionwould necessarilyentail substitutionof crop cultivationfor these old uses.

2.32 At the same time, however,there is a pressing need to reduce cultivationin other regions of the country, viz. the Winter Rainfall region (1.8 million hectares cultivated versus 1.5 million hectares available)and the Orange Free State (1.7 million cultivatedversus 1.0 million hectaresavailable). As

° Agicultural potental is a difricultconcept to measurebecause of wide differencesin viewsover what is sustinable in the long-trn, and of what congittes efficientresource n _agmt. The work of Schoemanand Scotney(1987) represns one of the first attemptsto measureagricultural potental for the whiteareas of the RSAby inrating climate,soils and topogaphy(slope) factors. Using data on crop performance,soils and rainfall,they qualitatively construct a map (1:2.5 millionscale) of the RSA, delimitingfour domainsof agricultumlpotential: high, medium,low, and unsuitable. High-potentialland is regardedas all areas with favorableclimuae, soils and terrain, whereactual or potentialperformance of cropsranks high comparedwith cropperformance nationwide. Landswere consideredunsuitable for drylandproduction if limitatonswere perceivedin one or more of the factors. The resultsof their studyare summarizedin Table2.2.

Chapter 2: Struct4re,Resource Endowment, and Performanceof the AgriculturalSector Page 29 South A,Itan AerluIe: Strutwur.Per orance & ODionsfor the Fu re a result of past incentivestructures, the use of marginalland for crop production is widespread. This is the case of the expansion of monocropped, mechanizedmaize and wheat cultivation in the Western Transvaal and the Orange Free State.

2.33 The suitabilityof land for dryland crop cultivationis highest in the Highveld(51.3%) and Natal (35.1%), and lowest in the Orange Free State (5.0%) and the Karoo (0.1%). Of the total land area estimatedto have some potentialfor cultivation(I4,268,700 hectares),21.9% is of high potential, 49.1% is of mediumpotential, and 29.0% is of low potential. Of the total surface area in the white seOt (105,373,400 hectarts), high quality land representsonly 3.0%, and mediumquality land 6.6%. Most of the high-potentialland is situatedin the Eastern Transvaaland Natal, and medium-potentialland in the Highveld. Two-thirdsof the low potential arable land is in the NorthernCape, Orange Free State, and West Transvaal. The Transvaaland the Orange Free State containthe highestfraction of arable land in the white areas (23.5% and 29.6% respectively). Only 5.6% of land is considered arable in Cape Province. Comparablefigures for the homelandswere not computedby Schoemanand Scotney.

Table 2.2: Cro, ProductionPotential and Present Cultivationof White-OwnedFrmland ('000 hectares) Total Area High Medium Low Land with 1986 Crop Cropping Potential Pattrns Transvaal 15,249.0 1,198.6 890.0 1,042.8 3,131.4 2,202.0 Natal 5.832.1 1,539.4 440.0 70.0 2,049.4 933.5 Higiveld 11,85.9 190.0 4,440.5 1,316.4 5,946.9 5,678.2 Fee State 20,482.3 68.0 946.0 1,014.0 1,683.6 Witer Rainfal 13.800.0 912.0 582.8 1,494.8 1,800.0 Eatern Cape 5,400.0 194.4 241.2 176.6 612.2 426.3 KAroo 29,060.0 10.0 10.0 20.0 192.0 Oher band 3,964.1 Total 101,409.3 3,122.4 7,001.7 4,144.6 14,268.7 12,915.6

Notes: Plantedpastures ae exluded from the definitionof cropped land. Other land is mainly constiugedby NationalParks.

Souse. Schomanaand Scotney, 1987.

2.34 Similar conclusionsare arrived at by other studies (McKenzie,Weiner and Vink, 1989). Using the developmentregions classificationas proposedby the DevelopmentBank of SouthernAfrica, Regions E (Natal, KwaZulu, and Northern Transkei) and F (Eastern Transvaal, KaNgwane and part of Simdlangentshadistrict of KwaZulu) account for 90% of all high potential and 22% of all medium potentialland, whileonly 2% of all crop land is of low potential. Croppingintensities are, however,only 62 and 55% In RegionsE and F respectively. Substantiallateral croppingexpansion could take place on high and mediumpotential land.

2.35 There also seems to be ample scope for substitutingmore extensivefor more intensiveland use. The size of a possiblelateral expansionin high potentialland-theoretically about 2.5 millionhectares-is

Chapter2: Sructure, Resource Endowment,and Perfornanceof the AgriculturalSector Page 30 South Afican A&rkUltare:Structure. Performance & Ontlonsfor theFature reported to be limited, however, because large areas of high-potentialland fall within nature reserves which make up some 24% (RegionE) and 31% (region F) of the total area.

2.36 Moreover,Schoeman and Scotney(1987) report findingsof a studyconducted by the Department of Agricultureand Water Supplyfor RegionF (1985) which indicatethat 20% of high- potential land is presently under forests. Priority areas designatedfor future afforestationin the districtsof Ermelo, Piet Retief, and Wakkerstroomalso largely consist of high-potentialagricultural land.

Arable Land Use in the Homelands

2.37 Given the legacy of past racially-basedpolicies, the resource endowmentof the homelandsis recognizedto be poor by design. Reliable arable land use data for the homelandsare scarce and often contradictory. The total arable land availablein the homelandsper capita is estimatedto vary from 0.08 hectares in Qwaqwato 0.27 in Boputhatswana(Table 2.3). Nonetheless,it is sometimesargued that a large proportion of arable land in the homelandsis currently underutilized.An estimated20 to 30% of the area allocatedas arable land is left uncultivatedevery year. Van Rooyen and van Zyl (1990) note, however, that very little arable land is unutilizedin the subsistencefarming areas. Whateverarable land in the subsistenceareas is not being utilized for croppingpurposes, they claim, is heavily stocked with cattle-at rates between 20 to 55% higher than in commercialareas.

2.38 Moreover, although these areas generally receive abundant rainfall, steep terrain reduces the amnountof arable land contained within them. In some cases land allocatedfor cropping purposes in subsistenceareas is in fact not arable, or is of low quality. Bophuthatswana,Lebowa, KwaZulu, and Transkei appear to occupy most of the presumablyarable land, but most of what is in the first two is of low productionpotential. It was previouslythought that Transkei containeda significantshare of arable land; but recent research indicatesthat only 50% of the arable land is of moderateto high potential, the balancebeing very marginalwith a high erosion propensityunder crops (vanRooyen and van Zyl, 1990).

2.39 Land degradation, especiallyon non-arableareas under crops, has also reducedthe quantityand quality of land over large areas. A studyof the agriculturalpotential of Ciskeiconcluded that cultivated areas had been demarcatedor retained in regions that were unsuited for crop production because of inadequaterainfl, inferior soils, or excessiveslopes. An assessmentof actual land allocationpatterns suggeststhat only KaNgwaneand Bophuthatswanastill have arableland availablefor cost-effectivecrop production (van Rooyenand van Zyl, 1990).

2.40 The combinedeffect of extreme land pressure and restrictions on rental and sale under the traditional tenure arrangementshave caused the renting out of land to others to be a risky proposition. At the same time, the homelandsconstituted until recentlythe only places where Africanscould legally own rural real estate.

2.41 Finally, the underutilizationof some homeland agricultural land merely suggests that the profitabilityof agriculturein the homelandsis relativelylow and the opportunitycost of householdlabor high. It can be argued that this is exactly what the policy-inducedincentive structure was supposedto bring about and that it has made the homelandsexactly what they were meant to be-cheap labor pools for other sectorsof the South African economy.

Chapter2: Structure, ResourceEndowment, and Performanceof the Agricutural Sector Page 31 Sough frican Aericulture:Structure. Performance& Optionsfor the Future

Table 2.3: Aable Land and Ponulation Density by Reeion

Per 1988 Rainfed Irrigated Total 1988 1988 Rural Resident Total Arable Arabc Arable Urban Rural Arabl Surface Land Law' Land Population Population Land Area (000) (000) (ha/res) (km2) (percent) (percent) (percent) (Per) (per) (per capita) A B C D B F A*D/F Cape 642,467 4.52 1.10 5.62 4,815 1,256 2.87 Natal 55,407 15.85 3.59 19.44 1,771 785 1.37 Transvaai 223,490 20.35 3.14 23.49 7,757 2,210 0.54 O.F.S. 127,292 28.80 0.79 29.59 1,249 1,093 2.38 Kwazulu 36,074 -- - 10.00 1,104 3,630 .10 Oazankulu 7,484 1.70 6.40 8.10 29 640 .09 Kwandebele 2,208 -- - 15.41 36 327 .10 KaNgwane 3,917 17.10 12.30 29.40 67 462 .25 Lebowa 22,137 15.10 0.60 15.60 154 2,310 .15 Qwaqwa 1,040 - - 18.00 34 228 .08 Bophutswana 40,011 -- -- 10.81 294 1,580 .27 Ciskei 8,100 - - 8.30 278 512 .13 Transkdi 43,654 - - 4.67 168 2,877 .07 Venda 6,807 10.35 0.62 10.97 18 490 .15 Total 1,220,088 - - 12.64 17,774 18,400 .43 a. DBSA, 1990, A Regional Profile of The Southern African Population, pp 25-68. b. Data on total surface area and arable land is taken from DBSA, 1991, An Inter-Regional Profile, p. 34.

Grazing Land

2.42 Livestockproduction systems in the commercialareas are basedon large-scale,extensive ranching models. Consequendy,livestock production takes place on specializedranches, where cattle is grazed on fenced-in areas, Such rangeland production makes up 81% and 87% of the farm land in the commercialareas and the homelandsrespectively. 'Carrying capacity"in this model dependsto a large extenton local rainfall and varies from2 to 5 hectaresper LivestockUnit (LSU) in the high rainfall areas to more than 30 hectares per LSU in the dry areas of the Cape province.

2.43 The concepts of carrying capacity and overstocking,however, imply a degree of accuracyand objectivitythat has not alwaysbeen matchedby empiricalevidence. Consequently,such conceptshave come under critical scrutiny in recent years. In South Africa, some estimatesof carrying capacitypoint to overstockingin certain regions, e.g., in the Karoo, under the current systems.The total area covered by natural "veld" (grasslands)amounts to about 81 millionhectares (includingall parks and conservation areas). Of this area, 0.3%6is in very good condition, 13.2% in good condition,76.3% in fair condition, 10% in poor conditionand 0.2% in a very poor condition. Currentlythe naturalveld in SouthAfrica can

Chapter2: Structure, ResourceEndowment, and Performanceof the AgriculturalSector Page 32 Sosth Afc Aericukure:Structure. Performance A ODvIonsfor the Ftoaure

Ocarry' about 8.1 million LSU. If veld conditionswere to be improvedthrough betterveld management, approximately8.8 million LSU couldbe carried in SouthAfrica. However, about 13.3 million LSU are now kept in South Af&ica--onethird of them in the overcrowdedhomelands. According to these estimates, South Africa's grazing resourcesare thereforeover-utilized by about 55%.

2.44 Stocking rates in the developingareas often exceed current definitionsof carrying capacity by more than 100%, leadingto productivitylosses and environmentaldamage. Althoughstocking rates in the homelandare high, they have been relativelystable over long periods of time; it has thereforebeen suggestedthat some sort of equilibriumseems to have been established.

Wateir Resources

2.45 Annuallyrenewable water resourcesfor SouthAfrica are not high. Over the period 1970-87they anounted to 51 kla3, or 2244 m' per capita per year. Table 2.4, however, lists several countries which are in a similar range of average per capita water availability,but are generally not regarded as poor agriculturalperformers. However, such comparisonsshould be interpretedwith care, since they do not take rainfall variability into account.

Table 2.4: Water Resources

Annual As a Total Annual Population Per capita Withdrawalof Percentageof Water 1970-87 total anaual Internal Total Water Resources Average water Renewable Resources (1970-87) resources Water Resources (1970-87) (1970-87) (km3) (percent) (km3) (millions) (m3) South Africa 9.2 18 51.11 22.77 2244 Denmark 1.4 11 12.73 5.054 2518 France 40.0 22 181.82 54.95 3309 3ermany 41.2 26 158.46 61.68 2569 Poland 16.8 30 56.00 35.59 1573 Spain 45.3 41 110.49 38.59 2863 China 460.0 16 2875.00 995.67 2888 India 380.0 18 2111.11 620.92 3400 Korea 10.7 17 62.94 35.91 1753 Peru 6.1 IS 40.67 20.74 1960

Inoc: World Bank. World DevelopmentReport 1992. Table 33, page 282-283.

2.46 Surice runoff constitutes85% of the total supply of water, and 15% comes from underground supplies. It is estimatedthat about 60% of availablewater can be economicallyexploited at the present

Chapter2: Structure,Resource Endo&*enr, and Perfonmanceof the Agricuttral Sector Page 33 SouthAfrikan Atricubure: Structure. Performance & ODrionsfor the Future time. However, actual withdrawalof internal, renewablewater resources2' over the period 1970487was 9.2 Iam'per year, or only 18% of the total of 51 km3. Total per capitause was 404 ni3 of which339 m3 was for industrial and agricultural use, and 65 m3 for domestic purposes (World Bank, World DevelopmentReport 1992, Table 33, p.283). While the above figures refer to national averages, they do not reflect the very high variability in regionaldistribution of water resources. For example, Natal occupies less than 7% of South Africa's surface area, but has 40% of its availablewater.

2.47 Agricultureis the largest user of water, consumingas muchas 50% of total demand. In 1981-- the latest year for whichdata were availableat the district level-824,000 hectareswere irrigated, or 5.9% of the 14.3 millionhectares with crop potential'. In 1988, 230,000 irrigatedhectares were in government water schemes, 300,000 hectareswere controlledby irrigation boards, and 70,000 hectares were in the homelands. Accordingto the Departmentof Agriculture,the total area under irrigation in South Africa today is 1.2 million ha.

2.48 In the past, commercialfarmers have benefittedfrom substantial amounts of public assistancewith respect to investmentsin water control. Moreover,property rights to water have exactly mirrored land distribution, creatingserious negative externalitiesand inequalitiesbetween agricultureand other water uses (e.g. drinking water for local communities). Current water use patterns, such as the irrigation of pasture land, reflect the low user cost of water to commercialfarmers. Since irrigation infrastructure, even in the homelands, is typically designed for large-scaleusers, important efficiency-gainsmay be realized by adaptingexisting infrastructurefor smaller-scaleuse.

2.49 Past investmentsin irrigation infrastructurerepresent an importantasset for the future, provided the allocationof water is made more equitableand efficientthrough the redefinitionof public and private property rights to water and the movementtowards a waterpricing systemthat reflectsthe scarcityvalue of water in its various uses (e.g., as drinking water) to all users (e.g., poor consumersin both urban and rural areas).

FinancialResources

2.50 Whitefarmers have accessto financecapital through a variety of institutions-themost important of which are the Land and AgriculturalBank, the agriculturalcooperatives (which fulfil an agency role for the Land and AgriculturalBank), and the commercialbanks. Total farm debt of white farmers amountedto 13.6 billionRand in 1988, up from 1.4 billionRand in 1970(Christodoulou and Vink, 1990, Table 1, p.3). The Land and AgriculturalBank and the agriculturalcooperatives together hold nearlybalf of this debt. Whitefarmers have been able to borrow at subsidizedinterest rates from these institutions, and their interest rates have been negative in real terms for virtually the entire 1970-1990period (see Chapter 1). Until recently,black farmers in the homelandsdid not have access to these financialservices. Substantialamounts of public financeis regularlyused to recapitalizefinancially-stressed farmers. For instance,in the 1992budget R2.4 billion was allocatedfor loan-writeoffs on guaranteedloans. Another

2' Annualwithdrawal refers to the averageannual flows of riversand undergroundwaters that are derived fromprccipitation fallingwithin the country.

22 lis compars to 33% in Asia..

Chapter2: Structure,Resource Endowment, and Performanceof the AgriculuralSector Page34 SouthA*can Aericufture:Structure. Performance& Ontionsfor the Future

Rl billion was set aside for drought-relief,which includedsome debt-relief.

2.51 White farmers also receive a wide range of subsidies. For example, in 1988/89,4,124 subsidy applicationsfrom white f^tners were successfulin securinga total of 682 million Rand, includingR121 million for input purchases,R104 million for droughtrelief, R88million for flood disastersubsidies, R75 millionfor stockfeedpurchases and incentives,and R52 millionon debt consolidation(Vink and Kassier, 1990, Table 3). Farmers in the homelands, equally affected by floods and droughts, get little such assistance. Subsidiesused in the administeredpricing schemesfor agriculturalcommodities constitute transfers to both producers and consumers,but it is difficult to determine the producer share in these subsidies. In 1988 a total of R583 million was spent to subsidizea variety of agriculturaloutputs and inputs (Abstractof Agricultural Statistics, 1993). By 1992, however, this total had declined to R165 million.

Performance

2.52 Agricultureoccupies an importantposition in the SouthAfrican economyin its own right and as a supplier of inputsto the agro-processingsector. But its share of GDP has declinedfrom around 20% in the 1920sto 4.7% in 1991 (Table 1.3); and its growth in the past 40 yearshas been uneven, averaging 2.5% in real terms since 1950, with rapid growth in the 1970sbracketed by slower growth in the 1960s and 1980s (Table 2.5).

Table 2.5: RealGrowth in Aariculture- 1950 to 1990

Period Aveage Annual GrowthRate 1950- 1960 3.1% 1960 - 1970 1.6% 1970- 1980 5.8% 1980 - 1990 1.4%

Sou: Derivedfrom Naional Accounts

InternationalTrade

2.53 Christodoulou and Vink (1990) report the following self-sufficiencyindices for the various commoditygroups for the period 1980-89in South Africa:

Field Crops: 134 HorticulturalCrops: 152 Animal Products: 97 Total: 130

2.54 Some 25% of the total production of field crops during the period 1980-89 was available for export-roughly 30% of the total value of field crop production;and, as noted above, more than one-third

Chapter2: Structure,Resource Endowment,and Performanceof the AgriculturalSector Page 35 Sowh Afican Arricubure: Striaciure.Performance A ODtlons for the Future of the annual horticulturalproduction is also availablefor export.

2.55 South Africa has participatedsignificantly in internationaltrade in agricultural commoditiesin recent decades (Annex 2, Tables 2.5 and 2.6). There is no clear trend in import volume, but they are rather limited in both volume and value, reflectingthe high degree of self-sufficiencyattained in most major food commodities;and annual totals are much affected by the cereals balance. Intermittent importationof cereals is required during periods of drought, and wheat is often needed to supplement domesticproduction. Coffeeand tea are importedto supplementdomestic production, and in most years the country imports red meat. Exportsgrew in volume duringthe 1970sbut declinedduring the 1980s, largely reflecting reduced availabilityof exportablecereal surpluses, mostly due to the drought.

2.56 In the internationalfruit market SouthAfrica, as a southernhemisphere producer, is well situated to meet the winter season demand in Europe. The major areas of export growth have been deciduous fruits, citrus, and sub-tropicalfruits; these continuedto grow during the 1980s, in spite of sanctions. Exports are growing rapidly in the 1990s. With a continuingfavorable exchange rate and removal of economicsanctions on SouthAfrica, rapid growth of fruit exports is anticipatedand orchards expansion has been proceedingfor several years. However, substantal concentrationin fruit marketingcould stifle future growth of the sector.

2.57 Grain exports, however,are likelyto be substantiallyreduced compared to the past two decades, as the area allocatedto cereals is reduced to reflect decliningprofitability and pricing policy reforms.

AgriculturalProduction in the White, CommercialSector23

2.58 Some 95% of the gross value of agriculturalproduction in South Africa, as reported in official statistics, originates in the predominantlywhite commercialsubsector. Over the past 30 years the compositionof agricultureby gross value has moved in favor of horticulturalproducts and, to a lesser extent, livestock. Using the threefoldclassification of field crops, horticulture, and livestock, the broad shifts are illustrated in Table 2.6 below. Horticulturalcommodities have generally been more buoyant than the field crops, with a rise in the total contributionto sectoralvalue of production from 14.4% in 1980/81to 20.9% in 1990/91. Stronggrowth in the horticulturalsector has been shown in particularby commoditiesdestined for export, notablyfruits, but wine and vegetableshave all risen sharply. The share of field crops peaked at 48.5% in 1980/81,but fell to 34.2% in the followingdecade.

I Althoughthe qualityof agricul statisties in SouthAfrica comparesfavorably to manyothes countriesin the word, a numberof deficienciesremuin. Fist, informationon area cultivatedin the commerci soctor under-reportstotal area culivated, as officialarea data are unavailablefor certaincrops and land uses(such as horticulture.vegetables, forages, and fallow). Second, areas cultivatedfor key growthsectors-horticultural crops and vegetables-arenot reportedin official statisties. Third,data on agricuural activitiesin municipalarcas on smallholdingsare not reportedin annualstaistes of the Depatm of Agriculte, but do figurein censusreports by the CentralStatistical Service. Theseshortcomings should be keptin mindthroughout the rport

Chapter2: Sructure. ResourceEndownmnt, and Perfornance of the AgricukuralSector Page 36 South AfricanAjrkcuhure: Structure. Perfornance & Ootionsfor the Fatlre

Table 2.6: Comoositionof AsriculturalGross Value

1960161 Share 1970/71 Share 1980/81 Share 1990/91 Share (%) (%) (%) (%) Pield Crops 365 42.6 700 46.7 3402 48.5 7422 34.2 Hoficulture 126 14.7 259 17.3 1010 14.4 4529 20.9 Anima Producu 365 42.6 541 36.1 2599 37.1 9767 45.0 Total 856 100.0 1500 100.0 7011 100.0 21718 100.0

Source: Abstact of Agicultural Statistics(1992)

2.59 Since 1965 total cultivatedarea (basedonly on principalfield crops)has fluctuatedbetween 8.5% and 9.7% of the total land area comprisingthe commercialsector (Annex 2, Table 2.7)-short of the 12.64% of agriculturalland consideredarable. The differencecan be attributedto non-fieldcrops (e.g., fruit trees, vines, vegetables,planted pasture, forests, and forages)24,land set aside for fallow rotations and conservation,and idle land. Over the past 10 years, there has been a marked decline in the area devotedto maize (Annex2, Table 2.7). Conversely,there has been a marked increasein the area devoted to oats and sunflowers.

2.60 Field Crp. Of the field crops, only wheathad growth in yields that approximatedthe 1970to 1989population growth rate (2.9%); maizeyields grew at an averageannual rate of only 0.7%, reflecting slow technicalchange. The strongest long-termannual yield growth has taken place in the productionof barley (8.3%), soya beans (6.6%), cowpeas (3.7%), wheat (2.9%), tobacco (2.6%), and dry beans (2.3%). Annex 2, Table 2.8 shows the trends in growth rates for the period 1970-1989.

2.61 baaze. The technicalefficiency of maizeproduction-in terms of higher and more stablephysical yields-seems not to have improved; and the area under maize declined as a result of policy measures which have decreasedthe profitabilityof maize production and stimulateda movementaway from crop productionin marginalareas. The economiccosts of failingto improvethe technicalefficiency of maize productionare considerable. Maizeoccupies more land than any other crop in SouthAfrica, representing 44% of the area of principalfield crops. Moreover,maize constituted some 75% of total grain production and 64% of the value of field crops; 56% of humangrain consumptionwas of maize products, and maize was by far the most important animal feedstock(Groenewald, 1989).

2.62 Annex 2, Table 2.9 summarizesthe area planted, production, and yield in white farming areas for the period 1981/82to 1990/91. Boththe area plantedand the yield have declinedover the past decade (by 2.7 and 1.6% per annum respectively)as a result of the decliningprofitability of maize production and because of drought, creating a decline in production of 4.3% per annum. The Land Conversion Scheme, which subsidized the conversion of marginal crop land to perennial pastures, considerably strengthenedthis trend.

1 lhe Abstractof AgriculturalStatistics provides area data on principlerield crops, but not for fruittree, vines,vegetbles, panted pasture, foremsand forages.

Chapter 2: Structure, ResourceEndowment, and Performanceof the AgricultaralSector Page 37 SouthAfrican Agriculture: Structure. Performance& Optionsfor the Future

2.63 Giventhe instabilityof physical yields, export volumeshave also been highly erratic (Annex 2, Table 2.10). Note, however, that given the high domesticprotection and subsidies, exports of maize occur as a means of surplus removaland constitutedfinancial losses to the industry.

2.64 Wha. Wheat, on the other hand, is another fieldcropthat has shown improvementin technical efficiency. Wheatyields have increasedat an averageannual rate of 2.9% since 1970and reacheda peak of 1.8 tons/hectaresin 1987/88. Yieldsin 1989/90were 1.1 tons/hectares. In 1991192,2.2 million tons were produced on 1.4 million hectares, at an average yield of 1.6 tons/hectares. The area planted in wheathas remainedsteady in South Africasince the early 1970s, at about 2 million hectares,or 22% of the total area of principal field crops. Total productiongrew at an average annual rate of 2.6% during the 1970s and 1980s, reaching a peak of 3.5 million tons in 1988/89, mainly due to growth in yields. Much of these increases in yields/hectaresmay have been caused by the increasedcultivatici of wheat under irrigation.

2.65 Qjt. Yieldsdeclined at an annual rate averaging5.9% during the 1970sand 12.9% during the 1980s,from about 260 kg/ha in the early 1970sto 61 kg/ha by 1989/90. After remainingrelatively stable at around400,000 hectaresthroughout the 1970s, the area plantedto oats rose at a rate of 5.5% per year during the 1980s,reaching 690,000 hectaresin 1989190(about 10% of the total area devoted to principal field crops). Yields declined at a faster rate than the expansionof area, suggestingthat oats are being targeted to marginal lands (relativeto wheat or maize). Productionfell at an annual rate of 5.9% over the period, from a high of 121,000 tons in 1970/71to an averageof about 50,000 tons per year in the late 1980s. Approximately42,000 tons were produced in 1989/90. In 1991/92, 31,000 tons were producedon 811,000 hectares, an averageyield of 38 kg/ha.

2.66 Sunflower. Yields have fluctuatedfrom year to year around a constant trend level of about 0.9 tons/hectares,but no significanttrend can be distinguished.The area planted to sunflowerhas grown significantlyin recent years, to about 8% of the total area under principal field crops. It increasedfrom 185,000 hectaresin 1970/71to 520,000 hectares in 1989/90,representing an average increaseof 8.7% per year in the 1970s, slowingto 4.4% per year in the 1980s. Productionof sunflowerseeds increased from 138,000tons in 1970n17to 559,000 tons per year in 1989/90,an averageof 11.3% per year. Most of this production increase came from area expansion. Yields in 1989/90 averaged 1.1 tons/hectare. In 1990/91,587,000 tons were producedon 575,000 hectares, an averageyield of 1 ton/hectare.

2.67 Sugarcane. Yieldshave remainedrelatively constant during the past two decades,averaging 48.6 tons/hectaresover the period, and 49.5 tons/hectaresin 1989/90.The area devotedto sugarcaneincreased by about 20% over the 1970s, reachinga level of 400,000 hectaresby the early 1980s. About 375,000 hectareswere plantedto sugarcanein 1989/90,representing more than 5% of total field crop area. Total productiongrew from i5.2 million tons per year in the early 1970sto 18.2 million tons per year by the mid-1980s-a growth rate averaging 1.3% per year--mainlybecause of expanding the area. A total of 18.6 milliontons wereproduced in 1989/90;and in 1991/92,19.7 milliontons were producedon 380,000 hectares, for an average yield of 51.8 tons/hectares.

2.68 HorticulturalProducts. The volumeof horticulturalproduction nearly doubledbetween 1970and 1989, rising at an average rate of 2.8% per year. Moreover, there has a been a major recovery of confidencein the industry since 1989, resulting from a combinationof the lifting of sanctions and the

Chapter2: Structre, ResourceEndowment, and Performanceof iheAgricultural Sector Page 38 &utahAfrican Aricuhure: Structure. Performance & Optionsfor the Future

strengthening of export earnings with the depreciation of the currency. New plantings are now intensively under way, promising a major increase in production in the coming years. South Africa is Africa's dominant supplier of fresh and processed horticultural products to the European Community, accounting for over half of such exports from Africa to the EC.

2.69 The volume of output of all horticultural products increased by 25% between 1985 and 1991, in contrast to a widely fluctuating growth in field crop output and a 15% increase in livestock production. Citrus production contributed most to this horticultural increase, rising by 60%. Other increases were noted in deciduous fruits (12%), subtropical fruits (16%) and vegetables (30%). Over this same period horticultural products increased from 16.2% of the gross value of farm output to 20.9%, and field crops declined from 43.3% to 34.2%.

2.70 The three main categories of fruit (deciduous, citrus and subtropical) also experienced significant increases in export volumes over the past few years. After declining from 468,000 tons in 1978, the volume of citrus exports increased from 300,000 tons in 1985 to 372,000 tons in 1990. Deciduous fruit exports have increased from 256,000 tons in 1978 to 336,000 tons in 1990 (296,000 tons in 1985). The value of all fruit exports increased from RI billion in 1986 to RI.6 billion in 1990, while fruit imports amounted to R83 million in 1990.

2.71 Animal Products. The total volume of output of animal products grew by more than 50% between 1970 and 1989, reflecting an average increase of 2.6% per year.

2.72 Catde. Moderate increases in catte production have been associated with low improvements in yield, given the relatively static herd, which has numbered between 7.0 million head and 8.8 million in South Africa, excluding the homelands, and between 10 and 12 million head in the whole country between 1960 and 1990.

2.73 Catte numbers grew at an average annual rate of 1.9% during the 1970s, from 7.9 million in 1970/71 to 9.3 million in 1977/78, and have decreased moderately since then, to 8.8 million in 1991/92. The number of cattle slaughtered increased at an average rate of 3.6% per year during the 1970s, from 1.7 million in 1970/7l to 2.7 million in 1979/80. Roughly 2.3 million catde were slaughtered in 1990/91. Total production of beef and veal rose from 449,000 tons in 1970/71 to 711,000 tons in 1979/80, representing an average annual increase of 4.3%. About 634,000 tons were produced in 1990/91.

2.74 Fg. Pig production has shown similar mediocre gains, averaging 892,000 the 1970 and then increasing at 1.5% per year during the 1980s, to 1.2 million by 1989/90. The number of pigs slaughtered averaged 1.4 million per year between 1970/71 and 1979/80 and then grew at an annual rate of 1.7% over the 1980s. Some 2.2 million pigs were slaughtered in 1990/91. Pork production followed a similar pattern, averaging 89,000 tons per year during the 1970s and then growing at 2.0% per year during the 1980s. Approximately 134,000 tons were produced in 1990/91.

2.75 Shee and Ga. Sheep and goat numbers have fallen slowly over the past two decades, from an average of 33.2 million animals during the 1970s to an average of 31.6 million animals in the following decade, representing a decline averaging 0.4% per year. Sheep and goats numbered 31.1 million (of which 28.6 million were sheep) in 1991/92. The number of sheep and goats slaughtered has

Chapter2: Structure,Resource Endowment, and Perfonnanceof theAgrictuktutral Sector Page39 South AfricanAlrrcuture: Structure. Performance& OD_ionS for the Future fluctuated around an averageof 7.1 million animalsper yea, with 9.1 million animals slaughteredin 1990/91. Total production of mutton has averaged 179,000 tons since 1970/71, with 188,000 tons produced in 1990/91.

2.76 Poulty. A strongincrease in technicalefficiency has increasedpoultry production in SouthAfrica rapidly over the past two decades. Some558,000 tons of poultry were slaughteredin 1990/91,nearly five times as much as in 1970/71. Most of the increasecame during the 1970s, when poultry slaughtered increasedat an average rate of 11.5% per year.

2.77 The concurrentslow growth in the technical efficiencyof cattle production has resulted in a decline in red meat consumptionover the past 20 years in favor of more efficientlyproduced white meat (Annex2, Table 2.11). Althoughthe overall levelof meat consumptionhas remainedat about40 kg/cap, this total concealsa fall in per capita red meat consumptionfrom 40 kg to 25 kg, while consumptionof white meat (poultry)has risen from under 2 kg to around 15 kg. As a result, the poultry industryhas witnesseda 10-foldincrease in the volumeof productionover the past three decades. These important shifts in consumptionand production relate to:

a. the reductionin relative real costs of white meat under intensivepoultry production;

b. the increasein the relative real price of red meat partly attributableto the monopolistic marketingstructure which increasedmarketing margins;

C. the increasingshare of the populationrepresented by the lower income(black) group; and

d. the increase in parallel, and therefore unrecordedmarkets.

A_riculturalProduction in the Homelands

2.78 The analysisof land use in the homelandsis hamperedby data problems. Reliable and complete informationon area and yield data for the homelandsand so-calledindependent states are completely lacking, which precludes any real analysisof the effectsof policy changes. Table 2.7 presents available production data for the homelands.

Chapter2: Structure,Resource Endowment, and Performanceof the AgriculturalSector Page 40 Soah gilcwtAnricubure: SrrucdUre Perfo,mance & Otilonsfor the Future

Table 2.7: Gross Value of Azriculturm1Production in tite Homelandsof South Afnica

Area Field Crop Production Horicultural production LivestockProduction (1985/86Dat) (R00) (R'000) (R'000) Total South Africa 2,583,600 2,100,900 4,690,800 KwaZulu 147,620 49,442 65,955 QWaq 464 479 5,288 Lebowa 27,714 1,230 18,706 Gazankula 2,833 3,358 5,044 KaNgwane 11,636 740 715 KwaNdebele 5,824 1,004 2,125 Transkei 45,040 61,862 98,436 Bophutwana 37,555 1,231 20,575 Venda 4,070 13,776 9,301 Ciskei 1,622 10,416 50,808

Sou,a Deveopment Bank of SouthemAfrica, internal data and 1987; Republicof South Africa, 1991; and van Rooyen, Fenmys,and van Zyl, 1987.

2.79 Only a smallproportion of agriculturalproduction in the homelandsreaches commercialmarkets (van Rooyen, Fenyes, and van Zyl, 1987). In the past, many of the marketingboards and other control bodiesrestricted access to markets throughthe use of quotasor permitsystems. In recent years, however, such restrictionswere progressivelyremoved; and there is considerableevidence of substantialgrowth in informalmarkets. Commercialproduction in the homelands accountsfor just over 3% of the total GDP of these areas, while non-marketor subsistenceproduction accounts for slightlymore than 7%. During the period 1970 to 1985, there was a marked real increase(190%) in the value of commercial aricultural production, from R56.2 million to R163 million, which was attributableto investments undertakenby the various parastatal developmentcorporations, as well as private sector investments (particularlyin KwaZulu). The net value of all agriculturalproduction was on averageabout R240 per householdper year in 1985 (Christodolou and Vink, 1990) (Table2.8).

hapter2: Sructure, ResourceEndowment, and Performanceof the AgrcuAturalSector Page 41 SowthAfto Aiirkukur: SAruaIre.PEgf2aM,Mz & ODeoasfor the hture

Table 2.8: AbsoluteVaue of Commercialand Subsidence AgriculturalProduction in the Homelandsand its Contnbutionas a PeraQtage of GDP. 1985

Teritoiy Commercil GVP Subsistenc GDP Total GDP Agricutual (R'000) (prcent) (RIOO) (percent) (R'000) Value Per Household1/ Ciskei 2,964 0.7 16,600 3.9 424,174 157 Tuan"kel 26,700 2.0 137,000 10.1 1,359,200 327 Kwazulu 75.000 7.1 133,000 12.5 1,061,768 285 Venda2/ 9,141 4.6 17,100 8.7 196,675 341 Lebowa 12,000 2.2 33.000 6.1 539,928 125 Gazankulu 3,S40 1.5 20,310 8.8 229,350 231 Bophuthatwana 22,400 1.9 30,000 2.6 1,162,650 183 Kangwano 8,S00 7.9 9,100 8.5 107,540 235 KwaNdebele 800 1.5 1,500 2.9 52,500 48 Qwaqwa 2,000 1.8 1,990 1.8 110,260 114 Total 163,045 3.1 399,600 7.6 5,244,045 240

Nots: 1/ Calculatedby divding totl agridulual producdonby the nunber of househols. 21 1984 figures. Soue: CobbeU1987.

2.80 As of 1985, agriculture in the homelandsemployed roughly the same number of people as the commercial sector-i.e., 1.1 million (van Rooyen, Fenyes, and van Zyl, 1987). In the homelands, however,very few farmerscan rely only on agriculturefor a livelihood.Quality of life is to a largeextent a function of accessto off-farmsources of income. Non-agriculturaleanings and, in particular, migrant earnings and remittancesand pensionsare a far more substantialsource of incomefor householdsin the homelandsthan earningsfrom agriculture (Cobbett1987; see Table 2.9). The statistics clearly indicate that with the exceptionof Transkei, Venda, and to a lesser extent Gazanalu, earnings from agriculture per householdare marginal in these territories. Althoughthe value of agriculural productionis greatest in KwaZulu, the high total populationhides its relative importance.

ChaP 2 S&nture, Rasonrce Ed_m, ad Permance ofth A.*xWr Sctor Page 42 SohdA fika_nAgriculture: Structure. PerformanceA Ontionsfor the Future

Table 2.9: Mimant andCommuter Earninas Compared witliAgricultural Eamines. 1985

Terrtory Total Migrant Total Aggregated Actual Agricultural and Agrcultural Household Household Eamings as Commuter Eamings Earnings Earnings percent of Earnings (R'000) (R'000) Household (R'00) Earmings Ciskei 449,625 19,564 294,464 2,356 6.6 Transkei 1,588,423 163,700 524,180 1,048 31.2 KwaZulu 3,028,582 208,000 2,384,339 3,266 8.7 Veuda 205,790 26,241 114,833 1,491 22.8 Lebowa 1,121,126 45,000 617,783 1,716 7.3 Gazankulu 307.168 23,850 151,970 1,475 15.7 Bophuthatgwana 1,671,247 52,400 962,227 3,353 5.4 Kangwane 382,180 17,600 253,825 3,384 6.9 KwaNdebele 361,651 2,300 174,952 3,645 1.3 QwaQwa 258,093 3,990 96,078 2,745 4.2 Total 9,373,886 562,645 5,574,651 2,381 10.1

Notes: 1/ These figures combinea monetaryvalue for subsistenceproduction with the value of commercialproduction. 2/ Calculatedby adding one third of migrantcarnings (assumedto be remitted in cash and in kiind)plus all the commuter arnings plus the value of total agriculturaleamings to give an aggregatedhousehold earning. 3/ Calculatedby dividingthe total number of householdsby the aggregated householdeamings. 4/ Calculatedby dividingtotal agriculturaleamings by aggregatedhousehold eamings. sourc: Cobbett 1987

2.81 Beyond the problemsrelated to insecureand fragmentedland rights, non-viableand smallfarm units, overstocking and the deterioration of the land, smallholder farmers face lack of support infrastructure, water supplies, transportationnetworks, financial support, and extension and research services. Institutionspresently providing farm credit are not alwaysproperly geared to providing credit to beginningfarmers, smailholdersand part-timefarmers; and researchto address problemsof technology in SouthAfrican agriculture has largelybeen gearedto the needsof large-scalefull-time owner-operators.

Chapter2: Structure,Resource Endowment,and Performanceof the AgriculturalSector Page 43 Soh Afrian Awukurre: Sructure. Perfrmce & o00bfor th Future

CHAPTlR. 3: EVOLUTIONOP 1HEJCOMMhRARY AGARN STP-UCTURE

introduction

3.1 'Me dual structure of agriculturein SouthAfrica and the comparativelylow productivityof small African farms described in the previous chapters is not the result of genuine economiesof scale In the large farm sector, but of decadesof governmentthat was policy guided-in large measure-by the general political and economicphilosophy of white dominationknown as apartheid. The results of these policies includedistortions in land and labor markets, output and input markets, infrastucture, agriculturalcredit and services, and the creation of large-scalewhite farms. Ihis outcomeIs not unique to South Africa, but is commonto colonial and post-colonialsocieties in most of Africa and Latin America.

3.2 This chapterhas two themes. The first is that African familyfarming was viable and successful in respondingto the increaseddemand for agriculturalproducts emanating from the miningcenters in the latter half of the 19th century. During that period, African owner-operatedor tenant farming proved to be as efficient as large-scale settler farming based on hired labor. African farmers adopted new agricutural technologies,entered new industries,and outcompetedlarge-scale settler farming in someof the emergingagricultural markets. Moreover,the ineffectivenessof colonialgovernment to intervenein agriculturalmarkets on behalf of seters, combinedwith the reluctanceof manorialestate2' owners-who relied on African tenant farmers for the operation of their farms-to support anti-Africanagricultural policies, resulted in a situation in which African farmers were able to accumulateagricultural capital, wealth, and farming skills.

3.3 The secondtheme is that the formationof a stronger, richer and unified settler state in 1910-the Union of South Africa-ushered in a policy environmentwhich suppressedand isolated Africanfarmers from mainstreamagriculture in order to facilitatetheir transformationinto rural and urban laborers. Ihe leverage of manorial estate owners-who previously frustrated farming policies aimed at constraining African tenancy-was greatly eroded by the miningindustry and junker estate2 ' farmers. The process by whichthe transformationof African farmersto laborerswas effectedinvolved the progressiveclosure of African access to most markets (land sale and rental, agricultura capital, inputs and outputs), the exceptionbeing a racially-segmentedlabor market. Currendy, African agricultureis associatedwith the economyof the homelands,where it representsonly a minor part of total income and in general fais to provide even the basic subsistenceneeds of the population.

IsA mnorial eae is an area of land allocatedtemporariy or as a permanet ownrship land holdingto a maal lord who has the right to tribute, taxes, or rt in cash,kind or in corvdelabor of the peasas residingon the edate.Mnorial can be rgaWizeas haendas or as lanord estat. A haciendais a manorialesat in whichpaut of the landis althvted as the home fam of the owner and part as the famiy fanmsof serfa, usufiutuary-rightholders, or tenantsin exchangeof rent in kind, labor or cash. Landlordesates are a ted entirelyby tenants(Bnw et al., 1993:1). South Afica's nanoial esttes wem assoted with seler landlordsand land companes.

2 A junker estateis a large landholding which protuces a diversifiedset of comnaodtiesopeated under a single managmt with hired lbor. Labors receivea plot of land to use for a houseand gardenfarming (Binswantr et al, 1993:1).

Chaper 3: Evoluion of the Contenpormy AgrarianStructure Page 44 uthAdkdrn49 £. SlWhfgutr-Pefonmanwa A O frAM Future PanningIn 19thCentury Soudt Affica 3.4 The agrarianeconomy of SouthAfrica in the mid-l9thcenr coisted of largescalewhite farms with hired labor, manorialsettler estateswith indigenoustenant farmers, and fee ndigenus farmingon black-ownedland. Ihere weretwo main geographic subsectors: coastal and Inteior faming. Coastalfaming-made up of all three typesof farms-producedhorticultue, livestock,and cp, and exported,wool, wine, and fruit to Europe. Interiorfarming, mainy by indigenousfarmers, raised livestockand cropsfor homeconsumption and marketedthe surplus. The lnd tenureaagemen of thesefarms varied from communal land ownership to privateor stateland ownership (both with nmous formsof tenancy,most often quitrent). Tradewithin and betweenthe interiorand oastal areas was dominatedby livestock,hides, and ivory in exchangefor guns, ammunition,texties, and tansport equipment. 3.5 The discoveryof diamondsin the 1870sand goldin the 1880sin the inteior changedilntior farmingsystem and the SouthAfrican economy and statecompletely. Large and rapidlygrowing urban and industi populationcenters sprung up aroundmining areas, creatingsubstanti marketsfor agriculturalproducts. The subsequentsupply response from African farmers confirmed for SouthAfrica whathas beenobserved elsewhere in the worldwith respect to the viabilityand efficiencyand of family farmingover large-scale farming based on hiredlabor at lowlevels of technology".Numerous studiese confirmthe efficiencyof Africanfamfly farming during the 19thcenury in responseto the growing demandfor agriculturalproducts from the miningareas. 3.6 With simpletechnology and plentifulland, laborwas the mostcritical factor in the successof frming. However,the relativeinefficiency of large settlerfarming implied low profitability,and the resuting difficultyof offeringwages sufficient to aa indigenouslabor awayfrom their ownfarms. Thisled to labor shortagesfor the settlerfarmers in someregions of the country. 3.7 Between1850 and 1870African farmers supplied the majortowns of the Englishcolony of Natal withgrain and exportedthe surplusto the Cape. In 1860,over 83% of the nearlyhalf million hectares of white-ownedland wasfarmed by Africantenants. Theiraccumuation of capitaland weth caused the NativeAffairs Commission (1852-3) to commentthat Africans were becoming wealthy, independent, and difficultto govern. 3.8 Manysettler farmers agreed with this view,because they wereunable to competewith Afica farmers,who producedhigher grain yieldsper hectareand cultivatedmore land tm they did. The setdersargued that laborshortages kept them from competing effectively with the Africanfaumers who had technology,equipment, farming skldl, and little or no needto workon setderfam.

fi Familyfarming is baed n fmily labora inc:ludesdrping, csh rem ad labo Unt fming- and ,idedat faming. Lre frming,on theother hand, is basedon atenayacquired labr whichincbludes i, indentured,ot wag lao.

2 Bundy,179 Trpido, 1975;Moms, 1976;KMegan, 1986; Benast, 1982 Wis, 1971;Iacq, 1982. ThIssectionul heavilyan Bundy(1970).

£7wr 3: Evbotionof * Coeora=7 Aarian Srucare Pae 45 $gWahAfrican Pf & ag for ON Eut

3.9 African fming was widespread both within and outside of settler manoril esta in the Transvaal and te Orange Free State. In 1870 the Trasvaal set aside 240,000 hect of treaty are plus an additional496,000 hectares as Africanlocations. However, the big landlordsand land cUmpanies continuedto enter into varioustenancy angements with Africanfarmers, and by the endof the century, half of the Africans lived on privately-ownedsetder lands despite a wide range of anti-squtr laws.

3.10 Short of labor, the large setter farmerspersuaded the colonialgovernment to intervee on ti behalf by: (1) limitingAfrican competitionin the marketplace; and (ii) settingup native reservesof tiny pieces of land to create an artificial land shortage in order to force African farmers to seek wo on manorial farms. To this end, various measureswere used includinglivestock, hut, and poll taxes; road rents; location, vagrancy, and pass laws; and confinementto the reserwes. TIese measures, while successful in constraining African owner-operatedfarming, did little to reduce competidon ong manorialestates for Africantenant farmers. Obviouslythe estatesoffering African farmers better tenancy conditionsreceived more tenants. In response to settlers' request to reduce such competiti, the stae then intervenedin the land rental market and sought 9o reduce the number of rent-payingAfrica tenan farmers on manorialestates.

3.11 When their problems seemed insoluble, some settler farmers in Queenstown (Cape Colony) proposedthat Africanfarming be abolishedaltogether. But a grup of white merchantswhose well-being dependedon marketingAfrican produce campaignedsuccessfully to have this proposa rejected. They argued that settler communitieselsewhere in the country which had adopted such measu ended up in a depressed economic condition, because Africans were better farmers an the white setders ad representeda substantialshare of purchasingpower in the local economy. Ihey warnedthat the forcible removalof Africanfarmers from the land and replacingthem with a smallnumber of setlers would result in economicdisaster for business and the white urban communityin general.

The Situationat the Turn of the Century for the Af= Fa

3.12 After the discovery of diamondsand gold, British imperial interests, had compellingeconomic reasons to bring the Transvaal and the Orange Free State under control in a federaion. The Britsh attemptsto take control of the Boer states ultimatelyresulted in the Boer War, beginningin 1W-a bitter and divisive conflict from which Britain emergedthe victor in 1902.

3.13 During this period many Africans,especially in the Trasvaal and the OrangeFree State bought land as individualsand in groups as land syndicates. Missionarieswere often used as 'fronts,' because buying land was difficult for Africans. Land buying was given a boost during the brief post-war period of British rule (1902-1910). Africanswere allowedto convert their labor tenanciesto sharecrpping or fixed-rent(cash) tenancies, or to purchaseland outright. No exactinformation Is availableregadig te amountof land bought duringthis period', but some settlers speculatedt1hatAfricans would succeed in buying back all the land they had lost during the colonialwars laasje, 1987). By 1904, of a total of 900,000 African households, 49% (438,000) lived on white manorial estates as tenants on hacendas,

I Onm eatneis thatby 1900,around 250,000 acs a bm buh by blackfame in theThaa a

Chapter3: Evobdion Qfthe ComenmporamyAgrarian Strucmure Pge 46 SouthAfrican Arlculture Symctr. PrformanceA Optionsfor Me Ftre where they receivedcultivation and grazing rights from ttie residet owner in exchangefor their labor; or squatters on landlord eswes, where they paid rent to the non-residentowner In cash or in shares of tie crop. Twentypercet (180.000)leased state land, 14% (130,000)farmed their own private land, 116% (123,000) were on governmentlocations, and only 6% (50,000) were hired laborers.

3.14 African farmers achievedtheir success not in an Ideal competlive market or a supportivepolicy envronment, but In a hostilesociety determinedto undermineit. A series of levies and fees imposedon Africansbetween 1903 and 1905forced them to pay higher incometaxes ta whites. Resictive taes and laws already passed in the 19th century were strengthenedby the 1908 Natives Tax Act which imposedtwice as much tax on sharecroppersand other fixed rent-payingsquatters as on labor-tena in an effort to reduce the attractivenessof the former types of tenancy which were more profitable to the black tenants than the labor-tenancycontracts. Measures to remove squatters by force were also undertakenin 1909and 1910. Tenantswere alsoincreasingly forced to sell off or removetheir allegedly excess catte, i.e. all catte other than the cattle that the tenants used to plow the fields. Measures were taken to restrict Africansfrom the profitable rural transportationsector.

3.15 The state, however, did more than hinder black farming; white farmers received substa l support from the governmentin the form of subsidies,grants, and other aid for fencing, dams, houses, veterinaryand horticulturaladvice, subsidizedrail rates, specialcredit facilities, and tax relief duringthe period 1890-1908. In 1908, the TransvaalDirector of Agriculturedeclared that during the past twenty years more moneyhad been spent per head of SouthAfrica's white firming populationthan in any other country in the world (de Kiewiet, 1942:260).

3.16 Despite all of these efforts to restrict African farming either to the reserves or manorial estate and to subsidizewhite farmers, the black farmerscontinued to maintaina competitiveedge in the market place. The chronic labor shortageson white farms were intensifiedby the emergenceof the minin and manufacturingindustries with their massive labor demands.

Agraian Development:1910-1947

3.17 The Unionof SouthAfrica was formedas a dominionof the BritishEmpire in 1910. At this time there were nearly six millionpeople living in SouthAfrica. More than two thirds of these were Africans, more than a quarter whites, about a tenth coloreds, and a very small mnmberof Asians. Ihe South AfricanParty, representng landlord and mininginterests, headedthe first govermenL One of the goals of the new governmentwas to ensure adequatesupplies of labor to the mines. To achievethis objective, numerousrestrictions were imposedon Africanfarmers, thereby forcingthem to serve as a much needed source of labor in other sectors, but chieflymining. Many of the manorial estate owners who had been supportiveof African farmers in the past and had often preventedhostile government policies were co- opted and became part of the new social order.

apter 3: Evohu*n of she ContVoravy AgrarianSructure Page 47 So Afax Amrkcutur.Structure. Perobrmance & Optlons for the Ftre

Table 3.1: African Populationson Various Classmsof Lands. 1916

Reser Mission Native- Crown ma Poa= TotalPopulaion Lands Lands owned Lands Lands Occupiedby Occupied Rual Ua Africans by

Cape 1,149.4 24.3 39.3 12.5 7.6 240.4 1,473.S 128.0 Natal 479.8 44.6 39.3 37.1 85.6 357.9 1,044.1 37.9 Transvaal 283.2 24.0 40.4 71.5 232.1 408.6 1,059.9 322.5 O.P.S. 17.2 1.8 4.7 - - 279.4 303.0 48.8 Union 1,929.6 94.6 123.7 121.1 325.2 1,286.3 3,880.5 537.2

Dre: BeaumontCommission Report, 1916.

3.18 Ihe Land Acts. 'Awakeningon Friday morning,June 20, 1913,the SouthAfrican native found himself, not actaly a slave, but a pariah in the land of his birth."'30On that date, the Native Land Act No. 27 drew a firm line between white and black landholding,prohibiting each from 'entering into any agreementor transactionfor the purchase,hire or other acquisition...ofany such land Cmthe area allotted to the other) or of any right thereto, interest therein, or servitude thereover". The Land Act of 1913 segregatedAfricans and Europeanson a territorialbasis by establishingcodified native reserves, referred to as "scheduledareas". Independentblack agriculture and cattle-raisingcould now only be undertak in the nive reserves. About 7.8% of the country's farm land was in the 'schedule" for the reserves. 'Iis becamethe only area where Africansubsistence farming could legallybe conducted. The Act was specificallydesigned to end both landlord and hacienda manor estate farming which relied on various forms of tenancyand to establishin its place large land-holdingcompanies (unker estates) which would operateunder a single managementwith hired labor. Laborerswould receive a plot of land for use only for house and garden farming.

3.19 The Act defined natives as membersof an aboriginalrace or tnbe of Africa and prohibitedthem from renting, buying, or otherwise acquiring land outside the reserves without the approval of the Governor General. It also excludedfreehold, mortgaging,and land sale rights within the reserves and stipulatedfurther that no person other than a nativecould acquireland rights within these reserves,except sites for trade or business, without state approval.

3.20 In 1916, after the Land Act was put into effect,about 50% of rura AIicans lived on the reserves, 2% on missionlands, 3% on native owned lands, 3% on state lands, 8% on landlord estates, and 33% on haciendas. Of the total population, 12% lived in urban areas (Table3.1). Afrikanerswere the white majority in rural areas, while the English were dominantin the towns and cities. Table 3.2 shows the ownership and occupationof lands in 1916.

Sol Plaaje (1982),Fuist Secretary General of theAfican NationalCongress.

Chapter3: Evoluion of the ContemporaryAgrarian Strutre Page 48 Soush fcan ArFrk4uur.:Structure. Perfomance & Ootons for the Fure

Tabl 3.2: Land Aras bv Lad Tnre Sytems in th Unimonof SouthAfrica )MM

Native reserves 9,538,300 Missionresves 460,000 Native-ownedlands 856,100 Crown lands oocupied 805,100 EOL: Occpied by Europeans 90,314,000 BOL:Occupied by Aficans 3,550900 Vacan Crown lnd, rserves and other 17,002,400 TotaL 122,526.800

HSft: OL: Eumpean-OwnedLand occupiedeither by (1) European; or (2) Aicans. a. Reportof the BeaumontCommission, 1916, pp. 3 and 4. b. 1990 SouthAfrican Statotics, p 26.2; and DBSA, SouthAfiica: An Iter-Rgional Profile,pp. 34. Exclutds Walvr Bay.

3.21 Ih.e Beaumnt Commissionwas appointedunder prvisions of the Land Act to organize the reserves. The Commissionreported that the scheduledland was only sufficientfor approximatelyhalf of the Native population and recommendedthat further land be released to the reserves to ensure territorial segregationand specifiedthe areas which should be added.

3.22 As Table 3.2 indicates, the reserves were limitedto 7.8% of the total land area. Outside the eserves natives owned only 0.7% of the land and lived in state and European-ownedlands (another 3.6%); thus the total land for native use was 12.1%. Yet the white populationstrongly opposed the BeaumontCommission's recommendation for an increaseof the reserves. Consequently,the situation remainedunchanged until 1936, when the Native Trust and Land Act, No. 18 was passed establishingthe NativeLand Trust. The Trust releasedthe recommended6,209,858 hectares (quotaland) to the original 1913scheduled reserve areas (non-quotaland) increasingthe size of the reservesto 13.7% of the country (Table 3.3 for the breakdownof this land).

Table 3.3: Land Ouots Delimitedby the 1936 Act

Arma pewst (hectares) Transvaal 4,306,643 69.4 Cape Province 1,384,156 22.2 Nalal 450,536 7.3 Orange Free Slate 68,523 1.1 Total 6,209,858 100.0

: Bantu Trst and Land Act No. 18 of 1936. a hp, 3: Evokafonof she ContemporaryAgrarian S&rwwuru Page 49 South African Airkcubure:Sructurtw. Perfonnwnce & ODtonsfor the Fmure

3.23 Other measuresundertaken to support the Land Acts includedthe Native Adminion Act of 1927 which intoduced yet another encroachmentby the state on the indigenoustenure system. Ihe Governor General (the State President today) was imposedon Africansas their 'traditional' Chief with decisive powers on African land matters. African chiefs were reduced to salaried officers of the government.

3.24 The need of a cheap supply of labor also led to several new goveramentacts. 'Me Mste and ServantsAct of 1911 and 1932tightened the grip of junker estatemanagers on farm workers: it prohibited breakdngcontracts, changing employers, or assigningof familymembers to other employers. The Native RegulationAct of 1911was modifiedand used to register with a labor bureau all African maleand female workers over the age of 16 seeking work. Once registered as farm workers, they could not switchto industrialemployment (United Nations, 1976). The Prison Act allowedestate fanners to serve as wardens on Africanprison farms and have the benefit of cheap prison labor.

3.25 African farmers were not allowed to join state-sponsoredmarketing cooperativesor farmers unions. Without membershipin these organizationsit was difficultto secure credit, market output, or obtain researchand extensionservices. Moreover,outside these agricultural institutions little Independent activity was permittedby law. (See Chapter 4.)

3.26 The combinationof all these measuresbegan to erode the developmentof African farming, and by the 1920s increasingpopulation pressure caused Africanhouseholds in the reservesto spend 60% of their incomeon food. Infant mortalityincreased and signs of malnutritionbegan to surface. Landlord estates relying on African tenants, however, managedto survivethe Land Acts for some time. In 1925 they still accountedfor 4.5 millionacres with fixed-rentor sharecroppingtenants. But as these changed intojiunker estates, Africantenant farming declined. Tenantshares becamesmaler and smaller, and labor tenancycontracts became longer and stricter (Morris, 1976). By graduallydestroying tribal instituons and closing many incomeearning opportunities-the exceptionbeing labor markets-the capital, wealth, farming skills, and informationbase that African farmers had accmulated over generationsbegan. to wither away.

3.27 The planned transformationof the agrarian structure from haciendasto junker estates did not resolve the problem of low profitability or labor shortages for large-scale farming. African owner- operatedfarming was effectivelyeliminated, but landlessfarmers often chose to work in the mines, in new manufacturingindustries, or in towns, rather than onjunker estates, whichalmost from the beginning were notoriousfor their low wages and bad working conditions.

3.34 When the 1932 worldwiderecession hit on top of these repressivelabor measures, the flow of workers from the farms to urban areas became a deluge. Between 1936 and 1951, the largest single sourceof newly urbanizedAfrican people was the white rural sector. Five times as many Africanscame from whitefarms as from the reserves. Ever increasingamounts and kinds of state support were required to keepjunker estatefarming alive.

Chaper 3: Evouwonof he ContemporawyAgrarian Structure Page SO Sou fricanA,ricultwe: Structure. Perfonnance& Ontons for the Future

1910-1947: Protectionof White Agriculture

3.28 The white farmingsubsector experienced a completelydifferent treatment from that of its Afiican counterpart. ITe support systemfor white farmers, alreadylong in place, was strengthenedsubstantally to enablethe effectiveoccupation and use of the land that was set aside for white ownershiponly. The Land and AgriculturalBank of South Africa (the Land Bank) was establishedin 1912 to give loans to farmers who did not have access to commercialbanks. The CooperativeSocieties' Act of 1922 began the cooperativemovement, and the AgriculturalMarketing Act of 1937 created marketingschemes and control boards to administerthem (see Chapter 4 for details). Research and extension services as well as the necessaryinfrastructure were providedto ensurethe success of white farming. Between1910 and 1935, eighty-sevenActs were passed in the UnionParliament rendering permanent assistance to farmers, financeddirectly and indirectlyby the revenuesfrom the miningsector. Between 1910and 1936,the State spent £112,000,000from revenue and loan funds on agriculture. Between 1932 and 1936 alone, the operationsof the Land Bank and the direct expenditureby the governmenton behalf of the white farmers amountedto £20,428,092. On export subsidiesalone, a total of nearly £11,000,000 was spentbetween 1931 and 1937 (de Kiewiet, 1942:253-261).

3.29 White Land Settlement. Whereasthe NativesLand Act of 1913restricted Africanaccess to lnd, the Land Setlement Act of 1912 standardizedthe acquisition,exchange, and disposalof state lands for white settlement. The terms and conditionsof the Act were updated in the Land SetdementAmendment Act of 1917 and many subsequentamendments. These acts establishedthe proceduresby which white settlers could apply for state and privatelyowned lands.31 The Ministerof Lands was empoweredon the recommendationof the Land Board (a counterpartof the Land Trust), to allot state lands and to use public funds appropriatedby parliamentto buy privately-ownedland for sub-divisioninto suitable agniculral holdings for white farmers.

3.30 The size of the farms acquired ranged from large cattle ranches to smallholdingsfor market gardens. The Land Board prepared a report on the holdings; if the results were favorable,the land was surveyed,divided into farms, appraised,and advertised. Listingsof farmswould be publishedfrom time to time in the GovernmentGazette and local newspapers, with particulars of the land offered, the governing Land Acts, the terms and conditionsof the lease or purchase, and when and how to apply. Applicantscould obtain farms by leasing for five years (renewable)with option to buy or by transfer of land purchasedespecially for them.

3.31 Leased. The applicanthad to be of good characterand at least 18 years old, possessqualifications sufficientfor utilizingthe land, and intend to occupy, develop, and work the holding. The land was initially leased for five years. No rents were charged the first year; in the next two years, 2% of the value of the land was charged and 3.5% the fourth and fifth years. The Minister could extend the lease for a further five-yearperiod at a rental of not less than 4% of the land value. The option to purchase

3 Theseamenmen amog ohers includefte LandStlemn Acts Furte Amedm Ad No. 28 of 1920and No. 21 of 192; theLand Settmet LawsFurth= Amaende Ad No. 26 of 1925;Ihe La Se_lmnt (AmendJm) Act No. 6 of 1928, No. 1 of 1931,No. 57 of 1934,No. 47 of 193S,and No. 45 of 1937;the Iad ScdlementRiefAct No. 25 of 1931;ad tholmd Setdlmnt Act No. 21 of 1956.

OYapter3: Evolutionof theConteMpory Agrarin Stetuare Page 51 South African A&rlcsuure: Sruaure. Perora-n_ce 4&OoMs for the Future the holding, by paying the purchaseprice over a period of 20 years, could be exercisedat any time with the Minister's consent, subject to fulfillment of the conditionsof the lease. Instalment payments amountedto about 7.33% of the purchaseprice per year, whichIncluded anterestat 4% with the residual applied to principal (1916 Year Book).

3.32 By 1941, the conditionsfor leasinghad becomemore generous.No rent was payablethe fit 1= ,yea; the other charges remained almostidentical. The option to purchase was extendedto 65 ymu althoughthe purchaser had to pay installmentsof at least £25 per year. Reducedrent and interest applied in certain semi-aridareas.

3.33 Pu,rch-a,sd.Ihe governmentset a maximumprice for the agricultual holdingsof £1,500 which includedsurvey fees, transfer costs, and two years of accruedinterest. The governmentloan was to be no more than 80%. Ihe buyer was required to pay 20% down to purchase the land, with no fiurher payment for the next two years; the repayment schedule for the next 18 years was In six-month installmentsof 7.85% of the remainingbalance. Four percent of this was interest, the rest principal. (1916 Year Book).

3.34 By 1941, the governmentwould loan up to £2,250 of the price, and the repaymentperiod was increasedto 65 vr By 1956, this was increasedto £5,150 and the purchaser's down paymet was reduced to 10%.

3.35 Bothmechanisms for obtainingland requiredpersonal occupation, mandated care of improvements and trees, specifiedadditional land improvementsamounting to 10% of the value of the holding witin the first five years, and prohibitedsub-letting, mortgaging, or transfer withoutpermission. Leases were subjectto cancellaton in situationsof non-compliancewith provisions in the Acts. The 1956Resettement Act further strengthenedresidency requirementson the holdings, but permted the Mister to gramt exemptionsfor a temporaryperiod on recommendationof the Lad Board.

3.36 Althoughthe Land SettlementAct of 1912 was the standardbearer, the allotmentof sate lamd to whites for agriculturalor pastoral purposeswas also implementedunder many earlier acts. Table 3.4 provides an overview of the various leases and purchasesgraSted under these acts in 1916 (1916 Year Book). Between 1910 and 1936, about 700 farmers per year were settled, supportedby massive state subsidies.

apte 3: Evoldon of de Contemporay Agarian Sruaurwe A 52 SWhAfrikan Aifcubre: Srnc:ure. Performawceh Odons for the Future

Tablo 3.4: Allotnent of Ariculuns Holdignsburinf 1916 No. of No. of SetUts Area Amourt Rat Holings (bccares) (f) (f)

Land Settlment Act 1912 141 210 168,636 110,053 -

Crown Land Diposal Ordiance 123 134 90,557 58,215 - (Transvsad).

Crown Land Dspoul Ordinance 26 26 21,414 10,654 - 1903 (Transvaal). Act 15 of 1887 (Cape): Sales 12 13 4,356 993 Act 26 of 1891 (Cape): Lases 24 25 19,291 - 523 Act 26 if 1891 (Cape): Sales 2 1 7,621 395 - NatalProclamaion 36 35 28,711 13,026 53 Inigation Setdlme Act 31 of 22 22 120 3,353 - 1909 Act 13 of 1908 (OPS): tscss 3 7 2,085 - 14S Totd Land Alia"ed 389 473 322,791 196,689 721

so: SouthAfiica Offici Year Book 1916.

3.37 In additionto help in acquiringland, loans were made to help white fanners obtain stock, seeds, equipment, and other items needed to develop their farm. Permanent improvemens such as driling operations could also be funded and the cost added to the rent or purchase price of the land. Ihese advances could not exceed £250 (£500 by 1941). Short and medium-termloans were also availe to lesseesthrough the Land Bank for fencing, constructionof dipping tanks, and other improvements.

3.38 One result of this period of strong governmentsupport was the growth of the number of white fams from 81,432 in 1921 to a peak of 119,556 in 1952.

1948: Apartheidand the Rural African Household

3.39 In 1946, the African populationliving in the reserves had droppedfrom 50% in 1916 to 40%. The Social and Economic Planning Council reported that the quality of the reserve lands was deteriorating, and the 1948 Fagan Commissionfound that many Africans were becomingpermanent inhabitantsof the towns.

3.40 Campaigningon a platformof more rigid racial separationfor blacks and power and wealth for whites, especially Afrikaners, the National Party came to power in the 1948 election. The previous government, although adhering to segregation, had tried to address African land gdevances in a conciliatoryspirit; the new governmentabandoned all such pretenses. Ihe new 'apartheid' went fiuther

C0Wer 3: Evlon of tOaContmporwy Aga Sucrwre Page Sou&thAfc Aeibre: Stwucture.Perfnnmawe O t2Loafor th FWure than previoussegregation policies by alsosegregating African ethnic groups from one anotherand forcing them to live in separate tribal areas.

3.41 The objectivesof the new policy were to: (i) facilitatewhite politicalcontrol by dividingAfrican nativesalong tribal and ethniclines; (ii) protectthe white workingclass from AfricancompeWtion through variousdiscriminatoq laws; CMii)further racial segregationso as to preservethe whites' cultura identity; and (iv) reduce the cost of developmentof industrialregions (Cobbett, 1987). Unstaed objecdves wero to retan the black agraian structure establishedby the Land Acts, to continuewhite land settlement,and to advancethe white agrarianeconomy by establishinglarge commercialfarms. Thesefarms would differ from junker estates in that the main supply of labor would be migrant workers. African farmers from the homelands would provide it. These objectives were largely achieved. Under apartheid, the old agrarian economydominated by junker estates was transformed into a new agrarian economybased on large commercialfa=ms32.

3.42 ConstitutionalFramnework. The NativeAuthorities Act of 1951 and the Promotionof BantuSelf- GovermnentAct No. 46 of 1959 artificiallycreated eight nationalunits out of the Pede, Sotho, Tswana, Swazi,Tsonga, Venda, Xhosa, and Zulu ethniccommunities. Ihe boundariesfor these *nationalunits,- not surprisingly, coincided with the reserve boundaries as defined by the Land Acts. The Transkel became the first self-governinghomeland in 1963. Nine other homelandsfollowed.

3.43 In 1970, the Bantu HomelandsCitizenship Act was passedmaking every African in the Republic a ciizen of some homeland. The Bantu Affairs Adm ition Act of 1971 transferred control over Africans, regardless of where they lived, from white local authoritiesto Bantu Affairs Adminiraton Boards. The Bantu Laws (Amendment)Act of 1972justified forced resetdementsof African people and stated that "a Bantu tribe, community,or individualcould be removedfrom where they lived withoutany recourseto Parliament, even if there was some objectionto the removal.

3.44 These Acts gave the governmentthe right to banish to a homelandany Africma considered "redundant"in urban or white areas and then abrogatedtheir consdtiutionalright to belongto the Republic of South Africa.

3.45 The Commissionfor the Socio-EconomicDevelopment of the Banto Areas within the Union of SouthAfrica of 1954, known as the TomlinsonCommission, placed the ultimatesize of the homelands at about 17 millionhectares (Fable3.5). This includeda) the originalscheduled land from the 1913Land Act; b) the land recommendedto be added to the homelandsat that time by the BeaumontCommission, but actually not released until 1936 (quota land); c) other land acquired by the Trust; land owned by blacks prior to 1936;and d) 'black spots," isolatedsmall areas of land occupiedby Aficans but located outside of the homelandsand of land planned for the homelands.

3.46 The Land Acts had long attemptedto consolidatethe "blackspots' by eliminatingsome that were 'poorly situated," i.e., in white areas, and combiningothers to make them into homelands. The

2 A Wrgeownr holdingwhich produces seeral difereit commodit. opezatingunde a singl nage wih a high degr of mechanii using a few long tm hired wores who mayreside an the farm and sooAlly hired workerswho do not reside pmanely on the farm (Binwager et al, 1993:1).

Chaper 3: Evowon f tdh ContemporaryAgrarian Sthrue pas 54 South A*c AMrckub: Strucue. PEnfonnac & Oto* tf *e Futu-e deidnatlonof the spots,however, meant that the landarea had to be subtractedfrom the amountavailable for the homdand.

Table 3.S Tarue Afiican as (ha) Dosinatd by OhmComrission

Scheduld black ars (1913) 9,190,101 Reease blak areas (1936) S,81S,460 Blanoo of quota land outstading (19S5) 1,630,692 Land Ownodby black sibued outside scheduledand 161,593 relased black armas(ie., black spots)(1955) Total 16,797,846

3.47 The TomlinsonCommission of 1954was concernedthat the qualityof the landIn the reserves couldnot support the 500,000African families living in theseareas. Ihe Commi.donproposed a drastic cut in the numberof familiesin the homelandsand a seriesof Bettermentor CloserSettlement Schemes to stop soil degdion throughland use planning,relocaton of peopleand livestock,stock-culling, fencing,contour ploughing, water conservation, and erosioncontrol. It also urgedthe Trustto haste te processof buyingthe oustandingquota land. ByMarch 1967, however, the Trust had onlymanaged to buy fivemillion hectares of quotaland; 1.3 millionhectares was stil outanding. 3.48 In 1968the administrationof the NativeLand Trust was handed over to the Departmentof Bant Affairs. This removedthe acquisitionof landand its allocationfrom the agendaof the 1936Land Act and placedit withinthe agendasof the homelandand industraldevelopment policies. This essentaly meantthe resettlementof Africanpeople in homelandsand a furtherallocation of land only if the homelandsopted for Independence.Table 3.6 describesin detailthe land situationof the homelandsin 1976.

3.49 AIthou the statedbasis for the exchangeof holdingswas the qualityof land, i.e., an African frmer shouldbe resettledon land equalto his formerholding; in practicethis wasnot the case. Much of the landreleased for the homelands,often bits on noncontiguousscrubland, certaiy did not meetany quality-of-landstandard.

3.50 The relocationof Africas to the homelandsfrom whiterural and urba areas, Africanowned areas, and fromone placeto the other withinthe homelandswas never voluntary.Ihe SurplusPeople Projecthas estimatedthat 3.5 millionpeople (predominandly Africans) were forcibly resetded between 1960and 1980for various reasons: eviction of blacktenas and uredundantworkers from white farms (32.1%), intra-cty removals due to the GroupAreas Act3 (23.7%),homland consolidationand clearng of blackspots (19.1%), urban relocation from white areas to homelands(19%), removal of infrmal and unauthrized urban settlements(3.2%), reocaton due to developmentshemes (0.7%), politIcay motivatedremovals (1.4%), andother factors(0.9%).

X le GroupAnn Act of 1959desiad all of SouthAfican land for theexclusive ue of one nacialgroup or andi. aAWr 3: EwAbon of kW Comaeaor, graa Stracie Page s5 Tab 3.6: (UmtRNW NM.sa Lan flhl. I 1976

(a) (b) (o) Ttar (d ( () T Tct Tnot Blk Bla Blak Tnit Tut .v. d d d owned d v_td -d Pd"to SON 311U.36 3118J36

Leboa 278fO3 1,081,361 230,921 1,590.786 222,32S 455 414,767 39,824 677,371 3Soth Ndcebd s60 69,143 4,902 74,606 S30 - - - 530 LI Gaktdn 387,711 21S,933 7.907 611,S51 18,350 44.483 63,013 S ISI,29 114,001 4.028 269,SS8 5,16 - 27.336 2.088 34.591 DOPhAaUwS 49,185 657,304 129,955 836,445 S06,489 18,619 139,80 IS,177 6S0.082 Venda 347,489 101,071 5.|62 454,422 17,271 . 177,807 700 195,778 l[umdo 3255 51,060 - S4,315 - - 13,843 - 13,843 TOa 1.218,232 2,289,873 383,576 3.891,683 770,312 19,074 S18,836 S7,789 1.665,211

Xc Pmaviec Ciki 53,966 115.462 2,867 172,29f 32,282 1.661 729,662 11,969 77SS75 Tiaei S,239 143.766 9,031 IS8.036 84.471 3,SS 3,482,708 24,893 3,59S,623 Dophutbatiwans 74.833 49S.523 11,986 S82,343 2S,5S2 4,433 1,386,481 194,911 1.611,377 TOd 134,038 754,751 23,884 912,674 142,305 9,645 5,S98,851 231,773 S,982,57S

Kwuazlu 48,022 3SS,372 17,847 42140 138,945 1,854 2.604,016 7,209 2,752,023

B *_ 53,972- 1,993 55,965 26,90 723 20.725 59,878 Q _eqwa - 5,417 - 5,417 - 42,827 42,827

Total - 59,389 1993 61,312 26,903 723 63,SS2 102,70S GyendTotal 1,400,292 3,459,385 427.30 5.286,979 1,078,465 31,296 9,054,4S 308.298 10,502,S14

~m: DDeputmu of B Aut iad Deovlpsat. U_AMhd dua. 1. QumaLAW (a) Trust vee hat Govemm lad in teo glsbhe own (Act. No. 18 of 1936) which w ted in doe Tndt inh of Sudin 6(lXb). (b) Twmt acqubedlea Lnd acqurd by do Trust dsme 1936outside do scebduled bla son (Act No. 27 of 1913). (c) Bl* aquire lad: Lad acqure by blak in 1936outside h seduld blak am. which ans sod in arelems an sdjwnnit o adjoinung scheduledblacksc. 2. Noe-Qot Lad (d) Black owned lad ea 311836: Lnd owed by bl b r to tedoSuh Trit and La At, No. 18 of 1936. (c) Blak aquied since 3136: Lad acied by blba in s l bk e. M()Tut td had: gan a bad in scheduledblack sun which wa vesed ir the Trit 4 (g)TnIt aqurd lad: Lad aqurd by Ih TbW in scheduledblack . 3. Due to _ndg we fur az bNae_ the - of fips my dir sigby fre t figu saly give i the to cehmn. SouthAfrkan Agrkgukurs:Strucure. Performance A Opilonsfor the Future

Ile SituationToday

3.51 SouthAfrican whites voted narrowly In October 1960 to adopt a new constitutionand make the countrya republic,which was establishedas of May 31, 1961. The Republicsubsequenty withdrewfrom the British Commonwealth.Racial strife increased,leading to internationalfinancial sanctions and to a reassessmentof apartheidby moderatepolitical leaders. Part of that reassessmentwas to decidewhte can constitutea viable and sustainablebasis on whichto build a new agrarian structure.

3.52 The land owned by the Trust prior to the 1936 Land Act and the amount of land owned and acquiredby blacks and the SADTFund from 1936to 1990 are set out in Tables 3.8 and 3.9. A total of 10,442,437hectares of non-quotaland was made availableunder the 1913 Act. The area of quota land acquiredby the Trust by the end of 1990 was 7,203,166 hectares,72.5% coming from land acquiredby the Trust, 21.0% from state land, and 6.5% from land purchasedby blacks.

Table 3.7: Non-Ouotn.Land

Non-Quota Land Acquired by State Land Non-Quota Land Land Owned by Blaks after Vested in thc Acquired by tho Blacks 31/U36 31/U/36 Trust (ha) Trust Total (ha) (ha) (ha) (ha)

T Iansvaal 748,754 19,074 818,042 74,338 1,660,208 Cape Province 141,148 9.824 5,514,404 264,389 5,929,765 Natal 135,757 1,877 2,598,657 13,468 2,749,759 OmangeFree State 26,727 723 63,552 11,703 102,705

Total 1,052,386 31,498 8,994,6SS 363,898 10,442,437

Sourc. Departnt of Developme Aid, 1990 Annual Report, p. 15.

Table 3.8: Ouota Land as of 31 Docember 1990

State Land Vested in Land Acquired Land Acquired the Trust by the Trust by Blacks Total Oa) (ha) (ha) (ha) Transaal 1,249,050 3,213,274 414,134 4,876,458 Cape Province 195,252 1,323,260 28,869 1,547,381 Natal 66,617 492,135 20,587 579,339 orne FPreeStat 253 197,742 1,993 199,988

Total 1,511,172 5,226,411 465,583 7,203,166

ac: D_epatent of Devlopment Aid, 1990 Annual Report, p. 15.

3.53 The present agradan structure has been systematicallyworked out since the turn of the centy. By the end of the 1980s, the African family farming sector had all but been eliminated, and African peasantshad been transformedinto wage workers on large firms, in mines, and in secondaryindustdes. apyler 3: Evoluon of dheConunorwy Arari Strutre PaSe57 South icatnA&dcutre: Strucwr. P-rM M OnAns for the Fhtur Nearly90% of the agriculturalland was in whiteareas supporting a totalruna populationof 5.3 million people,more than 90% of whomwere Africans. The remainingagricultural land was In the homelns and supportedover 13 millionpeople. Originallythe homelandswere justified as areaswhere Africans woulddo subsistencefarming; today up to 80%of somehousehold incomes comes from migrant eunings and pensions. 3.54 In 1916,every African except those on State(Crown) and Europeanowned and occupiedlands ownedat leastfour hectares. By 1990the individualholdings had droppedby 75%,to one hectare per person. Thishappened mainly because of thephenomenal population growth In the reserves,whose area hasremained relati. 31ystatic over the period. Todaymost homelands are peripheral,over-crowded, and poverty-strickenand lack comparablein&fastructure, despite some familyfarmer support development programs. Genuinedevelopmental strategies were frustrated by the overwhelmingproblems of these areas. Althoughthe proportionof statespending for agriculturein the homelandsincreased in the 1980s, onlya verysmall amount represented actual transfers to firmers. Statespending for servicesfor Africans Is also lowerin the homelandsthan in the whitenra areas. 3.55 Landtenure in the homelands.Communal tenure in the homelandsis officialydefined by the ProclamationR188 of 1969as 'unsurveyedland' or 'permissionto occupy". Underthis Proclamation, a maleperson holds rights to variousland allotments for residentialuse, arablefarming, and grazing. Landaccess is usuallyby virtueof membershipto a community,not throughsale, leaae,or rent. Only menare entitledto inheritland rights. Peopledo notlegally own their residentialand srableallotments. Ratherthey are allowedthe right of occupationand cultivationsubject to conditionsstipulated by the homelandauthorities (de Wet, 1987). 3.56 Accordingto Murray(1970), approximately 15% oftheland in thehomelands is heldonfreehold or conditional(quitrent) title. But Africanfreeholds rarely belong to a singleentrepreneur who farms usingeither family, hired laborersor tenants. Most freeholdpurchases were undertaken by etended familiesor syndicates.These farms are heavilypopulated by descendantsof the origi purchasersand extendedclientele. Todaythey are more peri-urbanin nature m rural firmlands.Many freeholders havebecome landlords to residentialtenants. Freeholds once represented a "progressiveideal of an area whereAfricans could lead a life-stylewithout government interference. There has beena steadyerosion of those ideals. Classdifferendation and conflictshave emerged. In somefreehold areas of Natal landlordscollect their rents under armed guard (Cross, 1990:22). In other areas, patron-lient relationshipsare exploitedto fiutherpolitical disputes. Violence in Natal'stenancy areas is exacerbated by 'politicized local patron-clientfollowings controlled by warlords (Cross, 1990:22). Pocketsof black freeholdareas have alsobeen involvedin the state'srelocation policy. Propertyhas been expropriated and the occupantsresettled. 3.57 Be1eerentplavWng. Bembridge (1986b) esdmated that about 70% of SouthAfrican black rural areas are officiallyunder "bettermentplanning". Bettermentplanning is SouthAfrica's attemptat villagization,i.e. plannedvillage land use. Before'Betterment', people lived in clustersof homesteads, alonghills or ridges,with their fieldsnear rivers and streams. Theygrazed their catdeon the hillsand in the forests,or furtherfrom home. With'Betterment' they changed to newfields and to newresidental areas. ne newland use system was inflexible; people found themselves with smaller fields and gardens thanbefore; had to walkgreater distances to fetchfuel, waterand thatching grass (de Wet, 1987). This

atr 3: EvowuIonof the ContemporaryArarian Struur Page58 SotdhAfrn Airukure: Strutre. Perfirmance Onions*r the Future was accompaniedby very unpopularstock cullingmeasures triggering peasant resistanceto 'Beterment' in the 1940s and 1950sthroughout the homelands(Cross, 1990).

3.58 Mjgwl,tbor. More than a century of contract-wagelabor in the miningand the commercial farming sectors, and more recendy in the ufactring and service sectors, have perpetuated migrant labor in SouthAfrica. Until the mid-1980s,workers who left the rural areas to seel' h etter opportunities in towns were preventedby law from taking their familieswith them. 'he removalof these influx control measureshas meant that black workers have theoreticallybeen permitted to sell their labor and settle where they please within the confinesof the group areas. Cyclicalrural-urban migration has thus become a way of life in South Africa, where the majrity of rural householdsare better viewed as membersof dislocatedurban communities(Nattrass and May, 1986).

3.59 Data provided by Halbach (1988) demonstratethe pervasivenessof migrant labor in the South Africa economy. In 1982, almost 60% of the black workers in South Africa were non-resident,or working in urban South Africa only temporarily. Lacking important manufacturing, mining and agricultural industries, the homelands have become primarily transfer and service economies-labor reserves for the rest of the economy. Social discrepanciesexist betweenhouseholds with and without migrant earnings, with and without land, and with and without pensions, which have led to a steady worseningof incomedistribution within the homelands(Simkins, 1984).

3.60 Demographic studies indicate that the rural areas outside the homelands (i.e., the white commercialfarming areas) were the main sourceof out-migrationin the 1980s.34Net out-migrationfrom white rural areas from 1980 to 1985 was 1.6 million Crable3.9), while net in-migration into the metropolitanareas for the sametime period wasonly around900,000. Metropolitanin-migration includes those from homelandsand white commercialfarming areas. Thus, the homelands'population increased as a result of natural populationgrowth as well as in-migrationfrom white commercialfarming areas. A significantportion of the out-migrationfrom white rural areas resulted from apartheidpolicies. For example, 180,000people scatteredthroughout the OrangeFree Stateand classifiedas 'unemployed' wete sent to Botshabelo-an artificially-createdsettlement with no economicbase now housingthree quarters of a million people.

> SimkinsC., 1990,eimates the black ruralpopuation outside the homlads as 3.8 millionin 1983,of which 3.4 mmion rodded on whitefarm.

hapter3: Evoalon of the CorntmporaryAgrrin Stct Page 59 S&Wh,AfW Ay,yrk&wre:StructIre. Perfonrance& Sons for te Future

Table 3.9: Black Out-Miration from the White Rura1Areas. 1980-198S

Net Outriigration Perentage Drop (persons) in Populton 198s8S (%) Orange Free State 220,000 2.4 Eastern Cape 110,000 4.1 Natal 530,000 8.9 Eastern Transvaal 400,000 5.4 NorthemnTranavai 210,000 7.5 Othe 140,000 - Total 1,610,000

Soume: Simkins,C., 1990, derived from an analysisof South AfricanStistics, 1986.

Conclusions

3.61 Two important lessons emerge from this chapter regarding the evolutionof the contemporary agarian structure in SouthAfrica. The first is that some Africanfarmers either on their own land or as tenants on manorial European settler estates were able to successfullyrespond to the needs of new and rapidly growingmarkets whicharose aroundindustrial mining and urban populationcenters in the latter half of the 19thcenury. Not only did commercialAfrican farmersparticipate in certainproduct markets, some also effectivelycompeted against large settler farming based on hired labor. This rise of African commercial farming took place under conditions of relative land abundance, weak and ineffective government interventions,and relativelyundistorted markets. Under those conditions,some African frmers accumulateda significantstock of capital, wealth, farming skldlsand experience.

3.62 The secondlesson is thatthe developmentand prospects of Africanfarming and rural development were crippledby a long list of governmentpolicies, including,but not restrictedto, creatingan artificial land shortagefor Afiican farmers;proscribing their participationin the sales andrental markets;confining thek access to land to the reserves; excludingthem from credit markets; blocking them from output markets; denying them access to marketingcooperatives and farmers' unions; refusing them extension servicesand accessto publicsector investments;subjecting them to rigid and authoritarianstate and state- made communalland tenure systems;forcing them onto the labor marketvia extortionarytaxes and levies; and forcibly resetting millions to densely populated reserves (homelands). The decline of African firming implieda gradualloss of agriculturalcapital, wealth, and farmingskills and experience. Farming ceased to be a window of entrepreneurialopportunity and managerialand technical developmentfor A*icans.

Captr 3: Dwltin of he Contenwy AgrarianStraure Page 60 SouthAfan Aericubure:Srutre. Peormw,e A Okton for theFute

CHAPTER4: MARKETINGAND PICING

l l troduct ion

4.1 More than 75% of agriculturalproducts in South Africa are sold under controlled marketing schemesset up under the MarketingAct of 19683 or related legislation. In 1990, poultry, meat, fresh vegetables,and fresh fruit for sale on the domesticmarket were among23% that was uncontrolled. A complicatedregulatory frameworkwas created mainly, it appears, to protect commercialfarmers from domesticand internationalcompetition. The22 controlboards and other marketinginstitutions that were established-with 15 still in existence-use a wide variety of control mechanisms,including:

productionquotas; * quality restrictions; * price cotrols; * monopolymarketing channels; * restrictive licensing; * import quotas; * producer price subsl cs for outputs and inputs (e.g., pan-territorialpricing); and * consumerprice subsid;esfor major staple crops.

4.2 In the commercialfarm sector, producersplay a dominantrole in the institutionsthat control and implementthe various marketingschemes for controlledcommodities. The marketingarrangements in the homelandsare more varied and complicatedthan in the commercialfarm sector, but the small-scale producers have very limited roles in the insfitutionsthat exercise control.

4.3 Significanteconomic costs have resulted from the distordons introducedin the marketg and pricing environmentin South Africa, which have:

* generallybenefitted producers and harmed consumers; * been partly responsible for the high degree of concentrationfound in marketg and processingwhich may harm both producersand consumers; * harmed the pooreststrata of SouthAfrica's populationin particular,because there is only a small subsistencesector in the economy with the result that the homelandsare net- importers of food from the commercialsector; * generally not benefitted small-scalehomeland producers, although informal marketng channelsmay have compensatedfor this somewhat;and * prevented the developmentof a more employment-intensivemarketing and processing structure.

4.4 Since the early 1980s the governmenthas attemptedto reduce the excessiveand costly public

IssT ogna prmulgto of this Ad in 19 was closly red to ote klslon, such a the Co-ptive Sodat Aos, the LAW Am an t LAdWard Agrca ank Act, as shownin Chers 1 and 3.

Chaepr 4.: AMar*eft and P81cbi Page 61 SouthA*ka Aa ge: Str . Prfnnance & Ontions or t*e FuturW support for th curr mark arramgements;as a result there has been a considerableamount of deegulation: severa control schemeswere abolished;some domestic quatiy and licensingrestictions were lifted; certain importquotas were replacedby tariffs; price conrols on milk, buter, cheese, bread, flour and maize meal were lifted; bread subsidieswere discontinued;direct producer price support for maize was witbdrawn cost-pluspricing for several major commodities(e.g., maize) was replaced by more market-orientedpricing mechanisms;and certainrestrictions placed on cooperativemarketing were liftd.

4.S The impactof these refoms has been uneven and awaitsfurther analysis. Agricultr producer prices of most commoditieshave fallenin real terms, yet consumerprices have increased. The resulting increae in marketingmargins poses somewhatof an enigma,caused in part by input cost inflation-the result of deregulatedoutput marketswitout concurrentderegulation in input,processing, and distribution markets. Other causes includedeclining productivity in the food chain, the withdrawalof consumerand producerprice ontrol and supports, and relativelyslow increasesin competitionin degulted industries becase of their high levels of concentration. Moreover,the bulk of farm produce is stil marketedunder controlledsystems Ihatregulate and reduce domesticcompetiion and protect domesticproducers from international compe Tion.The inefficienciesof the system are exposed not only by the increasing marketingmargins of the officialsysnem, but also by the rise in parallel-marketacdvities-a phenomenon which is relativelynew to SouthAfrican agricultura markets.

Overviewof Maor Madnlchemes

4.6 The Marketing Act6 is an enabling act that applies to all products listed in its schedule. it provides for the promulgationof subordinate legislationcalled 'schemes, which are insfttutedfor a product or group of products; and a Conrol Board is set up to administereach. The scheme empowers the relevant Board to carry out normal administrativefunctions, as well as specificfunctions which may Influencethe marketrin of te product to which it applies. These specific functions are derived from sections52 to 80 of the Act, whih state that a Boardmay exercisesuch powers with the approval of the Minis. Throughenabling clauses in the Marketng Act, the Boardsare fundedby levies on the produce handled. Despitethe wide range of authoritiesprvided in the Act, most schemesuse only a few powers which allow:

a. buying a product at an approvedprice (surplusremoval schemes, section 56 of the Act);

b. prohibiting the sale of produce except through the relevant Board (single channel schemesF,section 64);

c. fixing or infl ing the price of produce (section60);

d. contr the import or export of products; and

U6 &Thad is basudca seion 13 of theKaer Rpt (1993).

04par 4: Mo*hsg and Prcn Pae 2 Sosth Afican Afgcubure: Sr cu. Prfmnae & OQMbsfor the Future e. imposingor usinglevies and specia levies(sections 41 to 46)i7.

4.7 'he Marking Actprovides for fou maintypes of swhemes,taiored to the specific cr of the commoditycovered. These include:

a. single-chanmefixed price schemes. Producersare legallyobliged to markettheir productsthrough the Boardor its appointedagents, and pricesare fixedfor eachseason. Majordomestic crops such as maizeand winter grains (wheat, barley and oats)fall Into this category. b. sndePel Nr scheme Pducers markettheir productsthrough a poolconducted by the Boardswho pay advancepayments on receiptof the product. Deferredpaymeats are madewhen the final realizationof the pool, after deductionof pool expenses,is known. Cropsfing relativelyelastic demand, e.g., exportcrops, fall in this category (Table4.1). C. sulus val (also calledprice supportschemes). Producerssell their produceon an openmarket, although often subjectto stringentquality controls which couldrestict domesticcompetition. MTe Board intervenes when prices drop belowa fixedminimum price by purchasingsurplus for distributionand resaleat a later date (Table4.1). A dairyscheme was legally suspended at the end of 1992. d. mMrvisor schemews.The Boardacts in a supervisorycapacity and as a mediatorin arrangingprice and purchaseconutraCS between producers and buyers. Productsinclude caing fruit and cotton. 4.8 In 1990191,some two thirdsof thegross valueof agriculturalproduction was marketed under the Marketg Act. In addition,products such as wine(the Ko-operatieweWijnbouwers Vereniging Act); sugar (SugarAct); and lucemehay, wattle bark, and ostrichproducts (the Co-operativesAct) are reguatedby otherlegion, makingup some10% of the valueof farmproducts. Uncontrolled products includemost fresh vegetables(4.6%), fresh fuit for sale on the domesticmarket (1.8%), and poutry meat (14%of grossvalue of agriulural production).In addition,there existsan am of marketing arrangementsin the vaious homelands.Table 4.1 summarizesthe coverageof schemesby typeas of the late1980s.

n' Underto Maktg Ad, th SAAUca impos a levy on catin martd ouWia,ea on unwotolle outputs. his ha _gmei ld to do anomalyth blackfaus, by payingSAAU fee on unotoled product,wem effecovdy finain di SAAU.

auVtr 4: MaJ&*n ad Prci Page 63 Sgt Afin Aerkcturu: Src. egnmc A Oons hr the FWure

Table4.1: Maegn schemes:Pctage comnositionof the gosvalue of arnicutl wducts - l9!

Paecnt of Schcne/Product Vale S'meChmd Fbe re Maize 12.87 Winter Cerls 4.54 Shnle Cbannel Pool Olseed 3.09 Leaf Tobacco 1.86 Chicory 0.04 Buckwheat 0.0 Luoune Seed 0.04 Deciduousfut 5.04 Ciu fnut 2.33 Dded fruit (ionding nuts) 0.49 Banana 0.65 Roolbostea 0.02 wool 3.13 Mobair 0.39 SurDlus disoosal schemes Grain sorghum 0.33 Dry beans 0.71 Potatoes 2.U8 Slaugheed stock 16.S8 Eggs 3.71 Milk and butter fat 7.18 MMK= a KIMnhag Cannng apricob, pac and pns 0.39 Coton 0.S Contr in terms of seetion 34A Ka"au peb Tf TAL 66.83

1ov In {rms of Gumhef n Sugarcane 4.60 Lucrn hay 1.88 Wattlebark 0.22 Wine 2.99 Ostrichp Q4

TOTAL 10.14

Pxehbvagetaes 4.S6 Powls dm*Mtcd12.33 Sub-ropical fruit (excling bananas) 0.41 Odter uncontroled products S.73

C_ete 4: M*t&g ad hsg Page 64 Suh A*c Anricwdue:Stwulwe. Perfonanwe& OCado for Ok Futre

Ng: 1g follwing whemns wor recently diaoved. su&irhann pool schems-banana, chiory, rools tea aud wool; srphus disposalschema-dried beans, ega, and potoes. Woola milk axecurrnly controled under uuvio schm.

Somc: Ada_td from Abstrwt of Agricul Ssts (1992)

Role of Farmer Cooeraives 4.9 Ihe controlledmarketing environment is generallydominat by producerorganizatons. Farmer cooperaivesplay a major role in the marketingof bot crops and livestockand in the supplyof agricul inputs. Thereare three majortypes of cooperativesin the commercialagrculurA sector: a Regionalor geographicalcooperatives-locally based organizaions that undertakeinput supplyand general commodity haidlng and alsoprovide credit to their members,using the LandBank as their majorsource of fiuds. b. Specializedcooperatives, formed around the markedtgof a singlecommodity, such as wool, citrus, deciduousfruit, or tobacco.

C. Centralcooperatives, combining several cooperativeson eithera regionalor a commodity basis, to pool biying power for agriculturalinputs or to provide central markaetg for commodities such as meat, citrus etc.

4.10 Ihe Markn Act confers effective monopolypowers to the farmer cooperatives. They are appointedas sole agents of the control boards for marketingwinter grains (wheatand barley) under td e4ann_ fixed price shemes. They enjoy similar monopoly privileges for deciduous fruit Tnifireco),citrus (the Citrus exchange), and tobacco. The Co-operativesAct, the Wine and Spirit ControlAct, and other legisation also confer monopolypowers in the marketng of commoditiessuch as ostrich products and wine.

4.11 The posio of the cooperatives in the marketng chain is further strengthened by thei rpresentation in the South Afiican Agricturad Union, which advises the Minister on producer ereeaton on all the Control Boards. This has resulted in substandalcross-representation on the boards of cooperatives,Control Boards, and in some instancesprocessing and distributioncompanies.

4.12 There have been some recent legislative reforms to increase the economic flexibility of the coopeates, which are financiallystrpped. For instance, the volumeof business a cooperativecan undertakewith non-memberswas increasedfrom S to 49% of total businessvolume and vodng power has been tied more closely to volume of the individual'sbusiness with the cooperative.

Performane of the Varo Maketnng Schemes

4.13 The potenti for competitionfrom small-scalefarmers, traders, and internationalmarkets has alwaysbeen a significantconcern for SouthAfrican agriculture. Monopoliesin the marketng systems act as fornidable barriers against new participantsin agriculturalmarkets. Arguably, they assist many

Oaq*m 4: Marketbran Pricing ?q* ES SOh 4fficangAiwuw Strutue. Performance& Oon for sheFuture paricipating farmers, often preventg bankruptcyand an exit from the market. The pages ta follow describ the marketingsystems for the most importantfarm commoditiesand assessthe effects of recent policy reforms.

Sh2e Cane FixedPrice Schemes

4.14 Maiz. Maize occupies an importantposition in SouthAfrican agriculture. During the 198os over 40% of the area under cultivationwas plantedto maize, which constitutedsome 75% of total grain production, 64% of the value of field crops, 56% of human grain consumption,and by far the bulk of animal feed (Groenewald,1989). Reductionof implicitand explicit producerprice subsidiesduring the policy reformsof the 1980s resulted in a steady declineof area plantedto maize. Area planted declned by about 1.5% per year through the 1980sand yields are highly erratic. Maizeproduction Is particularly vulnerableto drought, as can be illustated by the 1992 crop year when the total crop fell to 2.2 million tons from 6.8 million tons in the previous year. Current estimatesof the 1993 crop are in the order of 8.1 milion tons.

4.15 The Maie Board maintainsa pan-territorialor unitary pricing system at the producer level for Is purchasesof maize, whichare largelyundertaken by primarycooperatives in the productionarea. The :ooperativeshave over many years establisheda comprehensivenetwork of physical silo facilitiesin the commercialfarming areas. Over 98% of the purchasesare in bulk form.

4.16 Ihe general impact of the maize marketng system has been to raise domestic producer and consumerprice relative to the world markets. Wright (1992)estimates the nominalprotection coefficient for yellow maize (the major traded maize commodity)at between 1.2 and 2.8 during the period 1986 to 1990. The reslting rise in domesticproduction was overshadowedby the welfare losses for consumers and Maize Board revenues and foreign exchangeearinp in excess of RSOOmillion per year on both white and yellowmaize. These losses exceedby a wide marginthe associatedproducer welfare gainson the order of R380 millionper year, implyinga subsmta l deadweightloss. Table 4.2 summarizesthe esated transfersbetween maize consumersand producersfor the period 1986-90.

Table4.2: WelfrVeEffect of Pfice Distoions for South Afbin MaimeProdon (ilion Rand)

Year Wefare Gainin Wear Gain in Maine Poduction Maiz Consumptin White Yellow White Ydlow 1986 255 263 -358 -310 1987 564 507 -S73 -522 1988 634 479 -613 -597 1989 564 239 -517 -549 1990 206 178 -554 -595

£guze:Wrdght 1992

'q*W 4: Marktng cd Prcig Page 66 SoiahAfro Ag rcJ SmtWe. Perfomance& O for theFuwure

4.17 Ihe system has also strong adverse incomedistriution effects for producers, since only 20% of the 15,000 registed commercialproducers accountfor 80% of deliveries.

Tabl 4.3: Domesto Pricesamd Exuort Realiation for White and YellowMaizm. 1986 to 1990 (Randper ton) Year ProducerPdie Epor Coswm Price White Yellow Realization Yellow 1986 240.4 225.3 187.8 285.3 1987 258.0 246.0 168.3 288.0 1988 260.0 255.0 326.8 295.0 1989 264.0 259.0 327.6 333.0 1990 302.7 302.7 317.0 360.0 1991 357.6 357.6 356.2 419.0 1992 455.0 455.0 475.0

Sorepartmetm of Agriculure,1993

4.18 Maize is producedprimarily for the domesticmarket. White maize is an importat staple in the South African diet, and is not generallyavailable elsewhere in the world for import Before 1987, the producer price was frequentlyset above exportparity, geneaing exportablesurpluses which needed to be sold at a loss. Consequently,export volumeshave been highly erratic. By the late 1980sthe Maize Board had built up a substantialdeficit on the export stabilizationfund.

4.19 The system was retrmed beginningin May 1987. Before 1987, the Ministerof Agrculture set the producer prices of maize. As of 1987/88the Mnister made the Board responsiblefor settig Is maize buyig price taking into accountthe long-terminterests of the industry. In practice, pre-planting maizeprices are made known to the producer. Pan-ritorial pricing is still followed,but the board no longer has the power to carry over surplusesor losses arising from exports. The Boardcannot make use of loansto financea particular marketingyear. Since 1987, then, the producerprice has essentiallybeen operated as a pooled price based on actual performances.

4.20 The principal effectof the pricing reformsof 1987has been to compelthe maize industryto meet the costs of losses arising in any season within the cost and price structure for that season. In other words, there is now viruly no scopefor inter-yearprice stabilizationthrough reserve stockmanagement. Given the requirement to balance the books each year, the producer pricing mechanism is clearly influencedby the proceduresused for dermining the Board's sellingprices. The Boardfaces a relatively inelasticdomestic demand for white maizefor human consumption,but a higbly elasticdomestic demand for yellow maize as a feedstock with a range of substitutes. The Board uses its monopolypower in the domesticmarket to charge 'what the market is able to absorb- in the long-terminterests of producersof maize. In practice this has meant relativelyrapid price increases for white maize comparedto yellow. In the period from 1986 to 1991 the white/yellowBoard sellingprice ratio moved from parity to 1.1.

aqw 4: Marketi andPrcing Page67 Table 4.4: TheIAremate Measureof &Sxnmt(AMS) for Maize in South Afric

Daw*tFtion Uok 1985t86 1986187 19M7M88 1968/89 1989/90 1990/91 I. LeIve of Ptuducion Too 8,295,325 8,520,814 753,361 7,014,264 12,035,253 8,708,774 In. Ptoducer Price Rond 271 275 290 298 301 352 I Ill. V"a of Produaion Rand 2,244,466,085 2,288,140,642 2,132,254,069 2,09302,812 3,622,972,211 3,066,794,764 [v. Diret ftP*ymtwa Rad 366,2M6,691 221,569,362 333,1J0,955 171,307,932 176,54,S75 91,225,266 V. JuatodValueof 4Rand 2,610,712,766 2.59,710,50, 2,46S,3S5,044 2,262,610,744 3,795,520,785 3,t58,020,030

VI. Pn_cy huidh A. Me"t Pric Suppot Read -21,117,371 618,310,696 494,938,3S -459,991,897 -348,780,798 -299,849,701 (WoiddReaf csFPc) Rond 273 201 223 364 330 387 S. DkectPaymcst Rand 366,246,681 221,509,862 333,100,955 171,307,932 172,548,75 91,255,266 C. RFaduc ofo lapot Coo Pad 10,011,912 10,372,679 68,461,382 47,891,690 56,126.333 52.779,931 D. Geasal Seviom Raod 104,68,630 10O,001,925 114,167,109 141,019,011 117,391,005 110,204,078 VI. Total AIS Rand 459,749,90M 950,276,162 1,010,668,482 -99,773,26 -2,714,886 -45,710,926 VII. Unk AMdS RUt 55.42 114.20 137.44 -14.22 4.23 -S.25 IC. Phresatagsc A IS b 17.61 37.86 40.99 4.41 -0.07 -1.45

NM: preduce priX= nt producer pricep eeoied plus macatling coas ad sdminrza" cosa of thdoMain Bosed. direct paymuwrs governfu- paymeaa toamao produocosezaoqt (IV) and (VI ) for loan ad awerosubsdies. mact rimes ppoe difflene b*_ dogma-I producer prie and wodd som pdie muldlicd by tota pr_etWo. rduo in 0upotCOIN: g_aumt edredics on UaQxpt, bmt , aN Ad. gIal seic: _AMof bodb of D _pamaof Appmr bau.ig maimepmdWa.

S : van Motr4a and va Zyl, 1993.

a Sowh AfrkcanAserkcl: Sb,~*. Padarmw & Oatons w dshFu tu

4.21 he time series of maize prices at th produc level and the Maize Board's soilingprice are set out in Table4.4. In real terms, after remainingfairly steadyfor 15 years between 1970and 1985, maize producer prices fell by one third in the latter 1980s. This coincided with the eliminaion of explicit subsidiesfrom the governmentbudget to the MaizeBoard and the emergenceof a substi positivereal markting margin in the Maize Board's price structure.

4.22 Recen trends in the Aggregae Measureof Support(AMS) 3' (Table4.4) Mlustratesthe decine In transfers from conumes and tapayers to produs. The decline is the result of the combinedeffect of a reduction in direct paymen by the governmentto producers (excludingsubsidies on loans and interest)and relative inmcreaseof te world referenceprices as estimed by the authors. Althoughthese results differ from Wright (1992) because of methodologicaldifferences, it can be concluded that transfers from consumersto producershave declined in recent years. This analysisof van Heerden and van Zyl (1992)shows the shap decreasein the AMSsince the dhangein maize marketig in 1987/889-he AMSdeclined from R 137.44/t (40.99%) in 1987/88to R -14.41/t (-4.41%) in 1988/89.

4.23 With the widening of the price gap between the Board's buying and selling prices the single- chanel marketng systemhas comeunder prese. In recent yearsthe Boardhas permittedan increasing level of direct transactions,and modificationshave been made to the levy structre to accommodatethese changes. As it stands,the systemprovides a major incentiveto use maizeon-farm as a feedstDckrather tan sellingit to the Board and incurringthe levies. Likewiseit has pwvided an incenive for large users of yellow maize as feedstockto invest in maize productione'.

4.24 The shift in conol instumes used in the maizemarketing system has already affectedfarmers in a number of ways. In the absence of safety nets or other exit options, many producers, especially marginal production areas, found themselvesin fiancial difficulties(see Chapters 2 and 5), despite a notceble shift to low-inputproduction systems. The expectadonis that productionwill generallyshift eastwardinto higher-potentialareas and that this trend wiMbe strengthenedas the uniay pricing system comes under pressure if cmrent GAIT negotiations are successful. A more open trade regime in SouthernAfrica and the rest of the world will induce further changesto regionalpatterns of production. The lack of efective protection against dumping and the high degree of concentrationin input supply, processing, and disbutio industries will also need to be addressed.

4.25 Wheat and other winter cereals. Wheat is grown in two major areas of the Republic - in the Western Cape (the Southernand Western areas, and to a lesser extent in the north-eastemdistricts); sad in the eastern and westn parts of the OrangeFree State. Total productionof wheat under irgation in the past decadehas varied from a low of 1.4 million MT in 1980/81to 3.5 million MT in 1988/89. The Cape Province growing areas produce an averageof some 850,000 MT, or 40% more ta the region requires. Surplusesare sent overland to the main milling centers in the OFS and PWV areas. Other

1 MmeAggregte Meaur of Suppor (AMS is basd oathe prodce subsidyequValai, nd imm=suremtasfrfron consum aduayer to peodus.

" it is rpoted tha th arge conneici prod of izc is the leadingpoxutry producer. Aordn to t ameds inomal eates on-fm tatiom ar ofthe orderof 0,8 to 1,0 millio andstM rsng Givo theincentives to avoidte leviesit is likly thatthe offici al tkaticundetavte the tue vohuneof maiz productioin the country.

Chaptur4: Markathagand h*fag Page a Som* Afria Ar&uhtr: StnWu. PE9*=W A ODIOW fOr th FWW* - wintercereals such as barleyand oat also fail underthe WheatBoard.

4.26 lhe Board'sallocation system aims to meetthe ualitY equmes of the millers,with the resolt that eachmiller receves wheatfm at least fivedifferent areas. Undertis systemthe into-millcost of wheatgrain is approximaelyequalized for al wheatmillers, irspective of location,by selectionof the sourcesfor eachmill. Millersbear the actal rail costfrom the supplyingsilo, but the allocationsystem aimsto equaizethe averagecost for each miller. Capeproducers benefit from the allocationsystem operatedby the Board,as a unitaryprice systemis followed. 4.27 Wheatimports are purchsed by the Boardon behalfof the governmentand sold at the domestic sellingprice. Importsof wheatare restrictedthrough quantiatve controls-possible because the importer is treatedas a producerfor purposesof controlof intemalmarketi. Importedwheat must match the qualitiesavailable domestically. 7Tis limits the potenal sourcesof imports,for eample catng a bias againstArgentine and in favorof Australianor Canadianwheat. 4.28 The wheat milling industryis significantlyconcentated, with six major operators. Thi concentraionis, moreover,more marked at the regionallevel witi onlytwo or threeoperators competing in severalparts of the country(two operatorsin WesternCape; three in Durbanand EastermCape, whereasaU six operatein the PWVarea). Thnenumber of bakerieshas increasedsubstatially since the deregulationmeasures of March 1991. 4.29 Themarketing system for wheathas createdsubstantia inefflciencies. The relatively small role of extr trade haspermitted the Boardto setproducer prices at levelssubstantily aboveworld prices *Cable4.5) withlocal prices between 11 and46% in excessof meanworld prices in the period1980 to 1987. On the otherhand, the Boardwas able to protectdomestic consumers from the sharprises in world pricesin 1984and 1985.

Chatr 4: Ma*ring ant P*S Page 0 Soqth Afrkan Auribur: SKr. Pfmance & Oa for she Fore

Table 4.5: Gross WheatPtoducer Prices and MeanWorld Ptices

Year Gross Psoduce World Market Price ProducerWorld Price (Grade DI) (US No 2 Dark) Ratio 1980 208.7 180.2 1.16 1981 234.2 189.6 1.23 1982 286.8 196.1 1.46 1983 266.8 241.4 1.11 1984 290.0 32S.3 0.89 1985 315.3 376.0 0.83 1986 364.2 292.4 1.25 1987 391.6 302.0 1.30 1988 341.2 352.5 0.97 1989 368.4 368.3 1.00 1990 493.3 342.7 1.44 991 591.5 363.5 1.63 i992 692.2 394.4 1.76

S_urc: Grocewald, 1989; Depatiunct of Agdrculure, 1992; Abstrac of Agricutul Staitis, 1993.

4.30 Nieuwoudt(1983) estimated that the consumerand producer share of the bread subsidy was 71%S and 29%, respectivelyover the period 1948-1981. With his estimatesof demand and supplyelasicities, he calculatedtha if the 1980/81subsidy of R162 million on bread were terminated,the consumerprice (m the absenceof imports)would increaseby 279%,leading to a fall in the consumptionof bread by 5.2% or 88,000 tons. However, since in reality wheat could be imported at a lower cost than the domestic producer price, he concludedthat South Africanwheat producersgained more and consumersless from the subsidy than these estimateswould indicate.

4.31 More recent estimatesby the Departmentof Agricultureshow that prices for wheat ontinueto be stanily above world marketprices. In 1991/92wheat could have been importedat R434per ton (CEFDurban from Argentina),and sold in the PWV area for R509. However, the minimumselling price for domesticallyproduced wheat was R625 at the start of the season-23% above import parity-and increasedby R8 per month during the remainderof the season-to around 40% above import parity by the end of the season. Since imported wheat is also of higher quality and the selling price for a comparablegrade was R744, consumerseffectively paid 46% over importparity at the start of the season.

4.32 A significantderegulation of the wheat milling and baking industriestook place in March 1991, resulting in substantal increasesin consumerprices. Under these reformsprices of mflledproducts and their derivativeswere decontrolled. At the same time, after over four decades, the subsidy on bread was finallyremoved. Maximumretail prices were then R1.20per white loaf (subsidygranted: R3.27 per loaf) and Rl.05 per brown and whole wheat loaf (subsidiesgranted: R7.88 and R5.52, respectively). White bread (but not brown) became subject to domesticVAT (at 10%) in September1991. From that date, the Wheat Board, accordingto officials, ceased to use its controls on the registration of bakeries in a

Chqpter4: Marketg and Prcng Ave 71 Sxoh A*fcc Ankcb: St o,ture.Peface At Ot Atd Fswe

restrictive way and now reqires reistration only for satcal purposes. Followingderegulation, the price of a loaf of white bread incresed by nearly 50 % between November 1991 and November199, with neary a third of the price increase attibutable to an increase in baker's margins. The expected effect-an increasein compettion at the bakery level whic& ouldlead to a reductionin consumerprices- is only recendy beglnningto be flt.

4.33 The prospects for area expansion in wheat producdon in South Africa are limited, with the possible exceptio of production under Iigaion. Ihis is generally also tre for the rest of the subconient, which, lflie SouthAfrica, faces a growingdemand for wheat productsas incomesincrease. Policy Isuvmen to managethis emergingdeficit in wheat productionwill have to be appliedwith extra care in the future. The unitary price systemfor wheat is also expectedto be dismantledwithin the next few years both for efficiencyreasons andto conformto GATr requirementson domesticsubsidy systems. This wil place pressure on wheatproducers in the Cape and some shift in the locus of productionout of Cape province, with an expansion in the OFS growing areas, can be expected. In addition, further liberalizationof the exchangerate may make locally produced wheat substantialy more competitive relative to impous.

leschannel Pool Schemes

4.34 This section will focus primarly on deciduousfruits, given their relative importancein tms of output and eport volume. Citus maetng, also an important export sector, is organizedsimilarly.

4.35 Deciduousmfit. Deciduousfruts-principally apples,table grapes, pears, peaches,and apricots- are grown in the Westen Cape, though peauchesare also grown in the OFS and Transvaal. Twothirds of the crop is for export Oarely to Europe) and generates85% of the revenues. Deciduousfiuit, along with citrS and wool, count among the farm commoditiesin which South Afrca has a clear natural comparativeadvantage ftr intenational trade. The sub-sector is supported by a highly developed productionand markeing system Consequently,recent performnnce and futureprowects of theseewport are good and have improved with the lifting of internationalsanctions and the devaluationof the Rand.

4.36 A one-channelpool schemeis administeredfor deciduousfrut on a weeldyandlor cultivar basis. Advancepayments at approvedprices are made to producers; and once all the frit has been sold and exportingcosts have been deducted,final paymentsare made. No price control is applicableon exports; products are sold on the world market.

4.37 Export marketing of deciduous frut has been a private monopoly of the Universal Frutrade (Cooperative)Ltd. (Unifruco),which in 1988took over many of the functions of the DeciduousFruit Board (DFB)I. Ihe Board compriseseigbt members, aU of whom are represeves of producersof deciduousfruit. Historicaly, andby comparisonwith other MarketingBoards, there has been littledirect interventionby Governmentin the DFB, whichhas alwayshad a strong commercialorientation. Entry

TIe rqilat ofthe pblic mnWolyof DFB(it contine noinay in atnoe for lqal sons,although with ordy Oneemployee) by Unifrucooccrd in pat to Allowmorm flle internal a an remuaneat andin pa to diace h muaetingof frit fiomthe South ican publesewor, whose organom we a patculr foalsof acton from he US

CA.pm 4: Mar*edgg and Pricing Page 72 SowthAfic Affkurs: *ruci. Perfomance A to*s FutumE to the export marketis reportedlyunrestricted for any grower able to achievethe high stadad of grading demanded(only Class I fruit is exported),as Unifrucois obliged to accept all fruit of this grade offered. The establishmentof Unifruco has W the effect that levies decreased, especially those coveaing administrtive expenses.

4.38 Unifrucosells in the export market on behalf of its members, deductingall market-relatedcosU and passingthe residualto members. It operateson a break-evenbasis, but makesa profit on non-market activites, principallya return on its fixed and liquidassets, and other trading. Is administrativeexpenses are fianced through a commissionwhich amountedto 2.5% of gross sales in 1990. In additionto its market-relatedservices to its members,Unifruco provides a first paymentto memberswhich it finances through borrowing on domesticand foreign markets.

4.39 Domesticmarketg of deciduousfruit is relativelyuncontrolled, although producers have to apply for a permit from the Board. Trade is subject to charges of 8.5% when sold in municipalmarkets and 1.5 to 2.0% when sold direcdy by the producer to the trade. There is also a grading system laid down by the DFB and now by Unifruco,that is enforcedthrough inspectors. Ihe extn of direct (or farm gate) sales, and therefore of total domestic consumptionis not filly known, as reporting depends on the producer, who bas some financi incenive not to do so correcdy.

4.40 Ihe merits of the single channelfor the export market are strongly defendedby the industry on the grounds that () there are economiesof scale in overseasmarketing and promotion; (iH)promotion of the Cape brand imagehas consolidateda major marketniche for SouthAfrican fruit; (ii)this has been reinforcedby the ability of Unifrucoto enforcerigorous qualitystandards; and Civ)Unifruco has been able to control its costs and marketingmargins. The latter have narrowedover time, in contrast to those for other commodities. The highly successfulrecord of this export industry so far, by comparisonwith its southern hemisphere competitors, suggests there is some basis for this view, although the KasLer Committeereported reservations about the high degree of cross representationbetween the boards of Unifrucoand the DFB, where the latter body gras export licenses to the differentmarkets served.

4.41 There are also opporunity costs inherentin restricting accessto export markets by settng very high quality stndards. Whereas South African fruits currently sell at a quality premium in European marketsover their competitors(e.g., Chile), the impositionof such high qualitystandards prohibits entry into the marketby producerswho would want to marketa lower qualityproduct and are willingto accept a lower price. Such opportunitycosts will increaseif a more varied range of producersemerges in the context of a less distorted agriculturaleconomy and a new South Africa.

4.42 An interesting case study in the current process of deregulationof the agricultura marketig system is presentedby the dissolutionof the Banana Boardin early 1993. In this case, the assets of the Board were transferred to a new cooperative,Subtropico. The existingproducer interest in these assets was convertedto a transferableshareholding, while shares are held in reserve for new producers. A further case in point is the considerablepressure from certan lobby groups to deregulate the wool marketingsystem, as extensivelyreported in the agriculturalpress.

Chape 4: Marketn and Pri8 Paeg 73 Sn*ahAnican Airkcufri: Sa iure. Permnance A Orlion fo the Fui

SkuMW-reovl Schemes 4.43 kkg. Livestockand related products constate over45 % of agriculturaloutput and are produced throughoutSouth Africa. The nationalcattle herd amountsto over 10 millionhead (includingTBVC areas)plus 1 millionpigs, 30 millionsheep and 2.5 milliongoats. Table4.6 summarizesthe levelof slaughteringin 1989/90.Apart from cyclical changes, the cattlepopulation has beenbroadly stable since the 1960s,while the ammaloff-take has increasedfrom around 1.5 million head in 1960to over2 million headin the 1980s.

4.44 The Meat Schemerelates to all producersand importersof slaughteranimals and all persons dealingin the trade withslaughter animals, meat, offal, hides,and skins. In timesof over-supply,as duringthe droughtin the early 1980s,the percentageof animalsoffered for salewhich are purchasedby the Meat Boardunder its surplus-removalmandate can reachsubstantial levels, e.g., 9.58% in 1983. However,the morerecent direct interventions have been minimal (Table 4.6).

Tab1e4.6: Price StabilizationMeasuros in Controled Amuas

Year Cate slaughtered Purmhas by Meat Paerct Board 1985 1,455,909 36,972 0.25 1986 1,377,333 1,3S7 0.10 1987 1,212,369 3,641 0.30 1988 1,094,729 1,087 0.10 1989 1,130,846 782 0.07 1990 1,346,645 2,922 0.22

Souro. Food Sotdies Group, 1992.

4.45 An extensiveand complicated control system was in operation to regulatethe marketin meatuntil early in 1993(Sunnyside Group, 1991). The marketwas segmented into controlledand uncontrolled areas, withthe formerincluding all the majorurban areas.41 In the conrolledareas mr it had to be slaughteredat Abacorabattoirs and ownership transferred at compulsoryauctions on the hc. ;. The flow of meatslaughtered in uncontrolledareas into controlledareas wasregulated by the Board. 4.46 The schemewas highlyinflexible and createdeffective barriers to entry. Meat priceshave recentlyincreased faster than those of other foods,despite a declinein producerprices as farmerssold off stockin the drought. Producerprices are erratic,and substantialseven-year cycles exist related to rainfall,supply and price expectations(Lubbe, 1990). The producer'sshare in the consumerprice decreasedfrom 62% in 1974to 48.5%in 1990.Per capitaconsumption of red meatdecreased from 40.6

4 In praudte the exatd of impncartaion of the contrl systan. fromwhich atisis on slghting derive i unomain. It is rqxwtd for cxampe, that there are substania flowsof lie aninds ino the black township espxecilyof sheep and goas. whichare sbughterd on unreisteed pmi. in additionthe eatt of on-farmsluhtering is not wal docunthed. Farnmes are pernifttd to slau&ghrup to 8 head of cattlepe monthwthount regsn, and it may be that extnsivensiy to localmuar*s prevals. espeialy in the Platdand. The domandstatists may. accordigyl. be sigificany underestimat& aupter 4: MarketIngand Pricg Page 74 Souh AfricanAwrksrd,: Sture. PenformanceA Ootionsfor theFuture kg in 1955/56to 25.78 kg in 1989/90(DIrectorate of AgriculturalTrends, 1990).

4.47 Ihe meat marketingand processingstructure is characterizedby a few large dominantfirms and the absence of free choicefor producers,traders, and consumers. For instance, in 1990, the top three fims appointd as abattoir agentsfor markeng held 84.04% of the market share of catle and 94.41% of the pig market. In the abattoirindustry Itself, one firm (ABACOR)held 64.95%42,while the top three fim held 88. 17% of the marketin 1989190. In the processingindustry, one firm had a 45.64% market share in 1988/89(Lubbe, 1991:256). Table 4.7 illustratesthe market share of the three major firms in each of the stages of marketingand processing. A commissioninto the price discoveryprocess in the red meat iustry (RSA, 1989) found that the three major firms were favored under existing restrictive reglations and that the systemof controls and trade barriers had tended to stimulateand protectverdcal and horizontal integration.

Table 4.7: ConmMton in tfe Livestockand Meat Industry (Perent of maket share)

Market Function Frm A Firm B Firm C Shares of the Big Throe Fums Combined Tanneries 24.8 8.0 46.4 79.2 Hidesand Skins 40.0 21.0 23.0 84.0 Reta nonontroled 4.3 1.8 6.9 13.0 Rtail Controled 2.4 2.6 8.0 13.0 WholesaleNon- 6.9 4.7 8.2 19.8 Contolled Wholesale 12.0 7.4 18.7 38.1 Controlled Pdmary demand 7.2 10.8 20.0 38.0 Beef Poessing mea 45.6 3.5 17.4 66.5 Aataoirs - Cattle 20.5 13.6 39.6 73.5 Agnt - Catt 50.6 18.5 15.0 84.1 Feedlots Capaiy 21.4 20.0 5.S 46.9 ource: Lubbe,1991

4.48 Nieuwoudt(1985) estmated the welfare costs of the quota scheme in the beef industry. Using a supply elasticity of beef at 0.8, he calculated that a 5% supply restriction from the calculated equilibriumposition would lead to a deadweightwelfare loss of approximatelyR 3.2 million. Mme decrease in supply would lead to a 6.5% increase In the retail price in controlled areas and a 6.5% decrease in uncontolled areas. Quota rents in the beef industry were estimated at R 81 mllion, or approximatelyR52 per animal. Ihe quota system was recentlyrepealed.

a Cumrray estted at 41%.

Chtr 4: Markeingand Prcn Page75 Soud i And"kure: &rawte. Peforr A onM brt*e nFture

4.49 The increasein the relative price of red meat with respectto white met (poultry)seenm directdy responsiblefor the seculardecline in red meat consumptionover the past 20 years (Annex2, Table 2. 11). While the overall level of meat consumptionhas remainedin the region of 40 kg/cap this total concels a fall in per capita red meat consumptionfrom 40 kg to 25 kg, while consumptionof poultry has risen frm under 2 kg to around 15 kg per capita. As a result, the poultry industry has witnessr' - 10-fold increasein the volume of productionover the past three decades.

4.50 Recenty (January 1993) reforms have been introducedaimed at liberalizingthe meat industry. The systemof control of the flow of animils from uncontrolledto controlledareas exercisedby the Meat Board was abolished. Each abattoir is now responsiblefor contolling the inflowof animalsto its pens; however, local authoritiesdecide on the hygiene standardsfor meat brought into their areas of contol. Giventhe inspectionfees and the condnuedimplementation of the existinghealth and hygieneregulations (which are arguably too stringent) the extent to which recent changes will increase competitionis questionable.

4.51 Grain sorghum. Before 1986 the markedng of grain sorghumwas administeredby the Maie Board as part of the summer grain scheme. After 1986, the autonomousGrain SorghumBoard has administereda sutplusremoval scheme. The adjustmentin the sector on bothproduction and consumption since this change has been dramatic. Grain sorghumis mainlyproduced by commerca farmers in the Transvaaland OrangeFree State. The area planteddeclined from 314,000hectares in 1986n to 135,000 hectaresin 1991/2,while the share of animalfeed consumptionin total consumptiondeclined from more han 67% to about 15% (Abstact, 1993). The share of human consumptionincreased constantly from the 1960sto 1980, then declinedthrough 1987 before rising to around 85%-or 200,000 tons-by 1992. Most human consumptionis in the form of sorghumbeer (56% of total consumptionand 78 of human consumptionin 1992). During the droughtin 1992, about 190,000tons of the total market requirm of 250,000 tons were imported.

4.52 The Board operatesa sutplus removalscheme. Becausethe Boardhas no physical control over the productionor marketingof grain sorghum,the producersare free to negotiatethe best prices for their products. The Board, however, announcesprices at which it is prepared to buy grain sorghumfrom producers who cannot obtain more profitable prces in the trade. These floor prices are normally announcedfor a fixed period, in practice when the major portion of the crop is delivered.

4.53 In pursuance of the intentionand spirit of the surplus-removalscheme, it Is the policy of the Boardto allow the marketingof grain sorghumto flow throughthe normaltrade and to purchaseonly the actual surplusunder its floor price scheme. To a great exteent,the Boardhas reachedthis objectivein is four years of existence,since its purchasesunder the floor price schemenever exceededaround 18% of total productiou.

Marketine Under SpecialLeislation

4.54 Sr. The main cane growing areas are in Natal and Kwazuu with smallerareas in the Eastern Transvaalscheduled to expandin the future with the erectionof two new dams. The cane sector includes both estates linkedto mills and individualgrowers. At presentthere are some41,000 growers cultivaing 378,000 hectares. These supply cane to 17 mills. Most cane is currentlycut by hand, and according

Chaptr 4: Marketingand Prcing Page 76 Sa* A_a AgdguMe: Sftwu. Pcfomuce & Oaions fiw the Fuure to Industry estmates some 141,000 people are employed. Ninety percen of the area under cane is rainded.

4.55 The productionand marketingof sugar cane, sugar and associatedproducts is reguated under the Sugar Act of 1936, which establishedthe South African Sugar Associationand granted the Industry statutowrypowers. The Act providesfor an agreementto regulatethe affairs of the industry, binding on aU those who grow sugar cane and produce sugar and associatedproducts. Ihe main proNJionsof the Sugar Industry Agreementcover quantitativecontrol over production by means of quota allocationsto growers, regulationof the supplyof cane to the mills, administrationof a two-pricepool system relating to domestic and export sales proceeds, cost responsibilityfor tie onof cane from fitms to mill, disposal of sugar eports, pooling of proceeds from sale of sugar and molassesand applicaton of the revemnesharing formula between growers and mils, de on of the Luc:rosecontent of cane, and impositionof levies to overthe costs of administrtion by SASA.

4.56 The SASA includes representationof the growers through the South African Cane Growers Asociation (SACGA)and the millersthrough SouthAfrican Sugar Millers Associaton Ltd (SASMAL). The two organizationsare equallyrepresented on the SASACouncil. The SACGArepresents all growers other than tose who also mill. Individualgrowers are representedon SACGAthrough their constituent growersorganizations. There are currently23 memberorganizations of growersin SACGA,representing some 41,000 individualgrowers.

4.57 The prime fimctions of the SACGA are to represent growers in SASA and to monitor the applicationof the revenue-sharingformula. Ihis involvesannual determination of cane productioncosts and coninuous monitoringof the growerlmillerdivision of proceeds. SASAalso takes responsibilityfor promotionalprograms and sugar industry research; it operates its own installations,including the Bulk Sugar Export Terminal, the Bagged Sugar Export Warehouse,the Bulk Sugar Depot In Germiston,the Experiment Staion, and the Industrial Traing Centre.

4.58 Overall welfare losses associatedwith the sugar scheme are estimatedin a simulationmodel by Ortmann(1986). He esfimatesthat under the single-pricesugar policy whichhad been in operationuntl April 1985, the welfarelosses were approximatelyR17.3 million. With the ensuingpolicy of a split-pool scheme in 1985/6these social costs were estimatedto have reducedto R11.5 million. Table 4.8 below summrizes the prices paid for sucrose (and cane equivalent)before and after the introductionof the two pool pricing system in 1985/86. Table 4.9 summaizes the final destinationand utilizationof cane and sugar for the period 1981182to 1990/91.

Chapr 4: Mwrk#g and Prcn Page 77 Sbnth Afltho AMw*cxbm:Svnactw. Perfoiance & OhtonsfIr thefuture

Tablo4.8: Sums. and Caneficos 1982to 1991 (Randper Ton)

Seasn Suaco cne 1982/83 172.33 22.16 1983/84 244.75 30.18 19W8S 192.64 23.64 APoo B FPOo A Pool B Fbol 1985/86 236.64 103.11 31.07 13.54 1986/87 298.90 162.06 38.26 20.74 1987188 299.68 132.97 3S.96 15.96 1988189 339.10 257.93 42.76 32S2 1989/90 387.32 3S8.92 51.01 47.27 1990/91 436.19 362.09 56.31 46.75 1991/92* 468.00 258.00 60.80 33.S5

MMo:the twopool pricing system ws Introducedin 1985/86. 50urc: SouthAfrican Sugar Associaion

Tabb 4.9: Tot CanelSunar Producton: 1981to 1990 (Minion Ml')

Sa Camn Suga Produced Crushed Total Naional Internaiomd Mast Masko 1981182 19.S 2.055 1.213 0.842 1982/83 19.3 2.125 1.181 0.944 1983/84 13.4 1.377 1.14S 0.232 1984J8S 22.4 2.690 1.307 1.062 1985/86 18.8 2.117 1.203 0.914 1986/87 18.3 2.013 1.269 0.744 1987/88 21.1 2.190 1.240 0.938 1988/89 19.9 2.168 1.348 0.819 1989/90 18.6 2.137 1.239 0.898 1990t91 18.1 2.038 1.314 0.713

ourcc:South Afican SugarAsoeiation

4.59 Befbre 1990 the quota system for the delivery of cane operated in favor of fte commercia growers. Anomaliesoccurred, such as the fact that small growers with low transport cost to mil were

0kpuer 4:, Marketin and Pricin PAve78 MON&AUa AgdcuL: Mnar. Pe awe A Omtk e, thg Ftue - xcludedin some aes fromdeliverang to their nearestmill, while established gtower at greaterditan had prefaenti accessto quotas. Thi was reinforcedby a transportsubsidy system which ceated a uniformtsport tariff that did not discriminaton the basis of dstance nd in some ices is supported cane haul distances of 70-90 kilometers. Under this system small growers nearer the mill subsidizedmore distan commerciagrowers. 4.60 Recentchanges in the systemfor regulatingsupply to the mMshave favored the expansionof the smaliholdersubsector. Ihese changesare leadingto a rapidexpansion of smalholderarea under cane, with the wmber of smal growersexpected to rise to 55,000 over the next5 years. Underthe new arrangements,introduced In Aprl 1990,smallholders (defned as thoseproducing less than 150tons of srose per two yeans)growing within 30 kilometersof a mIllrequire no quotafor mil deliveryand are guaranteedthe higherA pool price for their sucrose. Duringthe past decadethe expansionof the operationsof the Smal Growers Financia Aid Fund, which provides financialsupport for the establishmentand growingof caneby smallholdershas alsoprovided a stimulusto smallholderplaning. 4.61 Farmerrepresentation in SACGAhas traditionally been based on tonnageof caneproduced, with the effectthat commercial growers dominatd the SACGAboard. Recentlya newrepresentation structure was itrduced wherebyeach mill has a localgrower council with equalrepresentation of smalland commercialgrowers. Ihe Boardof Directorsof the SACGAhas been enlarged from 37 to 60 withequal smallholderand commercialgrower representation. 4.62 Wine.The bulkof SouthAfrica's wine production comes from the winterrainfl regionof the WesternCape, which is also a largeproducer of tablegrapes. The totalplanting of vinesis 99,700ha, moreta 80% of whichis plantedto winegrape varieties. Wmefarmers deliver their cropto wine cooperativesor direcdyto wholesalers,or theymake and bottle wine themselves (estate wineries). Until 1992wine was produced on a quot basis,with the quotagranted by the KWV,the controllingbody for the industry.This cooperatve sets miimum prices for good (table) wine and buys up sutplusproduction, mostlyfor dtilion. The directorsof the KWVare electedfrom the variousregions by farmer members.The KWV also providesadvertising, research and extensionservices to members.Production quotaswere suspended in 1992. 4.63 Guateed prices for low-qualitysrplus productionresulted in a incentivestructure that rewardedvolume at the expenseof quality.The gross volume of grapespressed has increased from 1,037 milliontons in 1981/82to 1,379 milliontons in 1991/92,and by value from R212,920,000to R750,000,000over the sameperiod. Thetotal production of winehas increasedfrom 772 millionliters in 1980/1to 970 millionliters in 1991/2. In the formeryear most wine produced was used for distilling purposes(418 million liters, comparedto 354 millionliters for tablewine). By 199112,however, this had changedto 516mfllion liters for tablewine and 457 millionliters for distilling,mostly as a resultof the liftingof sanctionsand the resultingincrease in exports. The effectsof the suspensionof quotasare not yet clear,but wouldprobably include both an expansionof thetotal area planted to vinesand a furter substiuon of the lowerquality high yieldingvarieties by noblecultivars.

Oahq' 4: MarkAlg and Pric Page 79 So"AAfricMlrcu : SIfucnure.Pe anceA OpbMfhr theFraur Concenution in the Marketingand Processing Structure

4.64 The extensiveregulation of marketingin SouthAfrica has contributed-alongwith the exchange contol regime,skewed income distribution, and other factors-to the concenation found in SouthAfrican marketig and processing(Sunnyside Group, 1991;Competition Board, 1992;Elliott, 1992;Groenewald, 1990; Lubbe, 1991; Wright, 1992). An assessmentof concentrationIn marketing and processing also needs to take into accountthe regional pattern of marketstructure. In several industries, such as wheat miling, the number of participantsat the nationallevel is moderate,but this hides effectiveconcentration in two or three companieswithin individualprovinces.

4.65 The growth of concentrationin South African agricultural marketing and processing can be attribued to a number of factors:

a. single-channelmarkedng systems;

b. economiesof scale in marketingand processing;

c. the absenceof effectiveand-tust legislation;

d. intenational competition;and

e. imperfectionsin financial capital markets.

4.66 The presence of single-channelmarketing systems for many crops and products has tended to promote concentrationin marketingand processing:

a in marketing, through the creation of statutory monopoliesor oligopolies,either in the form of the control boards themselves,or throughmonopoly or near-monopolyagency arrangements,such as those prevailingfor the cooperativesto marketcereal crops; and

b. in processing, through the economiesof scale created by the single-chnel marketing chemes. This has encouragedthe formationof concentratedpurchasing arrangements in the agro-processingindustries. In some cases, this has occurred through preferential applicationof control schemesand registrationrequirements, and in other cases through more purtly market mechanisms. In the livestockindustry, for example, concentration in the abattoir industry has been favored by the operationof the quota control system underthe meat scheme. The presenceof strong vested interestson the control boards has tended to lead to collusivearrangements favoring the emergenceof dominantcompanies.

4.67 A secondfactor generatingincreased concentration is the economiesof scale, generallycaptured by the larger firms operatingin certain sub-sectors,e.g., poultry production, where the companywith a dominantmarket share establishedextremely efficient, low-costproduction systems in the 1980s.

4.68 Imperfectfinancial capital markets have tended to increase concentrationthrough mergers and acquisitions,particularly during the 1980s. Ihese were often precipitatedby the threat of bankruptcy

Chapte 4: Markeing and Priig Page 80 SWArAfn Ao*iura: Slruvwr. Performa A Op*ow fr OmAgaru facing firms; the process was acceleratedby the concentratedcontrol over liquid funds exercisedby the lrger business groups. 'Te gove¢nmenthas in generalnot opposed acquisitionsand mergers wherethe alternativewas seen as bankruptcyfor the acquiredcompany. The level of merger activity was closely linked to overall macroeconomicconditions, notably the stagnationof real incomeswhich occurredover the past decade.

4.69 Internaional competitionhas also favoredconcentration in certain sub-sectors. Comparedto the major developedcountries, the South African market is somewhatsmall. As a result, firms which are large in relation to the local economy are in fact relatively small compared to their international competitors; this has created a problem in indusies with significanteconomies of scale, as well as a conflictbetween internationalcompetitiveness and domesticcompetition policy. In these industriesthere is understandablypressure for Government to permit "rationalization"of industries to improve internationalcompetitiveness. Ihese argumentsapply in industriessuch as fertilizer,chemical feedstocks, and-within agrictural processing-to malting and milling.

4.70 Government policy toward competitionis established in the Maintenance and Promotion of CompetitionAct of 1979, an enablingact whichprovides for the establishmentof the CompetitionBoard as an investigativeand advisorybody to the government. Ihe impact of the CompetitionBoard in the agriculturalsector, has, however, been limited. In many of the areas wherethe CompetitionBoard might be expectedto be most active, e.g., the marketingcontrol boards and the cooperativesacting as their agents, it is excludedby definition (one Act cannot overrule another Act). The Board lacks teeth in the form of penalties and is obliged to pursue deregulationas the means for removing restrictive practices arising out of the implementationof the MarketingAct. It also possiblylacks the political will to deal effectivelywith the large conglomerateswhich dominateSouth African industry and conmmerce.M

Marketing Strucures in the Homelands

4.71 As the homelandsgained authorityover agriculturalmarketing issues, they introducedvarying forms of control that emulatedthe existingstructures in commercialagriculture and the policy of food self-sufficiency. The four TBVC states and Lebowa created separate marketing boards. This led to general problems, mostly in terms of barriers to movementof commoditiesacross borders internal to SouthAfrica. The administrave costs of such controlmeasures was also high, as discussedin Chapters 1 and 2.

4.72 In generalthe yields from the small farm plots in the homelands are poor, and very little of the production is marketed. Food crops form the bulk of crop production in all homelandareas, and each

as membersi is hlf leeWdad hf ex-offico. he ex-offidomembersh was edarge in the mid 1980sto facilitat investtion of agricuural secor industie by fte incus of the Chairmanof the Coopative Counciland the Chairmanof the Noxal MarketingCouncil.

i In addition it is usualy difficult to prove colusion vcen whme ther is strong cuntial evidence. The recn investian of the Compein Board into the fetizer industryis a case in point. Tec study concludedthat collusionwas sUspeOed but could not be proven

CYqpker4: Mar*k and Pricng Page 81 ko*ahAfidan $,ricAibu: &r . Eran A o for*1w Frswe area is in deficitIn most yeas. Somemaize meal s likelyto be broughtto householdsby migrnts workersas remittancesIn kind. Tradersbring in foodto makeup the shortfallin effectivedemand. 4.73 An importantfeature of agricultre In the homelandsis the extent to which householdsare involvedin the market,whether for the disposalof surplus,or in order to makeup for food defict. Variousestimates have been madeof the marketpardcipation of smallholders.The table below gives a marketparticipation profle of uralhouseholds in KaNgwaneand the Ocumisaward of KwaZulu.

Tabl. 4.10: Market!avoheaent of RuralHouseholds MaKet inwlvment % of total Sal concentation Crop Net buyrs No net sales Net sLers pr % of total saia or buys 50% 70% 80%

% of householdi % % of households KaNgwa (n = 394) Maiz 68.7 7.4 23.9 62 2.8 7.4 11.2 Groundauts 81.7 4.6 13.7 52 3.0 6.1 8.6 Dry beana 96.1 0.3 3.6 - 0.1 1.3 1.5 Yuco beans 95.9 0.0 4.1 66 1.0 1.8 2.5

Kwa Zulu (n = 193) maize 95.2 0.1 4.7 49 0.5 1.3 2.4 Deans 84.0 6.2 9.8 54 3.0 6.0 9.2 Potatoes 93.6 3.3 3.1 40 1.6 2.6 3.6

Igiami:J. van Zyl and 1. van Rooyea, 1990b 4.74 Thedata indicatethe low levelsof sales,and the highrelative sales of the few householdsthat do sell food (e.g., that 2.8% of householdsin KaNgwaneare responsiblefor 50% of maizesales). Conversely,nearly all householdspurchase basic food staples. 4.75 Muchof the grainproduction by smallproducers is miled locallythrough service miling. Grain salescan go in manydirections from a homelandproducer: to millsor traders,to homelandmarketing boardswhere they exist,directly to the SouthAfrican marketing boards, to localinformal sales, and, in smallamouts (for itis ilegal underthe reguationsof the SouthAfrican) tD informalsales or giftst peopleoutside the homeland,such as migranthusbands travelling to urbanareas on a monthlybasis. 4.76 Similarlythere are few salesfor livestockdespite heavy stocking rates in all the homelands. Althoughcattle are seldomused for ploughing, they still have value beyond the market,in tenmsof milk, dungfor manre or burning,and asset holding and appreciation. No homeland marketing boards control meatsales, but there wereregulations on abatoirsand butchers. 4.77 By contrastwith commercialfaming areas, roads are often poor, marketsare distant, and marketinginformation is scanty. Withinthe homelandsthe poor roadsystem leads to isolatedmarkets andthe developmentof monopolisticand monopsonistic marketing conditions. There is littleinformation on marketdemand and market requirements outside the localarea, but the marketingboards in the TBVC areasand Lebowa attempt to providethis informationand protect producers against the marketingboards

Ca4. 4: Marie*ngand Prking Page82 SonuhAftc AMricitabMe:hnwe. PerfnuamceA OAtionsfor the Fmture set up by the large fam sector.

4.78 AMotherproblem cat be in reachingand dealingwith the SouthAfdican markeing boards fot productsproduced for markets outside the homelands.Furthermore, even where South African marketing boardsare not Involved,it is difficultto breakinto the whitemarketing structures outside the homelands, e.g., for horiculture, where vertcal integrationamong producers, distributors,and retailers is exceptionallystrorg.

4.79 The five marketingboards In the homelandseither set marketingcontrols or regulatethe marketingof the followingproducts:

Bophuthatswana: Statutorycontrols: maize,oilseeds, wheat. Sorghum:sets floorprice Eggs:registration of producers Cotton:Single registered agent Ciskei No statutoryregulations now (previously dairy, poultry, wool, mohair, watde).

Lebowa: Statutorycontrols: maize, wheat Meat:registration of butchers, Milk,eggs, poultry: registration and inspection

Transkei: Wool:single channel with basic sorting before export. Bread:registration of bakers,subsidization of breadprice. Meatand meatproducts: registration of butchers. Hidesand skins:now private agent appointed (recently decontrolled). Venda: No stautory controlson marketing.Sets floor prices. Marketingby the NorthemTrasvaal Cooperative. 4.80 Tbeboardshave to operatewitin the contextof poorborder controls and dominant South African boards. Therefore,products move in and out of the homelandswith relative ease, and this setslimits on prices. No board, for instance,can chargea veryhigh marketing levy, or all farmproduce tat is not sold infonnallywill pass out of the homelandinto the formalSouth African system. This phenomenon appearsto haveinculcated an exceptionaldegree of realismamong board staff in the homelands.Where statutoryconrols exist, all boardsclaimed to ignoreinformal sales (andthe loss of tax incomethus incurred).

4.81 Nevertheless,the existenceof the borders,the dispersionof tractsof homeland,and the statutory controlson movementsoften have the effectof iterferingwith regional trade, and increasing the distces overwhich goods travel to conOmers,thus also increasingthe marketingmargins. In otherInstances, evenwhen there are homelandreguatons whichfavor entrepreneurial development-e.g., for dairiesin Bophuthatswana-farmersmay be so far fromhomeland markets and so near to other SouthAfrican

C7aptr 4: Ma kting nd Pi*g Page 83 marketsthat they wouldhave to producefor the latter and thereforefollow the SouthAfrican reg ins, whose stringency makes entry extremelydifficult.

4.82 The boards, like others concemedwith agriculturalproduction in the homelands,tend to pursue self-sufficiency,sometimes at the cost of more rationalfood productionand marketingand often at direct cost to consumers,who lose when restrictionson 'imports' push up local producer prices.

4.83 In the case of export products (e.g., wool from Transkei and Ciskei, subtropicalfruits from Venda)the homelandmarketing boards providethe link with the SouthAfrican boards which export the products, often providing transport and some grading en route in return for a levy. Prices are set in the world market. The marketingboards attemptto ensure that the products meet export requirements.

4.84 The boards are financedin varying proportionsby governmentgrant, levies on handling,and, in the case of Bophuthatswana,interest on the accumulationof levy incomeover the last ten years. Levies often appear to be set at the same levels as those of the South African boards, regardless of the actual costs of the homelandboard's operations. Someboards charge levies on products that are passed to the South African market, so that the farmer essentially pays a double levy, as with livestock in Bophuthatswana,which has caused ongoingfriction with the SouthAfrican board.

4.85 The Lebowa MarketingBoard, on the other hand, uses its profits from charging the same levy as the SouthAfrican Maize Board to provide a higher price to farmers by giving them a rebate; it is also able to provide somesubsidy to transportersand cooperatives. It sells produce at the sameprice to local millers as the South African Maize Board, but is assured of sales due to its geographic advantage (NorthernTransvaal as a region must importfrom further south). Effectivelyits operationsthus reduce prices to the millers, relative to the prices of the S.A. Maize Board, but not relative to an uncontrolled price.

4.86 In the other five homelands,without marketingboards, regulationson marketingtend to follow those of the South African Boards unless the homeland has enacted new legislation. KwaZulu for instancehas been developingits own legislationon the utilizationand processingof food. It would appear that the homelandlegislation is oftenbetter suitedto local conditionsthan that of the SouthAfrican boards and therefore tends to keep down handling and processing costs, thus benefiting, for instance the producersand consumersof meat products.

4.87 However, the SouthAfrican boards tend to take advantagewhere they can from marketingin the (non-TBVC)homelands, such as in atempting to obtain levies on maie sold by KaNgwanefarmers to neighboringSwaziland. Such activitiesappear to provide cause for irritation in the homelands. The Lebowa Marketing Board has successfullytaken the Maize Board to arbitration in order to prevent Lebowa farmers being forced to pay levies on maize marketedoutside the homeland.

4.88 The Secretariatof the EconomicCommunity of SouthernAfrica (SECOSAF)provides the forum for discussionand negotiationof difficultiesbetween South Africa and the TBVC areas. In late 1990 SECOSAFset up a committeeto examinegrievances of the homelandgovemments against pracdces of the South African boards. Such items as the prohibition of the transport of bananas from Venda to Bophuthatswanathat did not necessarilymeet the standards of the South African Banana Board were

Chaper 4: Markedingand Pricing Page 84 Sg & AfcoanAfricu : $Sructur.Performace & Odonsfir tho&aurg negoted successfully(they can be Utrnported in sealed vehicles). Thus, while low-incomecoOSume in Bophutiatswana can now purchase less than perfect but cheaper bananas, low-incomeconume outside the TBVC areas cannot, includingthose in the other homelands.

MaUutin Margins

4.89 Trends in marketingmargins can be analyzedby determiningthe producer's share of consumer value. Table 4.11 provides data on the producers' share of whatthe consumerpays for grain, meat, fats and oils, dairy products and eggs, vegetables,fruit, sugar, and for agriculure as a whole. From Table 4.11 it is evidentthat there has been a sharp increase in markaing marginsover the last decade. 'his has resulted in higher increases in food prices than in ether consumerprices.

Table 4.11: Producers' Share of ConsumerValue )

Year Tota Grmin Meat Fats and Daity Vege- Frut Sugat ois and tables

1974 55.2 41.6 26.0 38.5 69.7 34.2 31.0 43.4 1975 53.8 42.0 57.0 40.1 70.7 34.S 31.8 61,8 1976 53.0 42.7 56.8 40.7 68.4 32.0 34.2 62.4 1977 48.9 40.1 54.4 39.2 60.4 29.S 36.9 41,5 1978 46.9 34.2 52.8 36.5 58.7 31.0 35.2 34.0 1979 48.9 35.6 53.5 38.1 62.5 33.S 35.9 37,1 1980 53.0 41.6 59.3 39.7 66.2 34.1 35.9 4S,0 1981 52.6 40.7 59.1 37.5 56.6 30.5 37.2 41.0 1982 51.3 38.0 55.5 38.0 67.3 30.7 36.8 38,5 1983 49.0 36.1 52.0 3'5.2 65.3 31.8 34.6 44,S 1984 48.1 33.7 52.7 33 6 64.6 26.6 35.5 38,3 1985 48.8 33.8 54.5 32.3 63.5 29.2 36.8 32,S 1986 47.5 33.t 51.6 33.0 62.3 30.1 36.7 33,2 1987 48.3 31.0 S4.7 38.8 61.7 32.7 35.8 27,S 1988 46.4 28.5 51.5 39.4 61.5 31.S 33.0 25,S 1989 46.4 24.7 51.2 35.9 65.5 27.5 35.0 27,9 1990 43.1 2S.0 48.2 37.3 56.7 29.4 31.4 29,0 1991 42.0 25.0 50.7 37.9 48.9 29.0 33.2 26,9

Sj: Dir_ecot of Apiultal BconoiAcTrends, DepaxineAt of Agrcuture, Abstdmtof Ariltural Stadstis, 1993.

4.90 An analysis of marketn margins In the food chain between 1980 and 1991 by the Board on Tariffs andTrade (BT, 1992)indicated a price divergenceof five percentagepoints per anum between the producer price index for farmers and consumer food prices. The price divergenceof S percentage points per annum can be broken down into the following components:

a. Input cost inflation (3.0 percentagepoints);

b. decliningproductivity in the food chain (1.0 percentagepoints);

Chaupt4: Mrkeng ad Pibg Pagp 85 Snuh Af,ca Aegrcuture.&t . PaM ce A Agiggs for thge&dtW c. price and supplysabilization (0.5 percentagepoints);

d. wthdrawalof state subsidies(0.3 percentagepoints); and e. liftingof price controls(0.2 percentagepoints).

4.91 Accordingto the BTT(1992), salaies andwages contribute up to 30%of expenditureson food. Inflationaryprice increasesin wagesand salariesat each level in the food chainhave contributed significantyto the consumerfood price inflation. Each level-i.e., primaty storage and distribution, food manufacturingand wholesaledistribution, and the retaillevel-contributed approximately equally to the totaladded inflaion in the foodchain. TheBlT's analysisof the structureand concentrationin the food dchinconcluded that:

a. the retailingof mostfood itemsis onlymoderately concentrated: about 45% of food is soldthrough 890 hypermarkets and supermarkets,the majorityof whoseoutlets belong to four groups,while many other outletsaccount for the rest; b. at the primaryproduction level the growingof mostfoodstuffs is welldispersed;

c. an analysisof the foodprocessing and distrbutionlevels indicates a fairlyhigh degree of concentation:the joint turnoverof the six largestfood groups is equivalentto 57% of the valueof processedfoodstuff production in SouthAfrica; and d. it is at the primarylevel of storageand distrbution that concenationtends to be biggest and that artificialbarriers to entryexist. 4.92 The introductionof the VAThas had a short-tm impacton foodprice inflado. In copari withthe positionbefore VAT, food inflationin April1992 increased by 6.8 peroentagepoints as a result of the changein the tax bace. Ihe sharpincrease in foodprice infladonsince the third quarterof 1991, reahn a high of 30.4%in July 1992,is mainlythe resultof the droughtand the short-termeffects of the introductionof VAT. However,the foodprice inflationrate without these disturbances is estmated at 15%,which is stillhigher than boththe non-foodrate of inflationand the rateof increasein farmgate prices.

4.93 The marketingpolicies described above resulted in a highdegree of overaI foodself-sufficiency in SouthAfrica; the productionof commoditieswas well in excessof domesticconsumption. The positive swpplyreaction was often created by high levels of domesticprotection against lower-priced world market imports. However,at the resultinghigh domestic price levels, domestic markets often did not clear,and surplusproduction was exportedat a loss. Table4.12 providesthe relevantaggregate data.

Chw 4: Maikng and Prcng Page 86 South Adcom Agdgab=F: *raweura.Par1dWnk A Ond foon1rthe Itbur-

Tabo 4.12: AMers. Poduton and Commwumtionof Sgelt I&UuWural Commdiftes in South Africa. 1985l1990

onumpIica Com _odt Impos Eots Poducion Toal" Human** SSI* (1000ton)

Wheast 94 449 2.612 2.262 2.119 115.50 mains (white& yclw) 484 1.689 7.422 6.127 2.615 121.10 Plotaols 5 8 1.042 1.039 872 100.30 Vegetables 4 27 1.739 1.717 1.54S 101.30 Sugr 63 863 2.044 1.258 1.258 162.50 Beef 81 16 S79 644 639 89.90 Mutton, sot'sm t aUmb 14 1 182 195 193 93.30 Fbrk 1 2 110 109 108 100.90 ChIcken 3 0 521 524 519 99.40 Eggs 0 3 181 178 169 101.70 Deciduous& subtropal fruI 0 466 1.37 897 808 152.30 Dairy producs 35 58 2.34 2.321 2.321 101.00 Sunfuwer sed ol 14 1 84 96 85 87.50 Cirs fruhts(fresh & 0 426 706 278 278 254.00

* Availablefor u=e - Openingstock + Production- Clofing stock + Inpxos - Expotss * Net humxan _ouupion - Availbblefor use - Ohewruses - Losses, and furthe adjusted for extaction rate *"" SYI(Wdf-sufilmaoy Index)A - TOtl production I TotEl COnSumpton zS 100 tbnnss Food balaoe shess of to Directorate of Agiltrnal Economic Theads of the Departmnutof Agricuture (as

4.94 Table4.12 showstota producion,surplus production for the expot market,and the degreeof slf-sufficiencyof SouthAfrica with respect to majoragricultural commodities during the perod 1985 to 1990. The fofllowingchrceIstics emerge. a. SouthAfrica self-sufficientin all the importantstaples. In spite of the periodic drugs experieAcedduring the 1980s,South Africanagriculture still succeed3din producngsurpluses. Staplecrop production can therefore drop (in totl) beforeSouth Atca becomesa net Importerof these productson a regul basis. However,some individ commodiiesin thisgroup are importedon a net basis (e.g., oiseeds). b. In hort u" production,pardcarly fruit, South Africa is notonly self-sufficient, but tDa larg de dependenton earningsfrom the exportmarket. he stuationis more faworabletuha tha of cropproduction. c. In contas to cropand horticultua products,red meathas a self-fficiency Indexof lowerta 100. Thisimplies tha SouthAfrica did not produceenough red meatduring

C7ew 4: MokT8 ad PWCh. Pae 87 SouthAfrian Aricubur: Structe. Peffomanca & O0*io for the FMure

the years 1985 tD 1990 to meet domesticrequiremets. These shortageswere supplementedby Importsfom, amongothers, Namibia, Botswana and some European countries. 4.95 Red meat,coffee, rice, vegetables,animal fat, and vegetableoils are the mostimportant food productsimported. The totadgross value of agriculturalproduction in SouthAfrica was almost RlS,000 millionin 1987,and food imports amounted to about R1200million. Food exports in the corresponding periodamounted to aboutR2400 million. 4.96 Pricingpolicy played a substantialpart in achievinga situationof surplusproduction and exports. Table4.13 showsthe differencebetween domestic producer prices and world(reference) prices for a numberof majorcommodities. The suggests that it is possibleto importa numberof products(admittedly onlyIn certainyears) below the domesticcost of producingthese products-e.g., meat products and wheat. On the otherhand, domestic transport costs also play a role, andhelp to isolatethe SouthAfrican interior markets:total transport costs of importedmaize to the majorSouth African market (PWV-rea) amount to roughy R220.00per ton-R95.00per ton forshipping and R125.00for railtransport from the port of entry. Moreover,these transport costs work in reversewhen commodities are exported,often rendering the exportof low-value,high-volume commodities unprofitable. At the sametime, it mustbe recognized that the over-valuationof the Randhas significandyinfluenced agriculural trade patterns. If a new governmentIs ableto operatean exchangepolicy that doesnot overvaluate the currency,many of South Africa'sagricuura commoditieswill become more competitive.

Table 4.13: Diffrence betwoenSouth Afian ProducerPrice. and the Wwrd (RefereneT)Prioe for Mahr Crove

World (Rcference)price Produce price Product Rlt 1989/90 1990191 1989/90 1990/91

maize RUt 330 385 301 352 Wheat R/t 375 336 457 503 Sorghum RIt 271 294 231 250 Soyabeans Rtt 698 723 591 608 GroundnutB Rlt 2,662 4,349 1,076 784 sunflower RUt 864 858 692 656 Beef RIt 4,380 4,597 5,493 5,401 Sheep (EEC) RUt 4,251 4,620 6,599 6,122 FPg (EEC) RIt 3,426 4,08S 3,479 3,628

Sbmco: Unpubished dat obtined from the Depautment of Agricuklre.

4.97 Termsof trade are importantin indicatingthe internationalcompetitiveness of a country.Souti Africanagdculture has beenexperiencing a weakeningterms of trade. Domesticdouble-digit inflation, whie major tradingpartners experienced single-digit inflation, undemined competitiveness on world markets.Some analysts believe that the competitivenessof agriculturehas improved recendy because the Randexchange rate deeased fasterthan inflationhad since 1983. However,if the weakeningof the exchangerate Is such that the risein costs(via imported inputs or highercosts to domesticinput suppliers) outweighsexport price increases, the competitiveposition wil weaken(Liebenberg & Groenewald,1990).

Chqkr 4: Mar*elg ad Pricing Page 88 SosthAfkn Alrcu SROWur.PErfonae dOno for e re

4.98 Th. terms of trade of SouthAfrican agricultre decreasedfor nine consecutiveyears since 1974; then party recovered In 1984 and 1985,only to fall back again. Ibis decreasewas strongly influenced by lower prices for agricult products in intrnational markets and higher prices for most imports. Exch e-rate changes were not able to realize more than short-term improvements. Thus economic pressure on agriculturalexports is increasing.

aqclusiors

4.99 Clearly many parts of agriculturein South Africa have been protected for many years. This is true for producers and for the marketingand processingactivities. This has resulted in a large transfer from consumerto producer and a highly concentratedprocessing sector with the associatedmonopolistic profits. Ihe one part of the sector that has been exposedto internationalcompetition-horticulture-has performed very well. The production of these crops is inherently labor-intensive,and the scope for substitutionis very low. The sector is rapidly growing and has given RSA an internationalcomparative advantage,with considerablescope for expansion.

4.100 One of the opportunitycosts of the curent marketing system has been the developmentof a relatively concented sector aimed at specific market segments. By catering exclusivelyto white fiamers, it preventedthe developmentof either black fanmingor informalmarketing systems which could have created substanti black employment At the same time, by emphasizingthe marketingof high qualiy, first-world agriculturalproducts, it lost or did not create jobs in marketingof lower quality produce for either export or domesticconsumption.

4.101 Current trends in marketingappear to be movingin the right direction-from a controlled, cost- plus pricing systemto a more market-orientedone (removalof domesticcontrols, reductionof protective measures,and tariffication). Giventhe high levels of concentrationin the agricultureand food processing sectors, the establishmentof clear policy guidelines setting out Government's objectives and policy instrumentsis a high priority. Three alternativepolicy approachescan be distinguished:

a. laisezfr: Under this policy the Governmentwould adopt a passiveattitude towards concentration, hoping that the possible benefits from economies of scale and competitiveness on international markets will outweigh the potential costs from monopolisticpricing. In recent decadespolicy has been rather close to laissez-faire.

b. acjwUaXi-trust aoach: Under this policythe mandateand powers of the Competition Board would be substantiay increased,with a view to reducing the existng levels of concentrationthrough induced or forced divestmentof holdingsby the major firms which dominateparticular industries. In practice, implementationof this policy approach is likely to prove difficult: the GovernmentwiU condnue to lack sufficientinformation to pursue a legal approach to anti-rust; given the concentrationof liquid funds in the economy,forced sales of subsidiariesare likely to lead to purchasesof divested firms by existingconglomerates rather than by new entrants to the market, and the unemployment implicationsof possibleclosures induced by an activeanti-trust policy wouldbe politicaUy difficult to accept. Finally, the existing conglomeratescan and wiU exercise strong

Ca*r 4: Meting ad Pricing Page89 SouthAHfrcan .rure: Sture. Performace& Ovlpons for thf Future

financial leverage to maintaintheir position if threatened.

C. an WI= gproacb to the economicenvirorment through deregulation Under this third approach, govermmentwould not actively attempt to break up the dominantfims, but would seek to promote changesin the economicenvironment which, over time, would remove the economicbasis for concentrationand reduce barriers to entry for potential competitors. Promisingareas for this type of approach include continuedtariffication (whichreplaces administrative control over importsby a marketmechanism), dereguation of legislated monopolies and restrictive agency agreements, active enforcement of competitionpolicy against restrictivepractices by firms in the sector, and the removalor reductionof regulatorybarriers to entry into the informaltrade. To adopt this approach requires recognitionthat the problem of concentradon cannot be resolved in a short period, but that over a longer period the reality or threat of competitioncan be the most effective instrment In combattingthe adverse aspects of concentration.

4.102 Concentrationhas important implicationsfor the policy of deregulation:in the extreme case dereguation may involve the substitutionof a public monopoly(wnich is in theory fully accountableto public scrutiny) by a private monopolywith statutory powers. 'his is highly undesirable.

4.103 Assistancein marketingto homelandfarmers is an integral part of agriculturaldevelopment, but it is not clear that this need is met by the current operationsof homelandmarketing boards. The major argumentfor their continuationis that they provide a mechanismfor defendingthe interests of some relativelysmaller producers againstboth the SouthAfrican boards and a highly concentrateddistribution system. The South African boards were set up by white producers in their own interests and show, at the very least, a bias toward large farmers, e.g., in the encouragementor requirementof bulk handling. Mecaisms will continueto be required for bulldngthe produce of smallfarmers, as well as to negotiate with the boards and/or with distributors, even in a deregulatedenvironment.

4.104 Income elasticitiesdemonstrate that there is substantialexpansion room in domesticmarkets for higher-value food products if the income of lower-incomegroups increases. This implies that a developmentstrategy aimedat raising the incomesof the poor, of whom a large proportionreside in rural areas, will have positive effects on the demandfor food.

Cuhe 4: Mai*ei and Pik8 Page90 South AfiJcanAgIricuiture: Struenre. Perfonnance & O fiOnsfor theFuture

CHAPIER 5: EFFICIENCYOF AGRJCUTRL PR

Introdu n

5.1 In order to analyze the broad subject of efficiency of South Africa's agriculturalsector, this chapterexamines several dimensionsof the efficiencyissue. As a startingpoint and to gain a perspective on the subject, a number of general remarks about efficiency need to be made. First, efficiency presupposesthe existenceof an objective, e.g. maximizingprofits or maximizingoutput. It is only against such an objectivethat it is possibleto evaluatethe activitiesof an economicunit, and ask whether or not there are less costly ways to attain the objective. Second, the evaluationof economicefficiency does not question the basic rules that govern economic activities-e.g., the existence of a certain distributionof property rights and a set of prices createdby marketsand governmentintervention. Hence, this examinationof efficiencymakes no judgementabout the merits of the presentdistribution of land and other resourcesin SouthAfrica. Third, efficiencycan be evaluatedfrom the perspectiveof an individua entrepreneur, in which case it is called private or financial efficiency,or from the wider perspectiveof society, when it is called social or economicefficiency. The differencein perspectiveis important. For instance, an individual may be efficientlymaximizing profits given the basic rules that govern their economicenvironment, whereas society seeks to create remunerativeemployment for as large a number of its citizens as possible. One can imaginesituations in which the individual'sbehavior can be deemed privately efficiert and sociallyinefficient at the same time. Moreover,since society can decideto change some of the basic rules that govern the economicbehavior of the individual,behavior can be efficient given the 'old' rules of the game, but inefficientgiven the 'new' rules of the game. In the contextof SouthAfrica, such distinctionsare important. In the past, farmers may have efficienty reacted to a set of rules-including distorted price-signals-whichwere related to society's objectivesat that time (e.g., subsidizingcapital in order to attainfood self-sufficiency).In the future, such rules may change andwhat was sociallyefficient in the past may prove inefficientin the future.

5.2 Large-scaleagriculture in South Africa has consistenly producedsurpluses, and as a result has typically been regarded as an efficient sector from a self-sufficiencyperspc.dve. In recent years, however, the persistent financialproblems experiencedby many farmers-largely as a result of drought and policy changes-have caused some important subsectors :n agriculture to be seen as fmancially inefficient. Moreover, given the high unemploymentlevels in the economy, agriculture's current contributionto employmentis below its potential. The issue of large farm efficiencyin South Africa is criticalto the developmentof a strategy for the rural economy. If, as some argue, the large-farmmodel is a valid one-characterizedby economiesof scale-then by removing existingpolicy distortions it will be possible to improve the financial efficiency of the sector as a whole. If, however, there are no significanteconomies of scale associatedwith large farm productionmodels, simply reforming policy will only yield a portion of the potental efficiencygain. Ihe balance of the efficiencygain could only be achievedby restructuring-throughdownsizing-the present farming sector.

5.3 The dominantmodel of agriculturalproduction in South Africa is that of the large-scalefarm basedon capital intensiveproduction techniques and relativelyextensive land use. Whilerecognizing that the questionof whether an enterpriseis efficientcan be answeredin several ways, the central questions to be answeredin the contextof SouthAfrica are whetheror not economiesof scale exist and, if so, what

ChapterS: Effickn ofAgural ProdAcio Page 91 S9uAhAfrican Agricu)urM. Structure. Performance & Optionsfor the Future are the sources of these economies?

5.4 Unfortmately in South Africa, a direct comparisonof efficiencybetween the large-scale and small-scalefarm sectorsis difficult if not impossible,because black owned small-scalefarms have all but disappeared. Ample historical and internationalevidence suggests, however, that when compared to small-scalefarms many large-scalefarming enterprises in South Africa, as elsewherein the world, are inefficientwith respect to profits, employmentand output generatedper unit of capital invested.

5.5 The move towards larger farms in SouthAfrica was strongly associatedwith farmers' adoption of mechanization,which was facilitatedand strengthenedby subsidizedinterest rates, tax breaks, and the support of output prices above border parity. Yet in spite of massive additionalfinancial assistanceto farmers, the profitabilityof agricultureremained low. By 1984, an estimated22,700 farmers (33% of the total) had debt burdensbeyond the criticallevel (SAAU, 1984). Most of these farmerswere concentrated in the Transvaal and Orange Free State-the regions where mechanzation of monocroppedcereals had been pushed to economicallysub-optimal levels and exposedfarmers to high risk levels.

5.6 The policy objectives that justified these extensive interventions were twofold-food self- sufficiencyand maintainingincome levels for large-scalewhite farmers. Agriculural production in the RSA increasedat a rate of over 3% annualy over the last four decades, while the populationincreased at an average rate of 2.8% (van Zyl et at., 1987). The social objectiveof maintainingincomes of full- time white farmers at levels comparableto the incomes in the urban sectors seems also to have been obtainedfor at least some farmers. From 1970/71to 1991, total net farm incomeincreased in real terms from R2,728 million to R3,469 million, but with significantfluctuations throughout the period and a c.nsiderably skewed distribution.

5.7 In order to analyzethe efficiencyof large-scalefarming in South Africa, this chapter examines three dimensionsof efficiency in agriculturalproduction.

a. Price efficiency. Givena certainoutput level, couldone choosea differentmix of inputs and achievea reductionin costs?

b. Scale efficiency. If one could expand all variable inputs proportionally, thereby producingmore output, could one reduceaverage costs of production? The answerto this question is often formulatedas increasing,constant, or decreasingreturns to scale.

c. Technicalefficiency. Given a certainoutput level, can one reduce costs by using less of all inputs? The answerto this questionis often linked to technologicalimprovements and growth in physical yields.

5.8 Price efficiec. The agriculturalsector experienceda policy-inducedshift towarda large-scale, capital-intsive farm technology,heavy relianceon intermediateinputs, and toward less dependenceon labor. Ihis reflects farmers' rational reaction to the prevailing structure of incentives, in particular, taxation and financial assistanceprograms. Althoughthe price of the input bundle associatedwith the prevailing technologyis one element in the structure of incentivesfaced by farmers, it has not always been the dominant one. Prior to 1973, large-scale farmers display considerableprice efficiency.

Chapter5:S. 1eny QfAgA*fwal Producio Page 92 Souh Afica Aaricure: Smtcugre PZgCrkre &.0,Wons bfrthg Fure

However, startng in the mid-1970sfarmers demonstratedless responsivenessto relative input price changes. This seems to have b3en the result of a combinationof output price distortions, financial assistanceprograms, tax incentives,and technicalimprovements that offset inputprice changes. In effect, farmers respondedto a distortedset of incentivesthat induceda diffeent productiontechnology tha that which wouldhave prevailed in a more market oriented regime.

5.9 Moreover, given an excess supply of African labor in the wider economy, the pattern of agriculturaldevelopment chosen was not price-efficientfrom an economy-wideperspective. During the 1980s, partal liberalizationof credit and outut markets, along with tax reforms and periods of severe drought, changedmany of these incentivesand seemedto have reversed some of the above trends.

5.10 Sle efficiec. The same distortedincentive stucture that affectedprice efficiencyalso created incentivesfor large-scalefarms. As a result, prior to the end of the 1980s,there has been an increasingly concentratedagrarian structure in SouthAfrica. The results of this chapter indicatethat a wide range of farm sizes (above 100 hectares) are efficient with respect to scale-in order words, a larger farm is not necessarilymore efficient than a smallerfarm. The crucialfactor in determiningthe greater efficiency of a large farm over a small farm is managerialexpertise. For farms smaller than approximately100 hectares,there is little empiricalevidence on scale economiesfrom South Africa, however, international experienceindicates that in an undistortedpolicy environmentsmall farms are often at least as efficient as their large-scale counterparts,provided the working capital constraintis removed.

5.11 Technicalefficien Trends in technicalefficiency can be grouped into three distinct periods. First, improvementsin the technicalefficiency of South Africanagriculture were virtually absent during the 1950s and early 1960s. During this period, agricultura production increasedthrough an expansion in cultivatedarea, supportedby the introductionoftractors after the SecondWorld War. Second,modest, but significant,gains in technicalefficiency appeared in the late 1960s and coniued during the 1970s. These gains are associatedwith the itesification of agricultureduring this period. Also during this period the relationshipbetween labor andmachinery changed from a complementaryone duringthe period of area expansionto one of substitution,especially as a result of the mechanizationof the harvesting process in cereal farming. Third, substantal growth in technicalefficiency was recorded ater 1983, as farmers adjusted capital inputs to dhangesin the structure of investmentincentives. For the post-war period as a whole, however, the trends in tecbnical efficiencydiffered dramaticallyby subsector with livestockand field crops performingwell belowthe average,while the export-orientedand labor-intensive horticure subsectorshowed dramaticincreases in technicalefficiency.

5.12 Hence, the picture of South African agriulture tha emergesfrom this analysis of efficiencyis a varied one, but one for which several subsectors can be characteized by: (1)a gradual deteriorationof price efficiencywhich in effectshows that farmers were compensatedfor the Impliedloss of efficiency through other channels; 00i)scale efficienciesthat are the result of a combinationof a distorted policy enviroment and managerialability; and (iii) unevenimprovements in technicalefficiency that are modest by intnational standardsand differ markedlyby sub-sector.

5.13 The policy distortionsthat have given rise to the inefficienciesthat characterizelarge parts of the agriculturalsector (e.g., subsidizedinterest rates and tax concessionson agriculturalinvestment) are not particularly difficult to reverse. Over the past decade, a wide range of policy changes in the general

Clner 5: FffieC of Agku&W Pfroni Page 93 Sosh Afia Aridet: Stre. RerforMncA Obdonsfor the FUre diection of greater market orientation have been introduced and have affected both agricultural inputs and outpus. The impact of these changes has varied considerably by subsector: debt-free producers of some export commodities have benefited, while profits for others, notably the main grain producers, have fallen. For the sector as a whole, however, the capital-intensity of production has been declining thoughout the 1980s and employment was increaing through 1987. More recently, however, employment has again declined. Although the cause for this decline is unclear, it Is likely due to political uncertnty over the future of white-owned farming.

Price Efflciency

5.14 Intervention in the markets for the main factors of production for more than a century have produced a highly segmented market structure. In agriculture, factor market segmentation produced a situation of relative land abundance for a minority of white farmers, who could draw on an excess supply of black labor at very low wages relative to the rest of the economy. Ihe resuting factor-price structure whih faced white famers before 1970 can succinctly be dchaized as relatively inexpensive land and labor.

5.15 Such factor market segmentation needs to be taken as a given in judging the price efficiency of the white farming sector. Price efficiency-the rationa response of a profit-maximizing farmer to price signals-will be gauged in terms of the prices that farmers actually faced, even though these were largely the outcome of insttutional arrangements which sought to suppress price-setting mechanisms predicated on free and competitive markets.

S.16 Figure S.1 shows the aggregate land and labor quantty indices in commercial agricuture for the period 1947 to 1991. I*nialy, the land index grew as cultivated area was being expanded through increaed mecanizaton. I grew at 0.3% a year untd about 1960 and te began aslow decline of about 0.25% per annum for the rest of the period. The labor index grew rapidly until 1959, at 1.3% per annum, then wavered until the late 1960s, before beginning a decline at 1.18% a year, from 1968 to 1980. Then there are three particularly bad years, followed by a recovery in the mid 1980s which took labor use back to the level of the early 1970s. However, since 1986 labor use has declined precipitously for reasons which awaitkfuher analysis, but may very well be related to changes in the political economy.

5.17 Prior to 1970, the cultivated area increased, particularly in maize production in the summer rainfal areas45, as oxen were progressively replaced by tractors. Larger areas could be managed and labor use inialy increased, pardy because of the greater utlization of fertilizer and improved seed vadeties (Payne, van Zyl and Sartorious von Bach, 1990). After 1970, however, the substitution effect Of m n dominated, especially with the lntroduction of the combine harvester, whih sharply reduced labor demand at harvest time.

* Te summer ami aeas foowd d patern in th winta reXo, whe the,expao of Xteutvated ar wa bg* complatby 1947.

Ohapr: S Fjekc, cfAiulra Pfrodte Page 94 SosahAf=n Aerkicuura:St.uctur PerOnwR A Onxonsfor the uf{r

Figure S.1: May Inuts (1947-1991)

JIG S-... , ,.:

as ------

Yew Source: Ihirtle et al., 1993

5.18 Ihe period 1960 to 1983 can be caacteized as one of substition of machinry for labor inducedby a policy environmentwhich includedcheap credit (negativereal interest rates) and tax breals for aicultre, allowing capital equipme to be written off entirely in the year of purchase. The Pass Laws conItutd to this trend, especily after severe enforcementbegan in 1968. By the end of the drug period of 1981-3 the ctedit and tax concessionshad been largely eliminated, the gold price plummeted,and the Rand was devalued. These eventsled to a dramaticreversal of the historical trend, as cheap labor was substitued for expensivecapital 4 . However,the increaseduse of labor between1983 and 1987 proved ephemeral. Labor use declined rapidly after 1987, probably due to dhangesin the political arena.

5.19 Figure 5.2 looks at trends in capita inputs. The machineryinput index grew at 7.57% a year until 1958, then at 0.76% until 1981; since then it has fallen4.56% a year. This leaves little doubtthat the relative price change that occurred in the early 1980s was enoughto cause a movementaway from machinery, as farmers were forced to become more competitiveand efficient in their use of capital. Similarly,the series for fixed improvementsshows the effects of policiesthat influencedthe real cost of capital4. That series grew rapidly until about 1967 (4.82% per anmumup to 1970) and since 1970 has

45 Van Zyl (19) showedth at elicids of subton i SouthAfriu agUclue we very low, a for he attrbutd to dxbility borncof lak of competdii Hd this s be donea fewyars late, so thatthis evaod be -pdup, th raunkmit. bodia

O Noectht thc sediespertais to e to landand bug, butdoes nat Inudeland vaues

Ch_$S. F: kffifApiu&wa Prod"= Page95 Skh icaa AtiutWbe: Srmaurt Pedfosa,we A ,* fonrdid iw beenfalling at 5.47%per anmnm.The cleardeclim since 1976can at least In part be attributedto the endingof subsidieson farm construction,especialy on housingcontuction for hiredlabor. Figure5.2: CapitalInputs (1947-199fL

260

2 ------20 ------§ 2180 X . 20 ._._ a 0-_ _\ _ __,_ 47-91555565 . _...... 56. 5788. 80 ------*It0.160 .1-/y...... I - /...... 140 .._-_---- ..-...... J!md ------3 120----r------, 0

sk OO ...... ______47 49 51 3 55 7 596163656c7 69 717375 79 8183 857 89 91 Yewr Source: Thie et al., 1993

5.20 IntermediateInputs (dips and sprays,ferizers, packing,fuel, etc.) appearin Figue 5.3. This indexgrewatS.39% per ammmuntil 1978andthenbeganto fail slowlyovertherestoftheperiod. There wastremendou growth in the use of dips and sprays,which grew at 9.74%until 1972,then at 20.56% untl 1980,before beginning to faUat 1.83%per yearover the rest of the period. This increasewas so enormousthat Figure 5.3 is reproducedwithaut dips and sprays, so thatthe movementof the otherseries canbe seenmore clearly. Packing,fuel and *other iqns recordedslow growth. However, ferilizer-a substitutefor land-grewat a rate of 7.73%per annumup to 1979,before declining by over4% a year for the remainderof the period. 5.21 In conclusion,the overalltrends in the use of inputswere the following: (1) two decadesof declininglabor use up to 1991,intrupted by a sharp,but apparently temporary,reversal in the mid 1980s; (ii) increasinguse of madhineryup to the ealy 1980s,followed by a declinethroughout the remainderof the decade;

(ni;) increasinguse of intermediateinputs Cm particular pesticides and fertilizers)up to the early 1980s,followed by stagnationfor the remainderof the decade.

ChaeS: S. cij_ny qf4i*ural Poduio Page f Soh Afijan A rw : St=WtrE. oance & OnIonhr te Fgture lse broad trends will be relaed to relative price changeswith respect to land prices, labor costs, and the costs of mechanizadon.

Figure 5.3: IntermediateInputs (1947-1991)

ma ...... _...... GM ...... ~A

;SO...... ,..g.

no...... ------w------Mw-A -4u---z 400...... >^..f...... ---P - 3Sa ------a-:::::::: ~ AA------*9**----

25 ------*--- ...... -

,O ------1 ------so ff - 47 49 5t 53 55 5759 6163 65 6769 7173 75 77 79 8183 85 8789 91 Year

Surce: ldre et al., 1993

Land orice

5.22 The 1913a and

5.23 Market prices of rural land are detenindr by the value of land as an asset in the portfolio of an investor-the net present value of all mutualy non-exclusive profit strems, which includes, but is not restricted to, the value of agriculfturalprofits. Thus, in considering the land price, one must distinguish between the market price and the price as deuined by the agricultural value only.

.. . o So5tAfricn Afrf : Srut. PeCo&mance& Opions for theFiure 5.24 VanWyk (1967)compared average market prices of landin whichthe LandBank was involved (1912-1965)and of all rural transfers(1927-1964), on the one hand, and averageagricultural value as estimatedby the Land Bank(1912-1965), on the other. His studyshows that: a. Between1912 and 1941,the increaseof landvalues-both market and agricultura-was gradualand consistent; b. Since1941, purdhasers throughout the periodpaid considerablymore for landthan the agricultu valueswould indicate. For example,in 1960,the averageBank valuation was R20/hecte, but purchasesfinanced by the Land Bankaveraged R27lhectare. In 1950,the averagevaluation was R27/hectare,while the averagepurchase price was R40ihecare;and

c. Since1951, the differencebetween market price and agdculturalvalue increased. 5.25 Thereis considerableevidence to supportthe hypothesis d muralland investment staegies were significantlyinfluenced by declinesin the cost of capital,as measuredby the real interestrate on rural landmrges. Theinfluence of trendsin itural profitability(on investment in rura land), which declinedsince the mid-1970s,seems less pronounced.For instance,whereas Behrmann and Collett,in an econometricanalysis of farmlandprices in SouthAfrica between 1933 and 1966, find a positive correlationbetween the increasesin real landprices and increasesin farmprofitability, they attribute the rapidincrease in marketprices after 1962to a highlyelastic reaction to the sharpdrop in interestrates. 5.26 Van Schalkwykand Groenewald(1992) attemptedto measurethe effect of land quaity characteriscson landprices from 1976to 1988.48However, while quality differences among Irrigated, rainfed,and grazing land were statistically significant, the arabilityindex was not fund tobe a significant factor. Instead,real debtload per hectarewas found to have a highpositive correlation to land prices. 5.27 Whereasland prices rose graduallyin real termsfrom 1970until the early 1980s,they generally declinedstrongly afterards (rableS.1). Realland prices In the summerrainfall region (column A) fell by 45% fromtheir peakbetween 1981 and 1990,in the cattlegrazing areas by 37% fromtheir peak in 1976to 1990(column C), and in sheepgrazing areas by 28% frm their peakin 1980(D).

Poold, provi.ncil MM40 data were usod fiom to rpeR i agriulur cuses of 976, 1981 and 1968.

Chprw5:S.,Fjcnkny OfAJ*Uk4( Produacon Page 98 Souh Afian Agrieghwru:Sowchu. erfm A Ooaw &Cr ONBro

Table 5.1: RealLand Pries. Real Cauital Formation.LonL-Term Debt. and Arrar.

ReadLand Prioe Indices Ral Capital Pomation, Real Value of Real Value land Bank: Paeent Fxed Maocinary& Moopago Loans for of Capil Owed in Improvements Implmeos Loamnto Land Ana SR WR ca sa (R million) (R milion) Famern Purchae t l Capita (R milion) (R million) (%)

A B C D P G H I J 1990 66 82 92 90 344.0 388.7 263.4 157.3 6.40 1.41 1989 68 78 86 88 372.7 524.2 322.8 206.3 5.75 1.47 1988 73 76 8S 87 345.9 536.6 297.5 198.5 S.4S 1.60 1987 79 87 78 89 360.6 410.5 299.8 178.5 5.62 1.63 1986 85 80 103 95 360.9 488.6 357.2 184.2 4.S5 1.22 1985 100 100 100 100 388.4 663.5 508.6 218.3 2.76 0.94 1984 114 97 100 107 414.2 737.9 656.6 245.1 1.94 0.95 1983 118 89 118 98 452.4 895.2 965.2 246.1 2.03 0.96 1982 121 76 116 84 582.8 1,131.9 266.3 140.1 1.26 0.92 1981 121 70 106 118 572.3 1,723.8 389.4 262.2 0.74 0.74 1980 105 59 113 125 568.5 1,294.4 365.5 265.5 1.14 1.06 1979 97 67 96 101 544.3 899.0 302.7 182.2 1.64 1.48 1978 100 77 103 102 556.3 961.8 231.5 143.7 1.64 1.48 1977 114 88 137 88 529.S 1,058.0 178.7 105.5 1.62 1.42 1976 103 83 145 91 581.9 1,056.7 204.4 125.0 1.49 1.27 1975 116 90 143 102 641.3 1,345.6 261.9 179.7 1.25 0.99 1974 105 8o 115 100 655.6 961.5 275.6 208.8 1.21 1.1S 1973 101 77 111 95 693.3 888.5 227.8 166.5 1.64 1.28 1972 103 78 102 104 645.0 846.0 176.6 113.8 1.8S 1.24 1971 107 85 85 111 571.6 920.9 313.8 163.3 1.92 1.12 1970 102 69 88 108 542.7 813.8 470.7 148.6 2.06 1.13

SR - Summer rain region;WR = winter ain region; CO = cattlegmzing region; and SO - sheep grazing region. lTe real prieindices forland iscalculted as the nominalland priceindexdivided by CPI and adjustedto makel985=100. The read value of fixed improvementsis cakulated as nominal expeadBitrs on fraxd improvementsdivided by the prie index for materialsfor fixed improvements"and adjused to consant 1985 Rand(1985 = 100) (Absmrctof Agriculturl St c 1992, p. 85, 97). Real fixed invedmet in tractors, machineyand implemenb equalsthe nominal investmen dividedby the price indexfor machiny and implementsand adjuted to equal 1985 constantRand (Abstract, p. 85,97). Percet of LAndDank long-termdebt in arrea is te from Table P.2, AppendixF.

5.28 Van Schalkwykand van Zyl (forthcoming)undertook the most comprehensivestudy of South African land prices to date, coverin the period between 1955 and 1990. The following conclusions emergefrom their work.

a. Real land prices were positivelyrelated to lagged inflation, suggestingthat investments in land were partialy made as a hedge against inflation;and

Cha S,Fcierwy: ofAgricadruralProducto pose 99 Sout*Afin Agrdur: Strtwe. Pffonrance& gdOon,for the ture b. real landprices rose as real interestrates decreased to negativelevels during the 1970s; correspondingly,real land pricesdecreased as interestrates increasedto more normal levelsin the 1980s;

c. real landprices followed the trends in real returnsper hectareclosely, except for most of the periodafter 1983when land prices and real retuns per hectarefollowed opposite trends;and d. rel land priceswere positivelyrelated to real debt per bectareexcept for the period between1980 and 1985,when real debt-loadrose whileland prices plummeted. 5.29 Van Schalkwykand van Zyl point to fact that as real Interestrates on creditfrom commercial sourcesincreased in the early 1980s,farmers responded by gainingincreased access to LandBank credit whichwas on soft termsand wasmanaged by their owncooperatives. Consequendy, during the period 1979-1985real debt per hectarewas allowedto increaserapidly as landprices fell and, as a result, solvencyrates declined precipitously. 5.30 On balance, it appearsthat prior to 1980land prices were mainlyinfluenced by the high profitabilityand the highly subsidizedinterest rates offeredto firmers. Ihe steadydecline in farm profitability,combined with higher real interestrates, are mainlyresponsible for the declinein real land pricesduring the 1980s.In termsof priceefficiency, these trends suggest that a largepart of the increase in real landprices from the 1960suntl the early 1980sreflected farmers' very reasonable reaction to the declinein the cost of capital. Thus farmerswere price-efficient as agriculturalentrepreneurs and as Investorsin rural real estate,taking advantage of the LandBank's preferential and subsidizedaccess to low-interestmortgages for farmers.However, during the 1980s,the decliningprofitability of the capital- inwensivefarming methods used by the farmersseems to havebeen compensatedby increasedaccess to softcredit from the LandBank which was managed by andextended through the agriculturacooperatives. In such a situationit is difficultto assessto whateatent farmerswere behavingin a price-efficient manner,given the increasedlikelihood for opportnmisticbehavior under a systemin whichdemand and supplyof creditare no longerseparable.

LaborCosts 5.31 Thesegmentation of labormarkets came about through a longseries of legislativeefforts similar to the legaldistortions introduced in the landand capital markets. It attemptedto solvewhat was refewred to as the laborproblem of the whitecommercial fuaming sector by meansother than simplyIncreasing wagesin orderto attractlaborers. The inability to usewages to attractblack labor reflects fundaenally on the unprofitabilityof large-scalesettler farming. In general,legislation enacted attempted to reduce competitionbetween employers by increasingthe transactioncosts of labor mobilitybetween the agriculturaland non-agricultural sectors, and between agricultura employers. For instance,the concepts of mastersand servantsin the commonlaw and a seriesof legislativeactions from 1911 onward (Cbapter 3) strengthenedthe abilityof white farms to retain farm workersand preventedfarm workersfrom breakingcontracts, changing employers or allowingfamily members to entercontracts with other potenW employers.The fact that the laborproblem continued to affectlarge-scale farming is demonstratedby the factthat until 1993,labor laws affecting other sectors of the economywere not applicablein agriculture.

mapr 5: fkuci of Agrkaral P°8oduco Palo 1zoJ Sogd Aftan Aeric*ru:Strucr . P,rfaR c A Obflonsfor g Futurg 5.32 Ihe economicImpact of the aboveand other legislation was extreme labor market aegmenaton. It ensureda spply of labor to whitefarmers at wageswell belowthose in the urbn sectors. White frmers basicadlybecame local monopsonists in the labor marketand were ableto set wagesfor black labor at very low levels. A substantialblack wage gap betweenmanufacturing and mining,on the one hand,and agriculture,on the other,was the resuIt4 Cable 5.2). For example,in 1975,it wasestimated that the per capitaincome of Africansin the four large metropolitanareas was 3 timesthat of Africans livingin the wbiterural areas (Wison and Ramphele,1989). In 1989,Afiican poverty-measured by the percentageof the Afiicanpopulation living below the povertyline-was also muchhigher in the white rural areas(well above 60%), than in the largemetropolitan areas (24.3%). 5.33 Suchwage gaps provided powertizl incentives for blacks to migrateillegally to urbanareas. Influx controlwas at best a sieve,but did littleto reducethe gapbetween rural and urbanwages. Conversely, illegalmigration of blacksto the urbanareas mayhave prevented the gp from growingeven firther, giventhat trendsIn averageannual earnings of all racesexhibit strong similarities across sectors.

Table 5.2: Annual Banins in the Fasn and Non-Farm Soctors at 1985 Constant Prices

Agrculte Manufactuing Mining Total Baks Othas Total Blacks Othens Total Blak Others 1988 793 - - 9,559 6,061 13,543 6,962 - - 1987 816 - - 9,215 5,664 13,294 6,577 - - 1986 821 - - 9,270 5,492 13,593 6,427 - - 1985 794 - - 9,644 5,607 14,252 6,595 - - 1984 - - - 9,950 5,829 14,728 6,7S7 4,628 22,614 1983 956 743 1,837 9,842 5,776 14,568 6,594 4,440 21,927 1982 - - - 9,636 5,637 14,898 6,606 4,386 22,652 1981 - - - 9,418 5,462 14,161 6,378 4,236 22,391 1980 965 - - 10,423 5,925 15,768 6,921 4,526 24,792 1975 707 524 1,609 8,289 4,347 13,163 5,405 3,024 23,752 1969 483 354 1,191 7,169 3,325 11669 3,264 1,157 20,292 196$ 579 434 1,365 6,453 3,069 10,472 2,906 1,101 17,075 1960 483 364 1,126 5,685 2,420 9,545 2,706 1,063 15,175 1957 415 326 941 5,481 2,333 9,259 2,733 1,089 14,578

Sou Roth et aL, 1992. Othes: Whites, Cloreds and Asins.

5.34 However,there is considerableevidence of a wideningof the blackwage differential between the

0 Some of the diffe_ence can be attributed to cost-of-liing differentials, and, in agriuu to the exclusion of inknd mancea that e difficult to compWltulycapture in salary or wae mesr. Howe, such adjutments aro unlikely to explain the full extent of the gap.

Chmar 5: FpciMf 4A8rlxural Prodon Page 101 South Afrcan Agriculture: Srunaure.PerfonMance A Ontiogsfor the Future farm andnon-farm sectors between 1972 and 1983(Simkins, 1990). From 1972 to 1983real wage levels in the miningsector rose rapidly (Table5.3), whilemanufacturing, starting fom a higherbase, came in a distantsecond. Uip to 1983,the real wagegains in agriculturewere considerably lower than in most othersectors. Wagetrends In agriculturewere roughly comparable to thosein the equallyless-organized tradingsec or. A growingwage gap between the farm and non-farm sectors has also been found by Merle UIpton(1985:388, table). Table5.2 showsthat the wageof blackfarm workersof R364 per annumin 1960was a mere 15%of that of blackworkers in manufauing and 34%of that of blackmine workers. By 1969,black farm workers earned 10% and 30% of whatblack workers were earning in manufacturing and miningrespectively, but by 1980only 9% and 14%of whatworkers in manufacturingand mining werereceiving. The situationhad not changedmuch at the end of the 1980s.

Table 5.3: Conmtative Sewotralnrease in Real Wages. 1972-1983

Sector Pere Incream in Real wages (%) Agriculture 26 Minig 254 Manufacuing 65 Tade 21 Govenmmt and private srvice 52

Soux: Sinkins, C., 1990, deived from an analysisof South African staistics, 1986.

mheeraordinary incrae in miningwages (254%) can be attributedto the vay lowbase fromwhich they strd, the gold pre spikein the lfte 1970s,unionizion and the decisionof the Chanber of Minesto aract more labor fromins SouthAfrica afte the withd of the very large MalawianMine Labor coningert by Bandain 1973.

Chaapr 5: EjIeiwy qf Agricniwal Producdon Page 102 Sowh Afric Aericultu: &Pwtu. Performarce & Ondn. for the FAture

Table S,4: AgricultuI Waso Income of Black Peopleon WhitwFarms. 1970-1987

Total Wages Wagespe Rgular and Paid to Fam DomestioWorwer Work (R/year) (RtmM=io) (198S constmt prices) 1987 811.0 1,017 1986 895.5 1,092 1985 924.2 1,120 1983 903.5 1,232 1981 878.8 1,276 1980 876.7 1,169 1979 905.2 1,169 1978 890.6 1,114 1976 86.2 1,076 1975 859.9 1,010 1974 800.2 927 1972 768.2 961 1971 736.0 876 1970 788.2 858

30ure: Simkhn, C., 1990, Table 3, deived from an analysisof Agriuul Census and Survys.

Note gI/:Wages must be interred with auton becaue of bigh paymens in kind. 5.35 The Increasein real wagesof blackfam laborersand of all farmlaborers (i.e., all races)seem to havebeen matched by a comparablerise in laborproductivity (of all racesof farm laborers)-atleast for the period for which data on black farm wages exist, i.e., 1957to 1988 (Cables5.2 and 5.4)51. 16 other words, real wage increasesfor black farm workersseem to have been justfied in terms of producvity Incrases. VanRooyen and van Zyl (1991)estmate that labor productivity has actally risen fasterthan real total remunerationper farmlaborer. Suchdiffreial trendsare probablythe resultof labormarket segmenaion, but also of the relativedecline of bargainingpower of blackfarm laborers, whoare excludedfrom the labor-relationslegislation governing other sectorsof the economy.

Therc is sme also evideca of rldieaion in relatv wag increases,at lt the pedod 1957to 1969,whe real wag of black fam laborersincreasd by a moderate8.6%, but real wags of aoh - increasedby 26.6lon

Chapw S: Fffl4eny qfA8rcu&al P*oda Page103 SoushAfrk= Agricuture: Strucre. PeM & Oons for X Fuure

Table S.S: Labor gMploymgnLWases and Productivity

Total Fam Ratio of Casual Productvity per Worker ORea Workers to Permanent Value OutputlWorker)(Randsr (000) Wo*ers 1988 1,144.8 0.74 11,722 1987 1,273.9 0.78 9,673 1986 1,270.6 0.73 9,351 1985 1,245.1 0.71 9,183 1983 1,OS6.6 0.72 9,999 1981 1,069.1 0.73 10,816 1980 1,152.8 0.72 10,311 1979 1,226.5 0.77 9,709 1978 1,196.9 0.76 9,013 1977 1,169.0 0.63 8,762 1976 1,184.3 0.73 8,684 1975 1,331.0 0.85 6,829 1974 1,344.9 0.89 6,842 1973 1,354.2 0.86 7,323 1972 1,389.3 0.89 6,099 1971 1.516.0 1.03 6,106 1969 1,609.0 0.94 4,591 1965 1,161.7 0.39 5,719 1964 1,122.1 0.48 5,869 1962 1,642.6 0.88 3,918 1961 1,S44.8 0.80 3,869 1960 1,348.3 0.78 4,376 1959 1,728.9 1.00 - 1958 1,605.S 0.97

Source: Roth et a., 1992

a. Grossvalue of agricktunl producion dividedby the producers' price index of agrikuluregoods (Abstrat of Agiculura Stastis) dvided by the total number of wors (manent and caual) employedin agriulure regardlss of race (South Africa Ststios 1990).

5.36 In spiteof the moderateand apparendylabor productivity-relatedwage trends, farmemployment declined in absolute numbers,starting in the early 1970s (Table5.5)52. his came at a time when the volume of field crop productiongrew at an annual rate of 4.7% and total agriculturalproduction grew at 3.9% over the period 1970-1980. Agricuture is the only sector in the economyto experiencesuch an absolutedecline in employmentduring a timeof output expansion,concurrent with growing numbersof unemployedblack labor in the wider economy.

2 Regnl dat on he mber of worker an farmsin the conmcial secw ar rnely limited, and effectly availale for only a few years. Dat ae someimes,but not always,desegegt into peranent and al worers within regions. Pairly compete dataon the numberof frm employeesar aivable by race and se (up to 1969)in te Absact of Agricull Sbtas, and by penmanentand c works for mostyes in the SouthAfican Stiics.

Crpsr 5: £ffickny ofAgricuural Produolon Page 104 Sgut an Awicubure: Stru . Perfomume & Opdtionsfor the Future 5.37 Additionally,the ratio of seasonalto permanentworkers has declinedsince 1970, from 1.03 in 1971to 0.74 by 1988. The sharpdecrease in casualworkers in 1964and 1965corresponds with the promulgationof the BantuLaws Amendment Act of 1964.Whether the number of casualworkers indeed actuallydeclined, or whetherfarmers stopped reporting them or startedcalling them permanentfor reportingpurposes would be difficultto ascertain.Marcus (1989) points to an increasingtrend of using womenand childrenas a temporarylabor force. 5.38 Labor productivity,defined as the real value of gross agricultu productionper worker (permanentand casual),rose steadilyfrom 1960through 1980 before levelling off in the 1980sC(able 5.5). From 1970to 1985,labor productivity increased by 5.2% per yearduring 1970-1985, for a toWal of 84% (vanZyl et al., 1987). 5.39 The stagnationin laborproductivity in the 1980sis matchedby a stagnationin real farmwages, as well as an economy-widestagnation and declinein the 1980s;average real annualearnings (average for all races)fell disproportionatelyin agriculture:from R965 to R793between 1980 and 1988. They fell onlyfrom R10,423to R9,559in the manufacturingsector and fromR6,921 to R6,962in mining. Gien the concurrentlong-term growth in unemployment,the increasedbargaining power of unionized blacklabor outside of agricultureseems to havelessened the effectsof the generaleconomic crisis on real wagesin the non-agriculturasectors. The effectsof extremelabor-market segmentation on the ability of farmworkers to organizeseems to haveaggravated the economiccrisis for them. As Table5.5 shows, the real wagein agriculturefell to below 1975levels by 1987. 5.40 However,there seemto be Importantregional and subsectoraldifferences with respect to wage trendsin agriculture.The devaluationof the Randfrom $1.29 in 1980to $0.46in 1985benefitted the sub-sectorswhich were export-oriented. In particular,deciduous fruit, mainly concentrated inthe Western Cape, increasedits profitabilityand technicalefficiency and translatedthis intohigher employment and higherwage rates. In Table5.6, employmentdifferences between regions in 1985and regionalwage trendsbetween 1980 and 1985are examined. The Western Cape (Region A) hasthe highestmalelfemale ratios, the highestpercentage of ful-time equivalentjobs, the highestfarm labor wages,the only substantiareal wageincrease, and the only increasein employmenbetween 1980 and 1985. On the other hand, the NorthernCapelKaroo (Region B), the OrangeFree State (RegionC) and the Western Transvaal(Region X), have aU experienced a drop in real wagesand a moderatedecline in employment between1980 and 1985.

Chapwr 5: Fcengyofdgncutwal Prodction Page105 SouthAfrcan Agricukure:Structur. Perfonnonce& Optionsfor theFsawr

Table 5.6: Wace Increasesand Variationsin Conbact Labor be Region

Region Males per 100 Ful-time 1985 Wage WageIncrease Employment fenals in equivalent(% Rate 1980-85 Increase 1985 of 1985 (Rl/year) (%) (%) popution) A 613 37.6 2,131 17 4 B 140 22.0 1,000 -23 -1 C 110 20.2 998 -15 -1 D lOS 22.7 1,227 1 -1 ED 121 - 1,303 0 - P 154 17.0 1,357 7 -2 G 124 21.9 942 S -4 J 168 12.8 1,056 -19 -11

Sources Simkins(1990), Agriculr Censues and surveys.

5.41 Labor marketsegmentaion lso caused an artificialscarcity of whitelaborers, thus increasingthe relative price of skilled, mainly white, labor. Moll (personal co cation) found that there were hardly any black farm managers in the Western Cape (wheat/sheepmixed farms), Northern OrangeFree State, and Western and SouthernTransvaal (maize farms). The tradition of black farm managers seems mainlyto exist in the deciduousfruit sector in the Cape, wherewhite owners often employblack foremen. Ihe reasons for this seem to be that the scope for the substitutionof machineryfor labor is much more limited in horticulture than in grain farming, that Cape land owners are more enlightened,and that the export orientationforces the subsectorto remam efficient.

Cot of Mehaab

5.42 The explamationfor the decline in farm employmentafter 1970 can be found in two related factors. First, the reative profitabilityof agriculturevis-a-vis the other sectors of the economystarted to decline, making it increasinglydifficult for farmers to competewith urban wages. Second,although the process of mechanizationstarted after the SecondWorld War, it was only after 1970that an absolute displacementof labor by capital took place. The various pieces of legislationseeking to prevent black farm labor from leaving the farm were accompaniedby a strong trend of labor substitutionthrough mechanization,which began to reduce farm labor demand after 1970.

5.43 After World War H commercialagriculture saw a rapid increasein mechanization. Two periods can be distnguished. First, until 1970, the increase in mechanizationled to area expansionand output growth. Labor followedthis expansionas a complementry factor. Second, after 1970 area expansion stagnatedand output continuedto grow, but labor was replacedby capital and farm employmentdeclined in absolute numbers. Ihere is considerableevidence to suggest that the second change was not price efficient, but led to under-use of labor and over-useof tractors and harvesters.

5.44 In 1981 the Land Bank's long-termreal interest rate, at a negative8.2%, encouragedfarmers to

Chaptr 5: Effcien of ACrutal ProdUCwon Page 106 Sow* AfricanArbwu: $&rucb. Prftmance & Odoas for the Fure investin durableassets (machiney and equipment).Tax provisionsallowing farms to writeoff 100% of capitalcosts againsttaxes In the year of purcae fiuter loweredthe cost of capital. Overvalued exchangerates fvored importsof machineryfor a sectorthat attemptedto maximizethe productionof nontradables.Apparently, the combinedeffect on the cost of capitalwas to lowerthe reladveprice of equipmentand machineryto suchan extentthat it becameprofitable to replacelow-wage farm labor with even cheapercapital. This is a remarkabletrend, given the alreadylow wagesof black labor In agricultureand the existenceof excesssupply of unskilledblack labor in the economyas a Whole." 5.45 Themechnization of laborin the maize-growingareas of WesternTransvaal is descrbedin detail by de Klerk(1985) for the period 1968-81(Table 5.7). Farm operationswhich used to be donealmost completelyby hand (e.g. planting, reaping,transporting, threshing, bagging, and gleaning)were increasinglymechanized' and methodsfor handling,transporting, and deliveringthe crop to the cooperativedepots became more mechanized as well. Sincethe early 1960scooperatives have invested millionsin bulkhandling and storage facilities. Bulk harvesting by combine,with transfer directly to a truck or trailerelimintes, muchseasonal labor. By 1968,55% of the maizecrop was alreadybeing deliveredin bulk, and by 1977virtualy the entirecrop wasreaching silos in the sameway (de KMerk 1985). Fertilizationand weedingwere also mechanized.Chemical weed control was not nearly as widespreadas mechanicalharvesting in 1968. Only 15% of the area plantedwith crops was being sprayedwith weed killers in that year. By 1981weed killers were being used on roughly95% of the total crop swrface(de Klerk, 1985). 5.46 Whenasked to explaintheir shiftto mehanizedharvesting, farmers mentioned the following factorsin descendingorder of importance:(1) labor is unavailable;(2) combine harvesting is quicer; (3) laboris unreliable;(4) combine harvesting is cheaper;(5) combine harvesting is easierto control;(6) bulk handlinghas madecombine harvesting easier; and (7) wageshave becometoo high (de KMerk,1985). Hence,the urban-ruralwage gap-reflecting the uncompetitivenessof white agriculture vis-a-vis the other sectorsof the economy-causeda laborshortage in whiteagriculture. When the costsof mechanzation wereartificially lowered, white farmers were given another solution for the laborproblem: substitAtion of mechanizion for workes.

Sif hol be noted,howeve, that effeadve labor at in agriculturemay hav finreasedduring the late I970 and 1980 beyondthe modesnominal was increasesbecaus of inang tensionsin laborreains andthe generalpolitcd ewionm

'" Plantingonce done by larg t of sea l wors is nowdone by farmes uing a fewwok wit tractor and plani eument. One manwveting begins,almost al fam i saona workes for rn, threshing,bagging, or deaningTwo harvesing methods am si pracied: (1) reaing by handsad treaing mechanically(hand harvediW; and (2) r and thesing with the samachin Untilthe 1950s,all maizewas reapedby hand by tms of works. C 1mbinveters,which ppearedin SouthJ-Aicain the 19, prented fiamerswith tie choiceof mchanizedharvaeting echiquesusingtratorpulledharveser. Self-propeed omb wae introducedin the 1960s.Even with mechanical harveting, seaonaiworkers are sil universallyhired for gleaning(collctg thesacks of headsleft behindby the combin). However,the c hasrelced mual reapig,tanp and eshing.

Chaur5: lFfemy ofAgricubwraPmducuio Page107 SoukhAfian Agrkilure: tructure. Prfonmunce & Onliousfor the Future

Table 5.7: Distibutionof HarvestingMe_hfods in the WesternTransvaal. 1968-1981

Hoivesdng Mechanized Hand and Meohnized by Hand Harestng Harvesting (%fans) (% tams) (% fams) 1968 81 16 3 1973 54 38 8 1977 11 81 6 1981 S 89 6

Sour: do Klerk,1983.

5.47 Mechanizationaffected farming systems in the various agro-climaticzones of South Africa in different ways: In the grain-pwroducingareas it is associasedwith a reductionin farm employmentafter 1970. Maize and wheat together account for about 60% of total cropped area. Payne.,van Zyl, and Sartorius von Bach (1990) examinedtrends In farm employmentby maize and wheat farms from 1945 to 1987 in the five major commercialgrain productionareas of SouthAfrica: North-WesternFree State, Western Transvaal, and the TransvaalHighveld in the summer rainfall areas, and Ruens and Swartland in the winter rainfall areas. They found that the number of employees per 1,000 hectares under cultivationrose significantlyin the summerrainfall regions from 1945to 1970, while it remainedconstant in the winter rainfall areas over the sameperiod. However, the numberof employeesdeclined over the period 1970-87 in all areas. 'he real value of total remuneration(cash plus kind) per farm laborer increased in all areas from 1945 and 1970; but although it also increasedin the summer rainfall areas from 1970 to 1987, it decreasedin the winter rainal areas.

5.48 In the summer rainfall areas (mainlymaize), a large expansionin cultivatedarea occurred from 1945 to 1970, mainly due to the substitutionof tractors for animal-poweredtechnology. This period witnessedan increasein the numberof farm employees(per 1000 hectaresunder cultivation)and in rea machineryinvestment (per hectare under cultivation),indicating a complementaryrelationship between machinery and labor. With higher production from area expansion and higher yields (from technical hange) came an increaseddemand for farm labor, especiallyfor harvesting. However, between 1970 and 1987the numberof employeesdeclined; and machineryinvestment increased, indicating a substitution of capital for labor in production. A rapid expansionof maize harvestng technologyafter 1970 also hastened a shift in the substitutionof capital for labor and a declinein employment(de KIerk, 1983). As indicated in Table 5.7 only 16% of farms in the Western Transvaal used some form of mechanized harvesting in 1968, but by 1977 81% of farmers were using it, and by 1981, 89%.

5.49 In the winter rainfal areas (mainlywheat) much of the rea expansionoccurred before 1945. From 1945 to 1970, the numberof employees(per 1000hectares cultivation) stayed fairly constant, while real investmentin machinery(per hectare under cultivation)increased significantly. Between 1970 and 1987, however, real investmentin machinerystayed constant;but the number of workers declined.

Price Efficiencyof Mechanizationafter 1970

5.50 South African agriculture makes sub-optimaluse of labor. Subsidizedlow interest rates and

Chqtr 5: 1F3icy of Agrkual Prodaon Page 108 Sg& Afrc Ariult Sture:Structre. Performan & Obionsfor theFure varioustux breals have encourged excessive substitution of capitalfor labor (Bell and Padayachee, 1984; Biggs,1982; de Xlerk, 1991). Additionally,labor-saving technology has beenextensively promoted by SouthAfrica's farm machimery industry and seemsalso to havebeen adopted by farmersin an effortto imitateNorth American and WestEuropean agricultue. Suchcapital/abor substitution has resultedIn decliningcapital efficiency (Volkkas Bank, 1991). 5.51 Consequentdy,meizion in SouthAfrica appears to haveexceeded the optimumlevel defined by the relationshipof relativefactor prices and the marginarate of substitutionbetween capital and labor. Overmechanization-useof machinerybeyond its economicoptimum-hs been widelydocumented for SouthAfrican agriculture (e.g. Brothertonand Groenewald,1982; van Rooyen, 1973;Hancke and Groenewald,1972). Severalstudies have found that the mostproduedve and profitablefarm enterprises had the least real investmentin machineryand equipmentper hectare(Viljoen and Groenewald,1977; lanse van Rensburgand Groenewald,1987). Ihe use of actors has been.so extensivethat their productivityappears to be zero (vanSchalkwyk and Groenewald,1992). 5.52 Simflarfindigs are reportedby van Zyl (Farmer'sWeekly, September 11 and 18, reportig on Hermanvan Sdhalkwyk'sM. Com.dissertation at the Universityof Pretoria)5. The studyfound that a straightorwardsubstitution of laborfor capitalwould lead to increasedprodution andthat labor was not onlyunder-used, but alsounderpaid. 5.53 Over-mechaizationwas foundin nearlyall regions. Exceptfor the arid Karooregion (which reliesheavily on irrigationfor the limitedplanting that takes place), the resultsindicated that tractors were appliedup to a point where their marginalproductivity has becomezero or, in the summerrainfall croppingregion and the grazing regions of Natal, negative. This suggestsconsiderable over- mehanizationfor largeparts of the countryas a resultof subsidizedlow interestrates. Moreover,two of the regionswith negative marginal productivity also had the highestrate of farminsolvency. 5.54 Laborwas underusedand underpaidin all regions. Its maginal productivitywas foundto be positivein all regions, suggestinggeneral under-use; and becausemargina-product values were constenty higherthan marginal costs, labor appears to besubstantialy underpaid-mostprobably because of the labor marketsegmentation achieved under influx control. 5.55 The switchto mn in maizeand wheatproduction during the 1970salso entaileda switchto monocropping;but large-scale,rainfed, monocropping of grains in the SouthAfrican agro- climateis high-riskproduction. A shift towardrainfed monocropped farming normally implies an increasedrisk exposurefor the producers;with price-efficient, risk-adverse farmers, such higherrisks shouldmean low adoptionrates. However,South African maize producers-and to a lesserextent wheat,

" Daft usd we fromthe Agitu nss ReponRs of 1976,1981, md 1988to whichreio Cobb-Dous podo funtdonswee ftt. The dqpndet vaiable uswdan approxi masure of productivity,ie. rgial ss fgrmincome deflaed by the inde of agiltural pwduc prie. Estatd coeffcent wee the used to evaluatewhth marginl valuoprodus equa margi co for th vu inputsused. Five homogeou region rc diatinguishe Natal; aunvn-infl cropping rgion, conistig mainlyof the mai triangle; Cape coastalareas; Kaoo-an arid to semiarid egion inuding the lagt past of the Cape Provinceand Waftn and Soutn Oange Free State; and the gazing rgion-4 Cape ano of the Onge Riva and Nothmu Trsval azringaeas.

MaWe l5:S., ckmy qfAriu&ra Prodwtion Pag 10 South AfricanAoricdu1e: S&te Prfonawc & Ontons for the Ftra havedeveloped a patten of o npecaonsbased on thegoverment's willingnessto guaranteeouW prices and provde financil assistanceduting drougt years. Indeed,farmers expect at if a droughtoccurs, thosewho are hit harst wll receivethe highestfinancial subsidies. Technically, insolvent fam ofte receiveloans at subsidizedneive real interestrates, whilesuccessu farmerspay commercalrts. In recentstudies, van Zyl (1989)and Brand et al. (1991)showed that a moreopen arcultur markeg systemwould cause maie productionto moveeast, towardplaces with higher and more stabler patterns. 5.56 Usingdata fromthe Wetem Tranvaal, lanse van Rensburgand Groenewald(1987) compared grainfarmers who did poorlywith those who did relaivelywell in the droughtyears following record maizeproduction in 1980/81and fund that those who did poorly had spent more than others on fertilizationper randof maizeeamings and had investedmore per hectarein machinery,equipment, and fixedimprovements. 5.57 In the policyenvioment of the 1970sand early 1980sthere waslittle incentivefor farmersto switchto lessrisky production patr. Again,although one expects fms toact rationally,i.e., price- efflcientiy,there is substantialevidence that manyof themdid not. Moreover,from an economy-wide perspectivethe shiftto high-risk,monocropped systems under rainfed conditions is probablyvery price- inefficient;and there are goodreasons to believethat this price inefficiencyhas contributedsubstantly to an acceleatingand unsustainablespiral of indebtedness. 5.58 The characerisis of financiallyInsolvent firmers confirm the inefficienciesassocited withthe over-investmentin capital-intensiveproduction prompted by subsidizedcredit Fullerand Darroch (1988) use discrminantanalysis to identifyfactors disingiin farmswith a high debtratio fromthose with a lowdebt ratio in the summercrop producing regions, i.e. the Transvaaland the OrangeFree State6. Their findings indicatethat higher debt levels were significantyaociated with seven factors-in descendingorder of importance: a. the ratio of medium-termto totalassets (.593), indicatng that high debtusers tendedto over-investin medium-term assets,mainly equipment and livestock; b. the subsidizedcredit to total liabilitiesratio (.380),indicafting farmerswith larger amountsof creditfrom the Land Banlc Agricultura Credit Board, and Cooperative Credit tendto be higherdebt users-reflective of the effectof interestrate subsidies; c. off-farmincome (-.374), implying that non-farmeaings help substitutefor credite'

Pmannesbhavingo above40% wea definedas high det users,while tbos bdow10% wemclassifed as lowdit uses Mh low det group _res hnmes with suffit liquidreserves to met cash flow shoxdalls. The hbi di groupdesbes farmuewbo fae protd finci andhay det pamnuftsch . TheSAAU sen out43,22 sv in b 1913 q _ediomah'a,11,080 of whichwere red. Th Fule ad Darroh sampleextrcced 937(447 low di and 490 hig del) summe cp producersfom this aW mpVec

Iws ngSave reationshp bas an impora aizig effect,and in;ae oppounis for di redbti thouh nonfrm ad puat-m actfties.

Chaps: Sfficenc faogrlubwa Prod.won Page110 M&nAfian AlrdcUEkn:Mrag P*MMnc & Ontlo for the Future d. the achievementof a standard10 education or higher(.334), implying that better educated farmerstend to acquirehigher debt loads; e. ageof 36 yearsor older (-.258),implying that older farmerstend to haveless debt; f. the profitabflityratio, definedas net fam incometo tot assets(-.245), indicatng t higherfarm incomes help substitutefor credit;and

g. the debt servicingratio, defined as interestplus capitalrepayments to gross farm income (-.229),implying that debtloads tend to declineas interestcosts become onerous, or as repaymentsreduce the debt load.

5.59 In summary,there is considerableevidence to support the hypothesisthat the reductionof farm employmentin the commercialsector during the 1970swas relativelyprice-inefficient, even froma narrow perspective. It was cetaly not price-efficientfrom an economy-wideperspective. The substitutionof capital for labor in agricultureseems to have been primarly inducedby extreme segmenat of land,labor, and capitalmarkets.

Rea= =nZ=Wd&M 5.60 In the secondhalf of the 1980s,the governmentwithdrew a nmmberof explicitand implicitprice subsides (vanZyl, 1991; Brand et al., 1991): a. Farmers were exposedto somewhathigher interestrates and considerablydevalued exhnge rates; b. Budgetaryalocations to farmersdeclined by morethan 50% in real termssince 1987; c. Thereal producerprice of importa commoditiessuch as maizeand wheatdeclined by 50% sincethe mid-198Qs; d. Controlledmakedng and processing systems were progressively deregulated since the late 1980s;and e. Favorabletaxation has been changed,reducing the implicitsubsidy. For instancethe writingoff of capitalpurchases in one year has beenchanged to hee years. 5.61 A relativelyprice-efficient reaction seems to havefollowed. The capital-intensityof production fell, and farm employmentinitialy rose steeply(Figure 5.1) in the late 1980s. Similarpolicy changes ffecteda fewother sectors of the economyas well,particularly small business. As a resuit,these sectors went througha period of adjustment,marked mainly by a rapid reductionin the capital-intensityof producton. Lookingat the sectoralcomposition of the disinvestmentthat took place, we see that agriclture led the way. Duringthe period1983-87, agriculture disinvested at an annualrate of 15.9%,

1a4o 5: Fjkkm7 of4A*rkar Produo Page 111 South fii*M Ar*uburg: Sruere. Perhbmace A Oeiion for theFuture significanly above the economy-widerate of 6.5% per year.58

5.62 The relative price-efficiencyof the farmers' response is also reflected In recent changes In cropping patterns. They have shifted away from low-profit maize and gmundnu to higher-value specialtycrops (horticulturalcrops and vegetables). Howcroft (1992) estimas that the adoptionof a free domesdc market for all cereals would bring reductions in the area under cultivationto maize and wheat of 13.4 and 19% respectively.

5.63 In horticulture, the price efficiencyof supply response to the devaluationsand the sanctionsof the 1980s are illustratedby the followingtrends:

a. an 81% growth of exports from 1981 to 1984, followed by three years of sanctions- related stagnation(the industryreports that perhapsone-quarter of overseasmarket sales were lost), and then by strong growth from 1987 to the present; and

b. a 161% rise in the volume of production for export from 1981 to 1992, comparedto a 64% rise in production for the domestic market (the latter possibly subject to under- recording).

5.64 Before the adjustmentpolicies of the 1980s were implemented,the price efficiencyof South African agriculturerelative to the U.S.A. was found to be low, in particular with respect to the flexibility to respond to changes in the prices of inputs such as machinery (van Zyl and Groenewald, 1988). However, in a later study of elasticitiesof substitutionbetween agriculturalinputs, Sartoriusvon Bach and van Zyl (1991) found that the price efficiencyof South African agriculturehas improved in recent years becauseof an increasein farmers' ability to substituteamong inputs in responseto changingmarket conditions. They attribute this progress to the easing of policies that restricted labor mobility and cotolled the prices of agriculturalcommodities. However, increasedpolitical risks may negatively itefere with the recent improvemens in price efficiencyof the sector.

Scae Efficiengy iSvidec

5.65 Agriculture is generallycharacteized by constant returns to scale, and the higher efficiencyof family labor compared to wage labor has often establishedan inverse relation between farm size and efficiencyin production (see text box). Internationalresearch has generallynot found evidence of any more than aminmum increase in retuns to scale for the cropsthat occupy more han90% of South Africa's cultivatedarea (i.e. maize, wheat, oats, sunflower, sugarcane, and groundnuts).5"

I Hover, do Kklk (1991)argues that some of the apparet redutionin capitaliety maybe a atat abeato causedby chaugs in valuationmethods.

I Economisof scae inprocess, notin produion per se, edst in sugarcane,while the coordinion problau of runmning sVWrmMi all year up to capacitymay spill ovwrinto the organition of producton.

0sqmpt 5: ffi*ncy of Agricurla Prodiucton Page 112 So&t*Afrian Avgricatre:&ruclre. Perfonnanc I ODSionsfor the Futre 5.66 bI South Africa farmsizes began to Increase in the 1950 andconl1 to Increaserutil te 1980M. After steadily Icesn nes aaiu. sn~ziewe~ until 1971, black farm employment began to i W c decline. Consequenly,it tohea i cmbaon wi ommoy obsed can be argued that scale iWdita gerts( ;a nc d efficiencies eared e aRer 1950,and Inparticular after ~vlns4...... tafrto auaict 1970, and were a main i1Q factor behind the steady $%Mh i ewdeais ae f gup(0to0h)te or1 decline of employment In aeu8lret~ i ags amsz rus(0 o$0h)I aoIyo agriculture. Hence,Sw agriculture was the only 1r iaa itsiie ui i rusiaihge rft e ntowdi major sector of the .a economythat experienced _i- X -- Th ein an absolute decline in i ao and mat s nearlydisappear . employmentbetween 1951 a-i md 198-despiethefact dwwgswr risin at a slower rate in agriculture = tha in other sectors.Ibis hMzg. eac.M1%,0lE history suggests that In e nty % d m i mo bovr . SouthAfrica, a numberof Interventionsin the markets =3 D for land, labor and capital W produced a structure of xveyitd ever*iy.prodcdiiceconom s inceives which induced rl t p ctain ts. the scale efficiencies, in ...... particularsince the 1970.l

5.67 From the beginning otflonre vei oefr 0 fhwrdspolaok,hoth of the cenuy unti therseda far edia f rion. 1950.the mnuberOf fiars and the tota area cultivated .. ... eycuQSS a atreKa!yu(99 dzwtaeta increaed, but the average . aimsem*t ayngeteivre farm size declined. After rdciiynpatclrare 19)fdstacetnfnnllmad 1950this trend is reversed; enerJfrsno ". and farm size grew _ r o r p an erst consistenty, acceleratingin thSds p oetat tqmlwetio sasoranaflt pr rted the 1970s before levelling t off in the late 1980s.. . Becaue the cultivated area

C7zepr 5: Fff1icn of grlcataral P1odu Page 113 &p A&=S Aoriutre: S&McturA. ane - OMlon fr the Future remainedthe same, the nmbr of farmsdeclined-from 116.848 uts in 1950to 62,084unit in 1990 (Abstr of Agrltural Statiscs, SADA,1993). Thepattern seem to continueunti the late 1980s, ithough the is some eviden of an increasingdifferentiaion in farm sizesbelow the 10O-ectae minimumwhich (in someareas) defines a farmin officialstatlstlcs.

5.68 Thereduction In the number of farming units shows an acceleratingtrend, beginning in 1950, whenthere were 116,848fIaming units (Hatngh, 1986.Tbis number declined at anaverage rate of 0.9% from1950 to 1960(1960: 105,859 units), by 1.4%in the 1960s(1971: 90,422 units), by 2.4% in the 1970s(1981: 64,430 unitS), and by 2.9% in thefirst half of the 1980s. Therewas a slightrise to 67,010unit by 1988,followed by a declineto 62,084by 1990(Abstract of AgriculturaStatistics 1992). Theprocess leading to fewerand largerfarms was apparentlynation-wide. All provincesewpriaenced substantialconsolidation of fiauland. From 1955to 1988, the numberof farmingunis in the OrangeFree Statefell 53%, in CapeProvince 41%, in Natal41%, and in Transvaal39 ton/hectare.

5.69 Averagefarm size increased from 738 ha/farm in 1953,to 867hecaes in 1960,to 988hectares in 1971,and to 1,339hectares in 1981,but declinedto 1,280ha/farm in 1988. TheIncrease in average farmsizes was a nadon-wideprocess, too. From1955 to 1988,average farm sizes by provinceincreased from1,284 to 2,663ha/farm in theCape province, 471 to 998 halfarmin theOrange Free Stae, 403 to 629ha/farm in Transvaal,and 390 to 609 ha/farmin Natal.As canbe seenfrom these data (whichdo notinclude the marketgardens around the majormetropolitan areas (Brand et al., 1992).the national averagefarm size bides significantregionrA variations. In 1988 the medianfarm size was about500 hectares,with farms in the high-potenti areas sigoificantlysmaller. Such qualificatonsshould not distract,however, from the fact that large-salefirms dominateSouth Afticar agriculture,and hatthe averagesize of thesefarms is extaordinaryby inteatonal standards(Table 5.8).

I Th daelin te numbarof frmg unitsmay msk n inceaseI the pption of sma fg units (main in pe-u areas). he o ofland ownersip seems to hav beenfared by ladtansfes fromth middlefarming cda Mny of fth farmersthat used to copise the middlefamng class keleft agrcul or expn dedr farm otha aL,

Cza$ 5: fficuny1 fAviula1 Port"cdos Page 114 Souh Afra Airluture: Structure. & donsLormancefor th£ Fdture

Table 5.8: Avrse Farm Sie ip thg World. 1970

Country or continent Averageopeaional fan size (ha/fanm) South Africa-total 988 Natal 523 Tranavasi 473 Orange Free State, 724 Cape 1,765 Africa (exl. South Africa) 2.9 North America 161 Europe 7.6 Latin Ameria 47 Asiat 2.3 World 14.1 hum: UN Food and Agicuur Oranizon (1981) Republicof South Africa (1993), Abstract of Agriuur Stawcs

3: Countries used in calculationsare the following. Africa: Agei, Angola, Botawana,Cameon, Cental AfricanRepublic, Chad, Congo, Gabon,Ghana, Mle d'lvoire, Kenya, Lstho, Liberia, Libya, Malawi, Reunion, Siea Leone, Swaziland, Tanzania. Togo Zaire, Zambia. North Amea: Canada, USA. Europe: Austria, Bem, Fane, West Germany,Italy, Malta, Netherlands,Norway, Poland, Porul Sweden,Unied IGngdom. Wodd: Also includes Eastern Europe and Oceania (not dsown). Oceania incldes r agi Australia (1,997 ha/fam) and New Zealand(302 ha/fm).

5.70 At present, there is mixed evidence for the existence of scale efficienciesin SoulthAfrican's commercialfarm sector:

a. Ihe dbton of ross farm income is highly unequal: In 1983, 0.9% of the farmers eared 15.9%of the total gross farm income; 5.8% earned 38%, and 27.5% eamed 72.8%;

b. Nleuwoudt(1990) cItes statistics indicatingthat 28% of SouthAfrica's farming sector supplies75% of its gross farm output;

c. Hattingb(1986 reports evidenceof a direct relationshipbetween farm size and efficiency in sheepfarming in the Karoo and in catle ranchingin northwesternTransvaal. Hattingh also reports that efficiencyincreased between smalland medium-sizedirrigated farms at Vaalharts and dryland grain farms in the OrangeFree State, before decreasingagain on the larger farms (size ranges are not specified);

Cspe :fArlukd5: F9rixny Prod*tiox Page 115 SoIahAfican ArFicturHe: Structure. Performance & Oz'donsfor theFut-r -

d. Analyzingthe Departmentof Agriculture'sProduction Cost Surveys, Moll (1988) finds no significant economies of size' both in maize-cattleregions (Western Transvaal, North-WestOrange Free State and the TransvaalHighveld) and in wheat-sheepregions (Swartland). Using retabulated1983 censusdata, however,Moll finds economiesof size but only in the maize areas and for 50-300 hectare range.

5.71 Conversely, there exists empiricalevidence from South Africa to suggest an inverse relation between farm size and efficiency. In 1988, 50% of farmingunits owning only 7% of the farmiand, with farm sizes of less than 500 hectares,were responsiblefor 43% of gross income,23% of net farm income, 36% of capital investment,and 33% of farm debt (Brandet al., 1992). Cbristodolouand Vink (1990), based on data obtainedfrom the Central StatisticalService which also covered the existingsmallholdings in municipalareas, come to the followingconclusions (Table 5.9):

a. the gross margin per hectare was R1514for small fanns (below500 ha), R87 for middle farms (500-1000hectares), and R36 for farms above 1000 hectares. Moreover, small farms employed632 workers per 1,000 hectares, comparedto 27 and 29 workers per 1,000 hectares for middle and large farms respectively;

b. the cash wage per 1,000 hectarespaid by smallfarms was on averageR1,189, compared to 13 cents and 7 cents paid by middleand large farms respectively;

C. farmworkersearned 16% of the gross incomeof small farms, but only 10 and 9% of the gross incomeof middle and larger farms;

d. smallerfarms' total farm expenditureswere nearly R5,000 per hectare, whereasmiddle and larger farms spent only R177 and R55 per hectare respectively.

nasMo lus oononie of i (all fadorsbut operaor labor obangin) a opposetto eononme of sae (all f4on Ow 5P

Ch* 5.- 1kqf OAscuklatrodu& PoJClAage8116 SouthAfrian A-ricubure: Structure. Performance& Optionsfor the Futur-

Table 5.9: Factor Intensites in Aniculturem1988

Fauming Gross Employees Cash wage Wages as Curent unit size Mar8in per worker % of gross ependkures groups income (ha) (R/ha) (No.11000ha) (R/1000 ha) (%) (Rihb) < 2 5,096.77 2,779.6 10,534.31 16.4 28,210 4 3,421.24 1,082.6 469.13 17.1 8,160 9 2,517.84 673.5 84.56 15.3 3,133 19 614.38 379.7 25.78 21.4 1,283 49 986.41 335.3 6.49 14.0 1,942 99 384.23 172.9 2.94 16.8 840 199 344.41 133.6 1.35 17.3 680 299 170.83 84.9 0.99 16.4 424 49g 91.45 45.8 0.35 12.5 277 999 87.55 27.1 0.13 10.2 177 1,999 74.55 17.7 0.08 10.1 117 4,999 38.46 7.8 0.05 9.6 56 9,999 18.06 3.0 0.07 8.6 25 10,000+ 11.71 1.4 0.11 8.6 21 Total: 50.11 14.3 0.01 11.6 91

Source Christodoulouand V=k, 1990, based on data provided by the CentmalStaistical Service.

5.72 Much of the evidenceon scae efficienciesin South Afiican agriculture cited above should be interpretedwith etreme care becausedata were not standardizedfor differencesin land qualityor for the paricular commoditymix of farms, and in some cases inappropriateanalytical methods were followed. In order to address these deficiencies,Chavas and van Zyl (1993) used non-parametricmelhodsQ to detmine scale efficienciesin sevenmajor farmingregions of SouthAfrica-Swartland, Rfiens, Transvaal Highveld, Western Transvaal, Eastern Orange Free State, Northwestern Orange Free State, and the Vaalharts irrigation area. Individual farm data for the period 1976 to 1989 were supplied by the Departmentof AgriculturalDevelopment's Directorate of AgriculturalEconomics.

5.73 A wide range of farm sizes in both extensiveand intensivecommercial farming seemto be scale efficient, dependingon how farmersorganize their specificvariable and fixed input mix, as well as the combinationof outputs they produce. In this respect, many relatively small farms in both rainfed and irrigation agriculture are scale efficient,but a rather larger numberof other relativelysmall farms seem to be scale-inefficient. Giventhe availabledata, Chavasand van Zyl (1993) could not investigatescale efficiency on farms below 10 hectares in the irrigated area (Vaalharts)and below approximately100 hectares in the other areas.

5.74 The comparativeefficiency of black, small-scalefarming versus white, large-scalefarming is very difficult to assess. More than a century of policy interventionshas suppressedthe profitabilityof black farming in order to protect white farmers from black competitionand to assure the white farm sector of

TMemmetbodolo used is desrib in del by Chas and Alibe (1993).

Ch* S.: 1ffldwyqf$JoAgricia Production Page 117 Soth Aian A,atur: Srutur. PEfma A ODtionsfOr e Futre low-wagelabor. Ite only areaswhere black farming was condoned were tihe homelands. Given their location,lack of infrastructure,and supportservices, generally poor soils, and extremepopulation pressure,It wouldbe unfairto comparesmall-scale farming in the homelandswith farming in the white ares. 5.75 Nonetheless,a few casesexist in whichblack, small-scale farmers were given access to support structres roughlycomparable to thoseof their whitecolleagues. The two casestudies preseated here, in bothof whichsmall-scale, black farmers equalled or outperformedlarger, white farms, come from tea and sugarfarming (van Zyl and Vink, 1992). 5.76 In the tea industrythe casestudy illustratesthat 'mini-farmingo(where an individualleases a smallarea planted to tea froma tea estateand is remuneratedaccording to the quantityof acceptabletea produced)shows an increaseIn yields,income, and profitabilityof boththe estaterenting out the land andthe mini-farmers.Compared to ordinarypluckers, mini-farmers obtained yields on their 0.5hectares plots averaging23% moreta whatthe large estateobtained. 5.77 Thesame applies to the sugar-canecase study in the EasternTransvaal, where black smaliholders obtained116.8 tons/hectare on their plotsof 7.1 hectares(on average),while largescale whitefarmers adjacentto these smallholdersobtained 102.9 tons/hectare on 68.6 hectares(on average). Total costs amountedto R3,286/hectarefor the smallholdersand R 3,448Jhectarefor the large-scalefarmers (unpublisheddata fromrepresive samplesgathered by the Universityof Pretoria). Boththese case studiesconfirm that withthe samesupport structures small-scale farming is at leastas effcient as Ilrge- sce firmingin these specificareas and typesof farming. Land Tnsactiom 5.78 Landtransactions constitute an importantmechanism to capturescale efficiencies.Since 1965, between7,561 and 14,889deeds, and between3.1 millionto 5.5 millionhectares of rural immovable property, have been transferredannually CTable 5.10). The total area of transfershas remained remarkablyconstant at around 4% of the totalsurface area in the commercialsector. 5.79 The averagesize of landtransfers has risenover time. The numberof transfersdropped in the 1980s,both nationally and for mostsize categories; but theredoes not appearto be a correspondingdrop in the totalarea transferred. Particularlyin the uppersize ranges of the marketfor nrualland, transfers have remainedrelatively constant in mmber whileareas transferred have increased. 5.80 Transactionsinvolving smaller parcels dominated: Of the 8,852parcels transferred in 1990/91, 26.8% were less than 19 hectaresin size, 26.7% between20-99 hectares, 17.9%between 100-299 hectares,10.4% between 300-499 hectares, 9.4% between500-999 hectares, 5.1% between 1000-1999 hectares,and 3.6% above2000 hectares. However,the frequencyof recordeddeed transfersof the smallestparcels is decliningin relativeterms, which may suggest that progressively more transfers are takdngplace off the record.

0 Note at 198142wu thebeat maim amoer.

Cuapff S: Xfficlencyoj4Arlcukaral hoduction Ave 118 Sh Afra Aarksbr: Stuwr. Per e A Onl= fir the Fun" 5.81 rans ivolving largerparcbls, however dominaed the totalarea nsferred. Of the 3.2 millionhectares of landtransferred in 1990191,0.6% were parcels less than 19 hectaresin size, 3.2% between20-99 hectares, 9.2% between100-299 hectares, 11.2%between 300-4Q9 hectares, 18.1% between500-999 hectares, 19.5% between 1000-1999 hectares, and 38.3% above2000 hectares.

Tible 5.10: Lad Madk Transacmionsand Prices. ebig of South Afnica

Toal No. of Tota Armaof Ave. Sizn of Percn of Ave. Land Transfers Transfers Pacels Farm Arma Priceha (No.) ('000 ha) Tranfed (%) (R/ha) (ha/pacel)

1990-91 8,852 3,235.2 365.5 - 798.1 1989-90 10,146 4,159.4 410.0 - 614.4 1988-89 10,132 4,206.3 415.1 4.90 496.5 1987-88 8,436 3,470.4 411.4 4.04 427.8 1986-87 7,561 3,166.9 418.8 3.69 375.0 198S-86 7,818 3,079.7 393.9 3.57 364.4 1984-85 9,437 3,S43.4 375.5 - 361.6 1983-84 9,642 3,477.6 360.7 4.04 336.7 1982-83 9,414 3,453.3 366.8 - 258.6 19814-2 13,544 5,269.2 389.0 6.10 237.1 1980-81 14,889 5,519.8 370.7 - 199.7 1979-80 10,406 3,924.1 377.1 - 145.7 1-97576 10,857 3,297.8 303.7 3.86 136.0 1970-71 13,123 3,482.9 26S.4 - 82.0 1965-66 11,399 3,599.1 315.7 4.10 47.6

8knnxx Roth t aL, 1992

5.82 Anotherland trmsactdon facilitaftig the capturingof economiesof scaleIs rental. In 1963total leasedland represeted only13.1% of tot landarea; but in 1988rented, leased, and share-cr ed land represented19.5% of the total surfacearea, wit considerableregional variation: 26.9% in the Orange FreeState, 22.9% In Tranvaa, 17.3%In the Cape,and 15.7%in Natal. Haingh and Herzberg(1980) found a fthosewbo lease land are mainlyfarmers who already own some. 5.83 Moreover,although the officialstadstics point to a relativelyhigh rent rate of nearly20% of totalarea, In fact mostrentals are betweenthe older andyounger generations of the samewhite family. Suchrenta aangements are de factopension schemes, and the proportionof genuinerentals can be as lowas 5%. It has beensgested tha the lowrate of genuinerentals at leastparty reflectsowners' fear that renterswill mine and destroythe fragileland. u of ScaleEffice

5.84 Ihe offici definitionof a viablefarm in termsof sizehas had a profoundnegative effect on the

Chqpw 5: Ffi*as of Agrilwrl fPocun Page 119 SPueadric ALrftu Stc, rfomn & O9t2onsfor theFeure relative profitabilityof farms smallerthan the viable size. Giventhe high levels of official assistanceand subsidiesto farmers, the viabilitydefinition became almosta self-fulfillingprophecy, because under the AgriculturalCredit Act all farms belowthe viable size were excludedfrom assistance. Moreover,under the Subdivisionof AgriculturalLand Act of 1970 it is not possibleto subdivide an existing title deed without minista approval. Permissionis granted only with proof that a reasonablenet farm income can be obtained with 'average' management The subjectivityof this requrement, together with the lending cteria of the official funding agencies,precludes systematicempirical analysis of smal farms in South Africa. Yet it is interestingto observe that despite the lack of assistancefor small farners, official record deed transfers show that the prices of small parcels of land increasedmore rapidly than the prices of large parcels since the 1960s.

5.85 Ironically, the benchmark for determining farm viability-farm size-has changed over time; during the 1960s and 1970s, expansion and mechanizationwere considered the solution to remain competitivewith non-farm incomes. However, in the 1980s, the high debt loads from capital and land purchasesreduced farm profitabilityand decreasedreturns to capital-intensiveinvestment. Thus many farns once thought to be viable by the criteria set in the 1970swere exposedas unviable in the financial crisis of the 1980s.

5.86 Farmersthemselves seem to view consolidationof farms as a rationaleconomic reaction captuning economiesof scale. For instance, Moll (1988) reports that of 55 farmers surveyed in Bredasdorpand Malmesburyregions who had bought land during the previous decade, 35 (or 64%) indicated that they had done so partly to take advantageof size economies.

5.87 De Klerk (1991) attributes the process of farm consolidation to technical change, viz. mechanization.Cw.nsolidation has generallyalso causeda reductionin farm employment,because the new mechanizedfarm did not need to employ all of the workers from the more labor-intensivesmaller farms that were acquired. While seasonalworkers bore the brunt of mnization, permanentworkers were most directly affectedby consolidation(de Klerk, 1985).

5.88 Sartoriusvon Bach, van Zyl, and Koch (1992) constructedan index of manageria ability based on indicatorssuch as budgetingand the keeping of records and found it to be highly correlatedwith both farm size and total farm income. By evaluatingCobb-Douglas production function coefficientswith the manageral ability index includedas an input, the authors found significantlyincreasing returns to scale among 34 farmersin Vaalhartslrrigation Area. When managerialability is excludedfrom the regression, however, results indicate constant returns to size. The same pattern holds true for a sample of 100 farmers in NortheasternOrange Free State.

5.89 Groenewald(1991) suggeststhat even beyond the indivisibilityof capital and managerialinputs, economies of scale may result from scale efficienciesinduced by the existing agricultural marketing systemthrough volume discountson the purchaseof inputs and volumepremiums on the sale of outputs.

5.90 Roth et al. (1992) econometricaly tested a nmmberof models explainingthe reduction in the number of farms between 1972 and 1988. Ihey found the number of farms to be positivelycorrelated with the ratio of real machinerycosts to real gross revenue, but negativelycorrelated with the ratio of farm requisites (mainly non-labor inputs) to output prices. Ihis suggests that scale efficiencies in

Ckpw S: Ficiew ofAgrkubwalProducion Page 120 SAuh Afcan A ft*uure:Strucurw. Perfoimne & Ontionsfor the Juure agrclture are strongly associted with a decline in machinerycost and an increase in the profitabilityof non-labor Inputs. Both correlationssuggest that the appearanceof scale efficienciesin South African agriculture is rooted in the policy distortions that led to the reductionof the real cost of capital in the agricuituralsector.

5.91 Chavas and van Zyl (1993) found a highly significant negative correlation between farm size efficienciesand debt burden, while size efficiencyand managerialability were positivelycorrelated. The results show that the issue of scale efficiencyis a complexone and is influencedby a variety of factors, of which managerialability-the basic indivisibleinput in agriculture-seemsto be dominant. A whole range of farm sizes in both extensiveand intensivecommercial farming was found to be scale efficient, dependingon how farmers organizetheir specificvariable and fixed input mix, as well as the combinaion of outputsthey produce. Their results are consistentwith the findingsof Sartoriusvon Bach and van Zyl (1992), who concludethat beter managershave larger farms. It shouldbe noted, however, that small frms will in general require less sophisticatedmanagement than large farms, which would explainwhy Chavas and van Zyl found efficientfarms in all size categories. On the other hand, these results can be interpreted to mean that farm size is not really the central, but rather managerialability.

Future Scale Efficiencv

5.92 The contiued removalsince the mid-1980sof the distortionsbrought about by price policiesand legal constaints is likly to bring about a more normal pattern of scale efficiency in agricultura production. Given inteional experience,an inverse relationbetween farm size and efficiencyshould slowlyemerge; but the current ownership-structureis likelyto act as a constrainingfactor for the efficient restructung of farm production. The structuruabias againstsmall-scale faming (blackand white)carries high opportity costs in terms of productivityand employmentforegone.

5.93 In 1991the racially-biasedland acts-responsiblefor a centuryof segmentedland-markets-were repealed. Becwse of imperfectionsin the credit market, however, the entry of new African farmers tirough self-financedland purchases wil be limited because of their obvious poverty. Conversely, increasedprofitability of small-scalefarming as a result of the removalof policy distortionsmay create an agrarianstructur chacterzed by widespreadtenancy-as exsed before the adoption of mechanized fiming methods. Such a return to tenancy, however, would considerablyreduce investmentincentives to black farmers, both in fixed farm capital as weU as in residential housing and other infrastructure. Moreover, given South Africa's fragile soils, such a strategy entails significantenvironmental risk.

5.94 Althoughhigher capitalcosts makecapital-intensive farming less attractiveinthe intermediaterun, the changingpolitical landscapemay make it difficult for farmersto return to the labor-hiringpractices that supportedthe sector for the two cenumiesbefore mechani Tionhus it may well be that the exodus of workers away from farms wil be difficult to reverse completelyunder the current agrarianstructre.

ChperS:Sffcetfnc qfAgrcukurlProdaction Page 121 SouahAfm A,raulre: Strucurw. Perfomrance& ODdonsfior the Future

TBal Factor Produgti5djy 5.95 Thirtle,Sartorius von Bach, and van Zyl (1993)undertook the mostrecent and comprehensive analysis of total factor productivitygrowth of South African agriculture.Indices of totl factor productivity(TFP) measure aggregate output per unitof aggregaeinput, providing an indicationof the technicalefficiency of agrictural production.Thirde, Sartorius von Bach, and van Zyl (1993)calculated a TFP indexfor SouthAfrican agriculture for the period 1947-91"(rable 5.11). 'here havebeen no previousattempts to measureTFP at the national aggregatelevel; but regionalproductivity was econometricallyestimated by van Schalkwykand Groenewald(1992), and productivity of grain-farmin wascalculated by van Zyl (1990). 5.96 From 1947to 1991total factorproductivity (TFP) increased by 1.26%per year-a modestbut significantrate by internationalstandards. There was basically no TFP growthuntil 1965,then 2.15% until 1981and fairlyrapid growth of 2.88%per aunnmsince 1981. TheTFP growthfor the 1981-1991 periodwas the resultof policyreforms that increasedthe costof capitalrelative to laborand sloweddown non-laborinput growth considerably.

5.97 However,aggregate TFP growthmasks subsial differencesin productivitygrowth among the threeagriculturd sub-sectors. The export-oriented and inherentiylabor-intensive horticultural sector grewat 2.42%annually over the entire1947-1991 period, whereas the moreinward-looking livestock and fieldcropsectors both recorded a verylow productivitygrowth of only 0.77%per year over the same period.

'The methodologyfoMlowed is prhaps closet to that of BaU(198S) and Thirle and Botomley(1992). Thirde o aL (1993) expla change in the TFP indexthrough the dermining variables,such as the dsockof knowlede (accumlted resarch caiL gSwAred by pAs reeach eqendie), extensionserioes and fanmr education. 7he basc argumet as that resarch and drdsvlpnamigenerats technologa,extensio difuse t and that better educatedfarmers are better at screenin new technology, so thy adopt fa nd alo adapttechnolg, so addn an emet of on-fum tacolog geradion. But, for SouthAfrica, intnaonl echnolog trnwsfermay alsobe smtat and the influenceof the weatheris coniderable

C1qgr 5: fcien of4grlcubuwalPrducon Page 122 Sogth Afiea A.rfcuwkr: Sbuetu. Prfomawe A Oniom fo the hiure

Tabl. 5.11: Outnut.Inmfl. TPP. Labor and Land Ptoduc6it Indicos

Yeaw Oulp Input TFP Labor Prod Land Prod 47148 100 100 100 100 100 48/49 110 102 108 108 108 49/50 104 118 88 100 103 SOl/5 114 108 106 109 114 S1/52 121 102 118 113 120 52VS3 112 112 100 104 111 S3154 129 123 10S 11? 126 S4JSS 134 129 104 120 132 51S56 13S 13S 100 118 133 56/57 141 139 102 123 139 S7/S8 138 142 97 124 13S S81S9 143 154 93 126 139 S9/60 150 18 9S 125 139 6061 161 155 104 159 151 - /62 172 166 104 1SO 164 62/63 178 163 109 l1O 170 63/64 172 165 104 161 166 64/65 179 176 102 174 169 65/66 183 185 99 179 180 66/67 227 186 122 201 226 67/68 201 182 110 178 201 68/69 208 18 11S 182 204 69/70 220 1 11tS 202 213 70171 227 407 110 217 220 71/72 257 211 122 257 249 72173 243 205 118 243 239 73/74 252 203 124 25S 247 747 272 199 137 281 273 7576 265 205 29 273 268 76M7 279 211 132 302 281 77/78 299 213 140 310 302 78/9 299 226 132 319 303 79/80 311 231 135 326 313 SQ1 317 226 140 3S2 318 S1/82 343 228 151 411 343 82/83 307 224 137 370 308 83/84 276 234 118 334 277 W8S 310 242 128 34S 311

£7qte 5. Fjckew ofAgHukal Producto Page 12S Sou&hAfricax Agriclrture: Srudcture.Perbnnon-e & Ornionsfor the FAaure

8S/86 315 231 136 322 315 8W87 319 218 146 323 321 87188 336 214 157 344 338 W89 360 210 171 401 362 89190 370 206 180 418 380 90/91 360 215 168 411 375 91/92 342 213 161 409 360 sogc: rniztle e aL (1993) 5.98 Outputgrowth has been consistently high, with the exceptionof the 1983and 1991drought years (Figure5.4). Overthe entireperiod 1947-91, the outputindex grew by nearly350%-a rate of 3% per annum. The quatity index of inpUtSmore than doubled,growing at 1.8% a year. However,the aggregatehides the fact that inputsgrew at over 2.5% per year until 1979,but sincethen havebeen fallingat 0.9% per year Crable5.12). 'he combinationof outputgrowth and inputreduction explains the growthof TFP overthe past decade.

Table 5.12: AnnualAverame Growth Rates by Pefiod. 1947-1991 (5)

Otput Inplt lTPP Labor Land 1947-91 1947-91 1947-91 1947-91 1947-91 3.02 1.79 1.26 3.60 3.13 1947-79 1947-6S 2.52 0.00 196S-81 2.15 197941 1981-91 -0.90 2.88

:W, Thte et al (1993)

5.99 Overthe 1947-1991period, the part productivityindex for labor grew at 3.6%a year,and land productity increasedat 3.13% per amnum. Labor and land productivityincreases confirm dth productivit growth with respectto the primaryfactors is higher than TFP growthbecause of the subttui of non-farminputs for labor and land (Groenewald,1964). This result canbe seenmost dearly in FigureS.S.

Chaer: SFJclkmy qfAgrclua P)odcto Page124 RnoahA**n Anrlcura: S&,L. P.tMee lbgft sh *h

Fim5.4 lnal E=c ftrviq. Oupu and lnM (1947-1991

380

------OlflpU 300------. . . I260 ------...... 240......

~200------.-

1.160- ~~~~~~~~~~~~~~~~E I------

5140----- ...... 120 ------:: 4749 5153 55 5759 61 63 6567 69 71 7375 77 79 81 83 8587 89 91 Yw 1mre: Thirdeet al., 1993

Ci'ye 5:Fjk*ea ofAS:wI Pr.dacte Page.125 SO.1k AgMen Aerbgd , StUNM. Psdmaw A Onhiowlfr sh FUtUrw Figure5.5: TFP. Laborand LandProduciiy, (1947-1991)

450

~~~4O0 ~~~~~~~------

U350400 ...... ------...... --- - ......

11352502t00 ------0 ----"--- .,...... Pro4&3cY...... 4..

Fe1005f '.. --- -E---...... -[-'''' ..

474951 53 555759 61 6365 67 6971 73757 79 81 8385878991 You

~: Ibintle et al., 1993

5.100 OuW growthIn horticture outpacedboth fielderops and livestock (rable S.13 andFigure 5.6). Livestockoutput showed the least growth. For 1947-1991,growth rates for crops, hordcultureand livestockwere 3.06%, 4.20%, and 2.39% a year,respectively. The increasing importance ofhorticulure is not a recentphenomenon. Groenewad (1965) noted a similartrend for the period194516 to 1962/3. heeffect of seriousdroughts, as in 1983,are clearlyshown in the fieldcropindex; irrigated horticulture Is less affected;and the livestockindex declines,but with a lag-since the damageto herds hits povky aer.

(kW4r S: Fj7ka.7y of4VrlcWal -1 rotduoa Page126 SouthAftan A.rLuburk: Sictur. Perbmnancce& Options for the Future

Table S.13: AvoraggShares in Rovenueand Costsand Annual GrowthRates. 1947-1991 (%) item Shan Chang Growth SpecificSub-Poriods of Interest

OUTPUTS: Crops 42 40 to 3S 3.06 199011used for Share Change Hordnlturo 16 16 to 22 4.20 Lvestock 42 44to 43 2.39 INPUTS: Labor 23 36to 15 -0.58 1947-S9= 1.30, 1962-80= -1.18 Land 9 7 to 9 -0.10 1947-59= 0.30, 1959-91= -0.2S InTRMIEDIATE 33 28 to 33 4.20 1947-78= S.30, 1978-91= -0.19 Pacing 13 16o0 10 3.02 Fuel 26 33 to 2S 2.09 Fetlizcr 28 24 to 20 4.73 1947-79= 7.73, 1979-91= -4.29 Dips & Spmys 10 3 to 16 10.02 1947-72= 9.74, 72-80= 20.56, 80-91= *1.83 Other 24 23 to 29 3.94 CAPITAL 3S 28 to 3S 1.67 1947-83= 2.S3, 1983-91= -3.79 Imrovents 1S 22 to 10 -0.12 1947-70= 4.82, 1970-91- -5.47 Machiney 47 50 to 38 1.10 1947-57= 7.S7, 57-81= 0.76, 81-91= -4.56 Animals 39 28 to S2 3.10 thnpe: Thirle es aL (1993)

S.101 Table 5.13 also reportsthe averageshares in total costs of the inputs, the change in cost shares over the full period, and the averageanSal growth rates. Ihe overall position is clear in Figure 5.7, which comparesfour aggregateinputs-land, labor, capital, and intermediateinputs-with the total input series, showsthat intermediateinputs and capital inputshave been substitutedfor the primary inputs.The cost share of labor in productiontumbled from 36% to 15%, and the growth rate of labor over the period was a negative0.58% per annum. Land has held its cost share, but shows no quantitygrowth. Thus, virtually all of the totalinput growth is accountedfor by the rapid increasein intermediateinputs (4.2% per amnum)and capitalgoods (1.67%).

S Now th te shar of labor,land. hitermedi and capital ips sum to lO0%. Within th itmeia and capita ategoris, the shaes of inputsagain awm to 100%.

Caqpter5: £ffieny of Agricubral Production Page 127 SoWhhAenh APVkEb S&nwluw Pemarow# Ornou for the F,urv

Figu 5.6: ZgruWture ut (1947-1991)

650 C55 ...... ,...... e 550 , ------...... I ----

X3504250 ------o* ..... -----

200*-oo ------f.a 150 ---- ...... 100 ...... -.

4749 51 3 557 59 61 63 6567 69 71 798183 8587 89 91 Yew

=Z: IhkTWeet at., 1993 Figure5.7: Agrcultre Iuts (1947-1991)

550

n450 -

i ------......

00S------...... '' 'l-- 2.002 ------SO - - 150 f - ..

4749515 3 55 57 59 61 6365 676971 73757779 81 8385 878991 Yew

Soaurce: lnjl e, et at., IMnt2

Crg S. 15ici ef ASkawa Pmdd ps 126 Anddfic An*abtue IMM PafmmAngWA Oviltns fr the Frturte

5.102 This general trend obscures the period-by-perioddifferences. Intermediateinputs grew at over 5% per annum before 1978, but declined at 0.19% a year since then; capital inputs increasedat 2.53% p.a. unti 1983 and since then have decreasedat 3.79% a year.

5.103 lhe growth rate In TFP is greater than would be expectedon the basis of van Zyl's (1990) prelimy study of productivityin grain production;this shows the Importanceof horticultre to the gicultural sector. The rapid growth of productivity since 1983 is In agreement with the regional econometricstudy by van Schalkwykand Groenewald(1992), whichfound evidenceof substantialgrowth In some regions since 1981, reflecting the increasing competitivepressures and the removal of price distortions caused by credk, tax and macroeconomicpolicies.

EfficiencyIssues in the Homelands

S.104 The poor natural resource base and the continuousbuild up of demograhic pressure since the begiming of this century tirough the racially segmentedland markets have made the homelands "fnetionally urban" or rural dormitories". Giventhese conditions,future migrantsfrom rural to urban areas are expectedto comefrom the homelands;but even steady out-migrationfrom these areas will not exceedpopuion growth (Urban FoundationReport, 1990).

5.105 It is difficult to get a grip on efficiencyissues in the homelands;the existingdata base is very weak and probably underestimatesthe importanceof agriculture, even thoughit is extremelyconstrained by the overcrowdedand poor resource base. For example, official estimates of homeland yields are consistendylower than case study data (e.g., Cairns and Lee, 1990:85,91).

5.106 However, the availablecase-study material does suggest that relativelyfew farmers are engaged in agculture full-timein the homelands. Accordingto Nicholsonand Bembridge(1991), the majority of households in SouthAfrica's homelandsdoes not have enoughland to provide for subsistenceneeds. Thus most nrur households enage in farming only part-time, and most of their output is kept for conumption at home.

5.107 Nonetless, a substantialamount of arable land seems to go uncultivatedeach year (Tapson, 1985; Benso, 1981; Lenta and Maasdorp, 1988). Three main reasons explainthis paradox:

a. Ihe portion left uncultivatedis in fact beavily populatedwith livestock. Investmentin livestock is a very profitable, even under overcrowdedconditions. The Native Lands Act of 1913 explicitlyforbade black tenants on white land to keep any cattle in excessof what was needed to plow the white owner's land. The only place for Africansto keep livestock,then, is in the homelands. Even with such restrictionsin place andthe extreme land pressure in the homelands,blacks contiue to own over one third of SouthAfrica's national herd.

b. Becausecrop cultivationreturns little profit, it is unable to attract labor, given exising opporunity costs of labor.

Chapt: E.-ickey ofAgricdiural Pduon Page 129 SouwhAfriaan Aarickure: SrcngW. Perfomiance& Com for Om Mm c. A significantportion of *arable"land is in factnot arable (McKenzie, Weiner, and Vink, 1989). d. Transactioncosts in the landmarket are high. becau selling,purchg, and renting land is tehically illegal. 5.108 Theprofitability of cropcultivation seems to havebeen declining becase of the worseningtms of trade. Whilefamily farmers in the homelandsalmost quadrupled maize output from 1970 to 1990,the termsof tradeof inputssuch as ferlizers andtractors declined by between40 to 60%in the sameperiod (Ducan et al., 1992). Nevertheless,homelands production Increases In maize,beans, sorghum,and groundnutsare outpacingthe commercialareas. Whenpartial relaxation of restrictionson blackfarmers occurred,smaltholder production in sugarand cottonincreased, as wellas marketgardens in vegetables (Ardington,1990; Wheeler and Ortmann,1990). 5.109 Severalstudies have also exploredthe relationshipbetween farm size and efficiencyin the homelands,where farms are in generalmuch smaller than thosein the commercialsector. Nieuwoudt (1991)notes that smallfarmers may use landmuch more intensivelythan do largerfarmers. Latt and Nieuwoudt(1988) used data on 140 householdsin Umbumbuludistrict of KwaZuluto conducta disbcmint analysisof inputuse. Theyfound that farms of lessthan 1 hectareapplied inputs much more intensivelythan farms larger than 1 hectare;thus they suggested that smaller farms may maximize retms to land (theirscarce resource), while larger farms maximize returns to labor or capital. 5.110 Moreover,even in comparisonwith the commercialsector, homeland agriculture is sometimes foundto be more efficient Casestudy material shows that in drylandcotton smallholders in KwaZlu are moreviable than largefarmers (Wheeler and Ortmann,1990:251). These trends are also confirmed by the two casestudies mentioned earlier where sma farms, giventhe samesupport structures in tea and sugarfarming, have doneas wellor outperformedlarger farms (van Zyl and Vink, 1992). 5.111 In a study of 60 farmersin Gazankulu,half of whomwere identifiedby extensionstaff as commerciy-oriented,Nicholson and Bembridgefound that ucommercialfarmers* cultivate more land than typicalsubsistence farmers" (12.2 hectares versus 2.3 hectares),own morecattle and equipment, are better educated,and are more likely than their neighborsto keep records. Similarfindings are reportedfor Trasel (Bembridge1991a), KwaZulu (Bembridge 1991b, Wheeler, and Ortmamn,1990, Nieuwoudtand Vink, 1989)and Lebowa(Coetzee). Ihe authorsof these latter studiesalso notethe importce of incomefrom non-agricultural sources, such as wageemployment, in providingworking capitalfor the purchaseof seeds,fertilizers, and otherproduction inputs. 5.112 Giventhe relativelyequitable distribution of landholdings in the homelandsand the lackof rights to buy or sell land,much of the increasein operationalholdings by successfiulfarmers represents formal or infomal acquisitionof temporaryuse rightsheld by others. Suchtransfers of temporaryuse rights maybe consistentwith improvement in efficiency,because they combine idle land with surplus labor and otherfactors. Suchtrarsfers are currentlylimited by the extremepopulation pressure under which the tenmrearrangements operate. Landholders fear they will lose long-term rights to theirland if theypermit othersto use it.

Chuqr: 1Fjkknc4VAgrifcuW hod&alon Page 130 Sowh A*ran Agriculwe Stntru. Parfoanca & O*ons hr dfor ura

TechnicalEffcieW of Agdiculre in th Homelands

S.113 Althoughdata on productionare generallyavailable, the lack of data on area and yield makes analysisof trends in technicalefficiency in mostof the homelandsimpossible. Such data are available, however,for Lebowa(see earae sectionin Annex3); they Indicaterelatively higher rates of growth in yieldsfor mostof the majorcrops from 1981to 1989,compared with commercial areas. Moreover, In the TBVC('independent-) states production of maize,dry beans,sorghum, and grundnuts appears to be oulpacingthe commercialsector. Unfortunately,lack of area data precludesanalysis of deterinants of this productiongrowth. Physicalyield data are discussedin Annex3.

Chqr : 1,4cieuqfsArkcft1 Produv*x Pae 131 SouthAfria A1erlcuure:Struetie. Prformawce A Ozons for theFuture

CHAPTER6: FARM PROEThABILr

6.1 As demonstratedin earlier chapters, the growth path followed by many parts of South African agriculturewas characterizedby large-scale,capital-intensive production methods. Chapter S examined the efficiency of this strategy by analyzingthree different types of efficiency and concludeddtat the mpugm scale efficiencyof large farms is due pdimarilyto the preferentialpolicy treatment a.corded these enterprises. The ultimatetest of efficiencyand financial sustainabiity of the enterprises,however, is the responseof large-scalefarms to the withdrawalof this preferentialpolicy treatment. If the structure of incentives(e.g., prices of inputs and outputs,taxes, and/or marketingsystems) changes, farms are unable to respond adequatelyand ensuretheir profitability,this constitutesa reliableindicator of their inefficiency in an undistortedpolicy environment.

6.2 Sincethe mid-1970s,agriculture in SouthAfrica was exposedto a cost-pricesqueeze that reduced farm profitability. This was compoundedduring the 1980sby: (i) unfavorableweather conditions;(U) reduction of farm subsidies; (iii) liberalizationof marketingpolicies that often resulted in lower real producer prices; (iv) and higher capital costs due mainly to higher interest costs. The failure of many firms to respond to the changingprice environment(recall the finding in ChapterS of only modestprice efficiency) meant that many farms continued to rely on the established, capital-intensiveproduction methods. Given the variable agro-climaticconditions to which South African farming is exposed, the continuedreliance on mechanized,monocropped farming also entaileda higher exposureto climaticrisk and increasedthe volatilityof farm profits.

6.3 As a result, the debt burden carriedby many farmers, in particular those in the summer rainfall areas (i.e. wheat and maizegrowers), escalatedto unsustainablelevels. Althoughreal farm debt reached its highest level in 1985 and has declinedsince then, the financialposition of many farmershas continued to deteriorate,as illustratedby a continuingincrease in the debt-massetratio. Moreover,the ratio of short- term debt to total debt increasedconsiderably. Higher capital costs, tighter credit standards,fewer farm subsidies, and the growing unsustainabiity of the debt burden should have positive effects on labor employmentin the sector and force less efficient farmers into foreclosure. However, the government contiued its atempts to aid the sector financially; and when the 1992 drought occurred, substanial amountsof debt which had been accumulatedduring the 1980s were effectivelywritten off through a massive recapitalizationeffort targeted to grain producers and totalingR3.4 billion.

6.4 The implicationof the downwardtrend In farm profitabilityand the resultantaccumulation of debt as a result-in large measure-of the changingpolicy environment,suggests that many of the large-scale farms are financiallyinefficient. To the extent that this inefficiencyis caused by size, it argues for a strategy that encouragesa down-sizingof agriculturalenterprises. This conclusionmust, however, be treated with caution. There is tremendousvariability in across the agriculturesector due among other factors to managerialability, and it would be incorrect to argue that a smaller scale of operationswould be appropriatefor all subsectors. At the sametime, a combinationof internationalevidence with regard to farm size and recent South African experiencesuggest that a smaller scae of operations in many agriculturalsub-sectors would be at least as efficientas the large scale farms once the picy distro har 6Fiminated.

Cagpter6:- FarmProjbabAft Page 132 Sout Afta Arrkubre: S&nwwy.Perfonnawe & Optior thoFing

DecliningFam Revenues

6.5 Farmreveumes, reflectng the combinedeffects of outputprice andquantty chges, showedthat aggrega gro farmrevees icreased up to 1980,but declinedslowly during the ne decade(able 6.1). For each of the major subsectors, tht followingtrends emerge: a. Red gross revenuesfrom fieldcrops rose steadilyfrom 1960through 1980, then stalled or declinedthrough the 1980s,faling 44% between1980 and 1990; b. Real gross revemnesfor animalproducts also rose steadilyfrom 1960to 1980,but stagnatedduring the 1980s;and c. Realgross revenues for horticultura products rose gradually until 1980.Ihe horticultral sectorwas the onlymajor sector to exhibitmodest real growthin the 1980s.

Table 6.1: GronsRevunues in AgnicukuralSubsectos (R milion, consJtn 1985 tem)s)

Gross Valic Gross Value of Gros Value (hrss Va1ue of Fild Crops Hortbuluro of Animal of Parm Crops Products Output 1990 3,646 2,224 4,795 10,665 1989 4,095 2,271 5,126 11,492 1988 5.041 2,169 5,630 12,840 1987 4,343 2,191 S,289 11,822 1986 4,416 2,104 4,870 11,391 1985 4,580 2,101 4,752 11,433 1984 5,522 2,071 5,159 12,752 1983 4,427 1,985 5,016 11,428 1982 4,520 1,956 5,055 11,531 1981 S5305 2,124 5,372 12,801 1980 6,565 1,953 5,017 13,535 1975 4,245 1,78S 4,142 10,172 1970 3,743 1,383 2,891 8,018

Sble: Abstra of Agricilurl Stistics, 1992, pp. 79, 82.

6.6 Producerprices of mostmajor agricutura commoditiesand of some agricultura inputs in South Af*icawere highly reglated and controlleduntil the mid-1980s.This applies specifically to cers, red meat, sugar, wine, and to machinery, fertilizer, fuel, and electricity. Producerprice movements, therefore,do not necessarilyreflect relative demand and supplyconditions, but rather incentivesto increaseor decreaseproduction. Given such price regulation,inenational marketsare basicallyused to

Osapw 6. Fn Prq1abWt P0* 133 Sout Afrkan Arculre Ssue. Perrormawe O&tinsfor the Fur dispose of domesticsurpluses or to compensatefor deficits.

6.7 Producerprices of agriculturu commoditiesincreased in real terms duringmost of the 1970s,but declinedduring the 1980s (T'able6.2). During the 1980sfield crops experiencedthe worst declinein real prices-a 32% declinefrom 1980to 1990. Price trends for animaland horticulturalproducts showed less marked declines.

a. The real producerprice index of field crops increasedmodestly from between 1970 and 1980, fluctuated in a fairly tight range from 1980 through 1984, and then fell to three-quartersof its 1970svalue by 1990.

b. The real producer price index of horticultural crops experienceda brief upturn in the 1970s, dropped sharply between 1980 and 1981, stagnatedbetween 1981-88, and fell moderatelythereafter.

c. Ihe real producer price index of animal productsincreased between 1970 and 1980 and then graduallydeclined throughout the 1980s.

6.8 High producerprices inducedrapid expansionof output durng the 1960sand 1970s,but aggregate output was relativelystagnant throughoutthe 1980s. The fall in real producer prices during the 1980s led to stnant agriculturalgrowth and a contractionof output for fieldcrops. Additionally,a series of regionaldroughts in 1983, 1984, and 1986" reducedoutput volumes,especially of field crops. The Land ConversionScheme, which subsidizedthe withdrawalof crop land and the establishmentof perennial pastures during the latter part of the 1980s and early 1990s, further contributedtoward this trend.

a. The volume index for fieldcrop production rose substanially from 1960 to 1980, however, from 1980-2to 1988-90,the three-yearindex declinedslightly.

b. During the 1980s, the volume of animal production index increased slightly, while horticilture crops experienceda significantrise in volumeof production.

iAb and uctuating naua codio (drout, flood and ot dMg) have drssd yie in the smmrai nanl _o'ngarea. n lag pab of oexteock gAn uisn theutemCape seriousdroug coinided with a cap of wool. mohair,and mat prim. See Anme=6 for data of rainfll.

Chatr 6: Farm Prq%biky! Page 134 Soth Aian AMiubm: Suctue. Peorfanow O&.do for dth Fur

TaSle 6.2. Prie and VolumeIndies of Seleted Atrigutuml ProductCateodc

Fied Crop Indice HotloulWnl Crop Indies Anmal Produt Indkoe AggrmegatInices NP RP VP NP RP VP NP RP VP NP RP VP 1990 1SS .76 101 190 .93 117 179 .88 109 170 .83 107 1989 136 .77 126 159 .89 124 177 .99 102 157 .88 115 1988 125 .81 107 154 .99 116 171 1.10 99 149 .96 105 1987 11 .86 102 142 1.03 109 143 1.04 98 132 .96 101 1986 110 .93 98 122 1.03 102 115 .97 96 114 .96 98 1985 100 1.00 100 100 1.00 100 100 1.00 100 100 1.00 100 1984 98 1.14 77 85 .99 103 91 1.05 103 94 1.09 92 1983 86 1.12 68 82 1.07 95 81 1.05 101 83 1.08 86 1982 71 1.03 101 7S 1.09 104 78 1.13 97 74 1.07 100 1981 61 1.02 138 63 1.06 96 75 1.25 87 66 1.11 110 1980 S8 1.12 112 63 1.21 95 60 1.15 91 59 1.14 100 1975 31 1.06 95 40 1.36 80 31 1.06 75 33 1.12 84 1970 18 .98 71 22 1.20 67 16 .83 66 18 .97 68 1965 16 .99 52 20 1.25 55 1S .93 S7 16 1.02 54 1960 15 1.03 47 16 1.14 43 13 .91 51 1S 1.01 48

NP - nominl price index; RP 8 roal - index (NP/CPI);VP volumeinhde Sorc AbstUrabof Agriultuld Ststics, 1992, pp. 88, 89.

Increing Farm Cms

6.9 Theanalysis of productionvolumes and prices in agricultureindicates that real gross revenuesin the agricltural sector grew duing the 1970s,but declinedduring the 1980s. Bec e profitscompris bothrevemies and costs, as well as the associatedquantities, trends in inputprices and volumes should also be considered.Table 6.3 presentstime series data on price and volumeindices associated with all farmrequisites, Including machiey and implements,fixed improvemen, and trmediategoods. Ibe real price of all farm reuiites increasedby about34% duringthe 1970s,but wasbasically stgnant duringthe 1980s. a. Realinput prices of machineryand implementsincreased steadily over the period1970- 1990,while the volumeof theseinputs increased from 1960 to 1981,but bavedecreaed steadilysince then;

b. Real inputpries of fixedimprovements increased considerably throughout the period, but the quantityindex shows a rapidand sustaineddecline after 1983;and

c. Realprices of intermediateinputs increased from 1970to 1980,then dropped somewhat, but remained fairly constant throughoutthe 1980s. Real expenditureson iant ediate

Cae & Farm frafaila Page 1- Sou M:ca, Arwtare. Perfonnawe A G-a forOm ure goodsand servicesincreased rapidly from 1960 through the mid-1980s,but hae trended dow"wardsi then.

Tabo 6.3: Ne d Volum Indigs of Panmnlseuitst

Machilem Fxed tAU Famng and movements GOOd Requisites

NP RP VOL NP RP VOL NP RP VOL NP RP VOL 1990 236 1.16 70 215 1.05 68 181 .89lo 198 .97 89 1989 212 1.19 73 187 1.05 75 163 .91106 177 .99 94 1988 175 1.13 82 161 1.03 83 135 .87 110 147 .95101 1987 152 1.10 86 141 1.03 89 121 .u8 98 130 .9s 94 1986128 1.08 93 124 1.05 87 116 .97 93 119 1.01 92 198 10oo 1.00 100 100 1.00 100 100 1.00 100 100 1.00 100 1984 81 .94 115 86 1.0011o 8s .99 106 84 .98 108 1983 71 .92122 76 .99 126 81 1.05 98 78 1.01 107 1982 62 .91 127 65 .94 123 71 1.04 99 68 1.00108 1981 52 .88 129 57 .95 126 61 1.03 107 59 .98 114 1980 46 .89 113 51 .98 135 56 1.07 106 53 1.02 111 1975 24 .82 110 2s .86 146 26 .88 95 25 .86 117 1970 14 .73 117 14 .72 156 1s .78 69 14 .76125 1965 12 .75 101 12 .78 134 14 .85 51 13 .80 94 1960 11 .75 100 11 .78 124 13 .93 43 12 .86 83

NP nominaln price index; RP - real pc index (NPICPI);VOL -volme indae. a. Trcto, lorries,implemens, and irition equipm b. Paeier,, fuelarmfeed, dips and sprys, packig matrial, andmainteance and repairwork. ac MAbsrct of Agriul Sttstics,1992, p. 96.

6.10 Since 1973agdiulture has beencharcterized by a cost-pricesqueee whichaccelerated during the 1980s(rable 6.4). Whilethe govermentprovided substantial supports in the form of consumer subdie on wheat(bread), maize, grain sorghum, dairy products, fertilizers, stock feed, and transport rebatesfrom 1980to 1986,such subsidiesfell shaply in real tems in the lae 1980s. Since 1987the growthof outputprices has slowedto wellbelow the rateof domesticinflation, as a resut principallyof policiesbringing domestic prices into line with lower border parity prices. No similartrend in farminput priceshas occurred,however, which suggesu that the liber rizaonof recent years has affectedimput pricesless thamoutput prices. The increasesin inputprices relative to commodityprices were also not fully counteedby increasesin totalfctor productivity.

0aqwte PamF Profl t Page1, Sout AfIc Awirkure: Sbuctmu. P&*o & Opnikmfor Fe Futwo-

Tabl 6.4: Averase Annual Grwth Rates in Catain Prices and OuRties. 1960MI (%)

Consum prices Producer Prie of LAndprice Faming Pedod prices of ag firming debt All iM Food product rqustes

1960-68 2.4 2.7 2.4 1.2 5.9 1968-73 6.0 6.7 10.4 5.6 6.1 7.8 1973-90 13.5 14.8 11.6 16.8 12.5 15.0 1960-1990 9.0 9.9 8.9 10.3 9.5

Not available 3Somho van Zyl (1991)

6.11 There have, however,been sharp differencesbetwun varioustypes of inputs. Thereis evidence, fbr example,that the rise in fertilizer prices has modeated recently (Table 6.5), with the following perenage annual risesbeing recorded. In part this is likely to reflect the lower demandfor theseinputs.

Table 6.5: Feutll Prices: Percnt Rise on Previous Year

Year X Rise on Previous Year 1985 33.4 1986 15.2 1987 -0.4 1988 23.9 1989 19.6 1990 7.3 1991 11.0

: Adapted from RtSA1992 (a), table 94.

6.12 Uroenewald(1982, 1986) attributesthe steady increasein the real costs of farm machineryand implementsto the existenceof monopoliesin the farm-inputmanufacring sector and to tariff protection of local industries. Since World War U, South Africa has followeda policy of industrialization,based on Importsubstitution and activelypromoted by protectivemeasures such as importcontrols and tariffs; the reslting distortios were passed on to farmers. The increasein real input prices has been linked to the protction of the domestic fum-input mu ing sector from international competition. For example,the fertilizer industrybenefitted from fixed pricing based on a certuinyield on capital, reducing incenives for efficiencyor capital saving. I was estimatedthat farmers paid an extra R272 million for Inputsin 1982because of these policies. Crop farmerswere particularlyhard hit, however,farm outputs were also protectedfrom competitionand there is someevidence that on balance farmers as a group have

Choter 6: Farm PoltabWy Page 13? Sqw Af0c AeIaim &fU*w, Perfowe & QatIo the Fu c gainedmox from protecionof outu then they have lost from protectionof inputs (Stadleret al., 1983F. It was esmated in 1982tat the cost of Intermediateinputs (cate feed, fertlizer, fuel) was 6.9%higher ta it otherwisewould have been under tariff-free iuterntiotal trade.

FinanciaIndigators of Profitability 6.13 Post-wartrends in real net farmincome go throughtwo distinctperiods. 'Te first period,up to the mid 1970s,is charterized by gradualimprovemets in net farmincome. 'he secondperiod, from ihemid 1970s until 1991,sees a reversalof histrend. Realnet farmincome declines, although the trend is chaatized by strongyear-to-year flucuions. lhe decliningtrend of net farmincome is the result of two countervaiingforces. Whereastechnical efficiency-as measured by total factorproductvity (rFP)-improves,it is insufficientto compensatefor the increasingcost-price squeeze-as measured by the termsof trade in agriculture(see Figure 6.1).

Figure 6.1 Realnet farm income,total factor productivity and agriculturalterms of trade (1947-1991)

!325' 300

i 250250 - Realnoi net farmy aC 225- 200-- 175--T

150

25 -

1947 1951 1955 1959 1963 1967 1971 1975 1979 1983 1987 1991 yeaw

s~: adaptd frm van Zyl van Sch and Tirtlo, 1993.

a Ws oideos is, howeyns, ionl and shoud be intprcted with evution as =port rafion instead of port parity wasused as measur of aiutua protet ahu, & Farm Proftabirsy Page 138 South Afra Asrrculure: Strtu. PeformawA & ons for she Future 6.14 As indicatedin Table 6.6, real net farm incomebefore interest payments gradually declined dtroughoutthe 1980s,from a high of R5.6 billion in 1981 to R3.5 billion by 1991'. The gradual deteriorationin the realgross value of farmoutput, combined with steady real costsof intermediateinputs andfluctuating other costs, has resultedin very unstable real farm incomes throughout most of the 1980s. Real net fam income(defined here as real grossfarm incomeminus real expenditureson intermediate goodsand services,salaries, wages, rents, and depreciation)generally increased during the 1970s,but trendeddownwards until 1985because of an increasein real inputcosts andbecause of a rapid increase in interestpayments. Net farmincome showed no consistenttrend afterward,but declinedin 1991.

Table 6.6: lncome andExenditur. Asses and Liablities of Ariculture. 1970-1991

1970 1975 1980 1981 1983 1985 1987 1988 1989 1990 1991 1. Real gross farm 6,76S 9,636 11,333 11,880 9,237 9,270 9,946 9,888 10,841 9,924 8,998 income Real expenditure: 2. Int goods, 2,235 3,081 4,264 4,445 4,422 4,144 3,806 3,879 4,102 3,972 3,802 services 3. Salaries,wages, 1,802 1,558 1,709 1,798 2,323 1,661 1,784 1,755 1,840 1,796 1,727 rent depreciation

4. Net fam income 2,728 4,997 5,360 5,637 2,492 3,465 4,356 4,254 4,899 4,156 3,469 (1-2-3) S. Interet payments 396 456 622 911 1,393 1,698 1,198 1,094 1,311 1,300 1,182 6. Net farm profit 2,332 4,541 4,738 4,726 1,099 1,767 3,158 3,160 3,588 Z,856 2,287 (4-5) 7. Reld value of 49,209 57,735 55,066 49,455 47,029 42,067 36,153 35,047 37,524 33,861 29,561 capi asses . Realshort-term 2,112 2,388 3,214 3,652 5,232 6,069 5,069 4,433 4,622 4,284 4,028 debt 9. Ral total debt 7,497 6,816 7,397 8,09 9,610 11,118 9,649 8,740 8,462 7,784 7,055 CPI (1985=100) 18.7 29.4 51.9 59.8 77.1 100.0 137.7 155.4 178.2 203.8 237.0 AnnualInflation 5.1 13.5 13.8 15.2 12.4 16.3 16.1 12.9 14.7 14.4 16.3

Real values are nominal revenues,expenditures, profits, assets and debt divided by the ConsumerPrice Index. Nominalvalues are taken from de KIelk, 'ne accumulationcrisis in agnculture,' pp. 200, 201; and personal conversao with the Depatment of Agrculural Economics, 1992. Short term debt is loans from commercialbanks and agricultual cooperatives.

0 Gross icome in Table 5.6 includesprofit repayments on intenationaltransacions in maizeand wheat (exluded in Table 5.7). Ohewis, the sourceof data is the samefor the two sets of estimates.

Chutfr 6: FannProfiabilty Page139 SouthAfrican ARrkuture: Struaure. Peffornance A Optfons for the Future

6.15 However, once interestpayments are deductedfrom net farm incomein Table 6.6, trends in the resultingreal net farm profit exhibit dramaticswings. From a peak of R4,700million In 1981, net farm profit plummetedto R1,100million by 1983before gradually reviving to a rangeof R2,300-R3,600 million between 1987 and 1991. Thus during the 1980s interest paymentsseem to explain most of the volatility in net farm profits. Real interest paymentstotalled R396million in 1910 and R622 million in 1980, implyingan annual increaseof about 4.6%. However, from 1980 to 1985, real interestpayments rose by 173%,to R1>698million, at an annualrate of 10.6%. Aft 1985interest payments fell in real terms to R1,182by 1991.

6.16 Liquidity governs the capacityto repay short-termdebts. A close proxy for farm liquidityis the ratio of net farm profit to short-termdebt. As indicatedin Table 6.7, this ratio fell by 68% from 1975 (1.9)to 1991(0.6), indicating both the downwarddecline in farmearnings and the rise in debt servicing needs for short-term debt. Trends in financialprofitability are illustratedin Table 6.7 by the net rate of return on assets (the ratio of net farm incometo the value of capital assets) and on equity (the ratio of net farm profit and equity-equity being the differencebetween debt and the total value of assets. The net rate of return on assets and on equity increasedbetween 1970 and 1980, decreaseddramatically between 1980 and 1985, and improvedand remainedconstant at around 11 to 12% between 1987 and 1991. The reason for the upturn in the late 1980s is related to the falling capital-outputratio after 1983.

6.17 The decline in returns between 1980 and 1985, together with a move to more market-oriented interest rates, constrained the supply of capital for agricultural investment. The capital-outputratio declined from 5.1 in 1983to 3.3 by 1991, and the real value of capital assets fell by nearly half between 1980and 1991 (from R55.1 billion to R29.6 billion).

Table 6.7: AericulturalFmancial Indicators

1970 1975 1980 1981 1983 1985 1987 '1 1989 1990 1991 Net return on assts (%) (4/7) 5.5 8.7 9.7 11.4 5.3 8.2 12.0 12.1 13.1 12.3 11.7 Net return on OB (%) [61(7-9)15.6 8.9 9.9 11.4 2.9 5.7 11.9 12.0 12.3 11.0 10.2 Nominalnet worth (R b) 7.8 15.9 24.7 22.7 28.9 30.9 36.5 40.9 51.8 53.1 S3.3 Real net worth (R b) (7-9) 41.7 50.9 47.7 41.4 37.4 30.9 26.5 26.3 29.1 26.1 22.5 Debt burden (%) (9/7) 15.2 11.8 13.4 16.3 20.4 26.4 26.7 24.9 22.6 23.0 23.9 Short term debt (%) (8/9) 28.2 35.0 43.4 45.1 54.4 54.6 52.5 50.7 54.6 55.0 S7.1 Avg interest rat (ANIR)(5/9) 5.3 6.7 8.4 11.3 14.5 15.3 12.4 12.5 15.5 16.7 16.8 Avg real interest (ARIR) 0.2 -6.8 -5.4 -3.9 -2.1 -1.0 -3.7 0.4 0.8 2.3 0.5 NFP/STDBTratio (6J8) 1.1 1.9 1.5 1.3 0.2 0.3 0.6 0.7 0.8 0.7 0.6 Capital-outputratio (7M1) 7.3 6.0 4.9 4.2 5.1 4.5 3.6 3.5 3.5 3.4 3.3

OE o=wneesequity. ARIR = Average nominalrate of interest less rate of inflation. NFP = net fann profit. STDBT= short-tem debt.

6.18 The largest part of this decline resulted from falling land prices since around 1983 (itself an indication of declining profitability) and declining real net investment in fixed improvementsand machinery since 1984(Table 6.8). The aggregatedecline in land prices also masks importantregional

Cawter 6: Farn Profaabiry Page140 Souh Aftn A*cukmwr: Scurw. Perfomnae & Onons br the Fphre differences.The decline in landprices, and hence profitability, is mostmarked for the summerral areas, followedby the winterrainfall areas. Giventhe predominanceof mechanizedmaize and wheat cultivationin the formerand the Importanceof horticulturalproduction in the latter,these trends confirm the hypothesisthat the inward-looking,capital-intensive subsectors lack the efficiencyof the export- oriented,labor-intensive sub-sectors, once certain policy distortions are graduallyremoved, as theywere in the 1980s.

Tabl. 6.8: RaLan pL i. Real Capital Fonrntion. Lon-Term Debt. and Arrears

Land Price Indios Real Capital Formation, Real Value of Real Value Land bank: Percent Pixed Machinery& Motgag Loas for of Cpt Owed in Imprvements Implements Loans to Land Anrs SR WR CG SG (R million) (R miLlion) Farmers Purchases Ise Capital (R million) (R milion) (%) (M) A B C D B P a H I J 1990 66 82 92 90 344.0 388.7 263.4 157.3 6.40 1.41 1989 68 78 86 88 372.7 524.2 322.8 206.3 5.75 1.47 1988 73 76 8S 87 345.9 536.6 297.5 198.S 5.45 1.60 1987 79 87 78 89 360.6 410.5 299.8 178.5 5.62 1.63 1986 8S 80 103 95 360.9 488.6 357.2 184.2 4.55 1.22 1985 100 100 100 100 388.4 663.S 508.6 218.3 2.76 0.94 1984 114 97 100 107 414.2 737.9 6S6.6 245.1 1.94 0.9S 1983 118 89 118 98 452.4 895.2 965.2 246.1 2.03 0.96 1982 121 76 116 84 582.8 1,131.9 266.3 140.1 1.26 0.92 1981 121 70 106 118 572.3 1,723.8 389.4 262.2 0.74 0.74 1980 105 59 113 125 568.5 1,294.4 365.5 265.5 1.14 1.06 1979 97 67 96 101 544.3 899.0 302.7 182.2 1.64 1.48 1978 100 77 103 102 556.3 961.8 231.5 143.7 1.64 1.48 1977 114 88 137 88 529.5 1,058.0 178.7 105.5 1.62 1.42 1976 103 83 145 91 581.9 1,056.7 204.4 12S.0 1.49 1.27 1975 116 90 143 102 641.3 1,345.6 261.9 179.7 1.25 0.99 1974 105 81 11S 100 655.6 961.5 275.6 208.8 1.21 1.15 1973 101 77 111 95 693.3 888.5 227.8 166.5 1.64 1.28 1972 103 78 102 104 645..) 84.0 176.6 113.8 1.85 1.24 1971 107 85 8S 111 571.6 920.9 313.8 163.3 1.92 1.12 1970 102 69 88 108 542.7 813.8 470.7 148.6 2.06 1.13

SR = Summerrain region;WR = winter rain region;CG = cattle grazingregion; and SG = seew grazing region. he ral price ndies for and is caldated as the nomil land prc indexdivided by CPI and adjuatedto make 1985=100. Ihe real value of fixed improveentb is caeulated Ps nominalexpenditues on fixed improvementsdivided by the price index for 'mateials for fixed improvements"and adjuted to constant 1985Rand (1985=100) (Abstractof AgricukuralStatistics 1992, p. S5, 97). Rea fixed investmentin ractors, machney and implementsequals the nominalinvestment divided by the price index for machhiy and implementsand adjustedto equal 1985 constantRand (Abstraet,p. 85,97). Percent of Land Bank long-termdebt in arars is taken from Table F.2, AppendixF.

Chpwrr 6 Farm Projkabity Page 141 SouthAfrian Arulure: Sreure. PXt&== A Cetlonsfor uheFsoure eb Burten and ParmLending Policies

6.19 The abovetrends illusta a declineIn profitabilitythat causedsevere cash-flow problems in agricult (vanZyl ad van Rooye 19). Liquidityproblems have affectedthe finacial standingof commercialagriculture in three ways: (i) debt loads increased;(ii) loan arrearsmounted; and (tii) sequestrationsincreased. 6.20 Since the mid-1970sthe real total debt of farms in the commercialsector has increased substantially.The real net worthof farminghas declinedsince the mid-seventiesand after 1980the farm debt becamefinancially unsustainable because of a steadyerosion of real net farm income.However, based on data in Table 6.7, the averagedebt burden(total iaebtIvalue of capitalassets) would appear reasonable.It increasedfrom 15.2% of asset in 1970to 26.7%in 1987,before declining to 23.9%by 1991. Thesedata, however,do not takeinto accountimportant regional differences. The SAAU(1984) conducteda studyof the financialconditions of farmersIn 198319and foundthat: a. about 15,200farmers or 22.4%of the totalnumber (slightly above the 20.4%reported in Table6.7) had criticaldebt burdens in 1983,based on SAAU'scriteria. Thesefarms werelargely concentrated in the Transvaaland the OrangeFree State(de KMerk,1991); and

b. the most damagedsubsectors were summercrops (52% of farmsbeyond the critical level),and wintercrops (22.6%). By the end of 1984,these estimateswere expected to havegrown to 22,700farmers (33% of the total), with65% of summercrop farmers and 38% of wintercrop farmers beyond the critcal level. 6.21 Moreover,the declinein farmprofitability seems also to havecaused a substitutionof short-term for long-termdebt from 1970until the mid-1980s.The ratio of short-temto totaldebt increased from 28.2%in 1970to 54.6%in 1983,and peakedin 1991at 57.1%.

6.22 The evolutionof the debt-burdenIs closelylinked to the lendingpolicies which prevaied in te agriculturalsector. A numberof creditsources are opento farmers,Including: (i) the Land Bank;70(i) the AgriculturalCredit Board;t (ill)agricultural cooperatives;" (iv) commercialbanks and other private

0 Basedon prevaling economic _cimoesin that year, thfeSAAU calculated the aida dbt bwden for s of the lwrge subcetors: 16.7% for sxnmw crops;34.1% for winter crops; 10.2% for red meat; 17.2%for milc; and 14.2%for woo.

"' lhe Iand Bankis a statuborybody establishedto pvide for the edt needs of fnn. The bank advanes moneyto fares, agicultur coopeaive, contol boards sablished under the provisionsof the 1968 MAetine Act, and stautory a Ficulturainutons. Long-rm s are ganted to farmes againstmortge of land or the reOjtrion of a cha aginst land. Shoutand internediatetm loam are grantedin tms of secdon34 of the lnd A.t

nThe(Agiulua Cmd Boardwas etablishd by an Aot of Parlimant in 1966 and is admisered by the Dqmat of Agricul Dedopnent. Up to around 1983, creditby the Boardwas povided only to white com fams, after which aedit servces wae minimallyedd to In&ian,coloreds and blcks

Chtper 6: Farm Projlabiity Page 142 &SWhkl J Ack*u Alrsct,r. Prfmmancu d ons for the FWtre sectorfinnca iitutions; and (v) privateindividuals. 6.23 Short-termcredit is generallyprovided by the commercialbanks and cooperatives, while the Land BankIs the primay supplierof long-termcredit for landpurchases. The Agricultura Credit Board is the lenderof last resortto Insolventor nearlyinsolvent farms (Table 6.9).n 6.24 At the end of 1991,total debt owedby individualfarmers totalled R16,489 million. However, thesefigures cover only direct fin ing toDt-ers. TheLand Bank also extendscredit to cooperatives enablingthem to grant creditto their membersand to purchaseproduction and farmingrequirements. At the end of 1990cooperatives owed the BankR6,023 million under cash credit advances, including an amountof R3,780million (ndirect credit to farmers)owed by membersto the coopeatves. If directand indirectchannels are aggregated,the Land Bankis the largestlender to farmers. Combiningcredit in both directand indirectchannels, the Land Bank's financingto farmersin 1991amounted to R7,813 milion, or 45.6%of totaldebt outstanding in the agriculturalsector (Land Bank data). 6.25 Accordingto LandBank policy, it doesnot extendcredit beyond the assessedcollaterai of farm property,which Is valuedon a productionrather than a marketbasis. As landrepresents the largestpart of farm capitalassets, trends in real land priceshave an importantbearing on the farmsector's debt- carryingcapacity, and as indicatedin Table6.8, real landprices fell in mostof the countryin the 1980s. However,Issuance of long term mortgagecredit by the Land Bank(G) spikedupward in the earlyto mid-1980s,before it fell through1990.

n MmAgricultura Coopeati ae regardedas thefarmes' ownindepende business orantos. Asof 1991,thre wer roughly245 p=imay Coo veswith 1,300 banches throughout the county. Tes cooprat smpplythir memberwith most produce goodssuch seod,feilisw, fA rep services,credit and tnsio sevces. eboy handlemot members' produo. The ancalso 41 eentralcoopera thatprovidc the primaycopea withspeciffo fSring reqtes, serices swuhas proeuig and marktingof agicultal prdu, and hnur see s for crps andshort-term damage. Coopeve moveramounted to R20,929.2million in 1989.Of fts total,produce accouned for R15,340 million, faing rquisitesR4,720.S min, andseviaes R868.7 miln (1991Year Book).

' TheLand k and Agricu l edi Board dsngi am g tro e typesof farmsin assing end-use of credt awtics categoyono farms are thosethat are well off financiallyand are verysound commecially (blue-chs); categoy two fam he m dim finanal standng;and ategoy 3 farmsame thoso t havepoor financid arnings and are nsvet or nearly so. Commeci bankstend to lendto caegor onoperaio. M Land Bankseric caegoriesone an two. TheAgiculura CraditBoard is the upier of astraort for ctqpy threefirms.

Chapur6: Fenn pr*4iavy Page143 t~~~~~~~~~~~~~~~~~~ASLandfl..& CoauumbT -6A9: Camnaci Paim Dmun AAetas" yCeiw:I

Yw 7 _ _ I CApco__ _ _

t 1"1 3,512 21S 5,tl6 31S 4.03 26% 972 6% 1,07243 S S 2S t990 3,441 22S 4,950 31% 3,7S 24* t,Otl 6S 1,t55 7% 1200<2> es 324 2S ISA4 MO 3,149 21S 4,650 3tS 3,5S7 24S 97 6S 1,072 7S 1350 <2> 9S 301 2S IS,0 19SS 2,924 22S 3,47 26* 3,412 25S 921 7% "9 7S 159 <2> IIS 2S0 2S 13,5t3 t9S7 2.801 22S 3,355 26S 3.224 25% 789 6% 940 7S IS <2> 12* 26 2S 12.1 1986 2,649 21% 3,437 2S% 3.01 25% 684 6% 690 7% 1,420 i1t 251 2% 12.413 198S 2.338 2IS 32335 30% 2,754 25S 549 5* 792 7% 1,128 10* 241 2% 11.119 1984 1.923 20% 2,ff9 33* 2.234 24% 443 5% 720 s ,000 1% 207 2* 9,49 3983 1.331 IS* 2.54 30* 1.780 24S 309 4% 670 9% 883 12S it5 2% 7,410 1982 989 37% 16400 21S 1,368 24% 247 4% 634 I3S 775 13S 174 3b 5.787 198I *56 ISS 10OSS 22% 1.130 23% 202 4% 601 12% 833 17% 162 3% 4.4 1980 676 IS% 802 21% 667 23* tSO 5% 560 15% 612 16S 122 3% 3.40 1979 592 I6% 691 21S 654 20% 164 5% 486 35% S23 16* 108 3% 3,220 3978 537 19% 620 22% 556 19S 156 5% 452 16% 453 16% 95 3% 2,671 1977 508 19% 560 21% 462 38% t50 6* 415 16% 426 16% 86 3S 2,60 1976 479 21% 485 21% 344 35S t50 7% 369 16S 393 175 77 3% 2,299 1975 429 21% 455 23% 247 12% 1351 a 327 166 324 16% 70 3% 2.005 1974 390 22% 384 21% 198 1H* 149 a% 292 36% 311 17% 63 3% 1.7,7 1973 368 21% 373 22% 159 9% 155 9% 290 17% 36S 1S* 62 4% 1.726 1972 360 24* 275 ISS 146 3o% 160 ItS 235 15S 291 19* so 3% 3.53S 1971 346 24% 272 19* 128 9% 154 ItS 232 16% 279 19S so 3% 13.f2 1970 296 21% 282 20% 113 J* 136 10% 243 17% 287 20S 45 3% 1.403

Saun DOA.Dsp_aui of Asna kga BooomicTrein.c Ab_a ofAgdmIxm SMidic., 1992 <(I> nbadaadhcot waans.. mrebawbau*a idarmmay inuhueim. ua.. couyamdapwamlom Lind taut coaqiaks,al.nmme4oaa books and mmugameta. pagwcApaimmaw&Vg.bonds. <2>r banci iswfiftm <2 > DOAaalamt. &Sh Afcn Aicuur: S&ructure.Performance & Onions forOe Fwuwe

6.26 Intest rates. Startingas early as 1973,the Land Bankbegan significantlysubsidizing long-term credit in agriculture. Real interest rates were posidve (0.9 to 4.2%) between 1965-70, then turned negative for a 15 year period starting with the upturn in the rate of inflation in the early 1970s. Real interestrates fell from -0.4% in 1971 to -6.5% in 1975,then declinedto -4.0 to -4.2% between 1976 and 1978, before again dropping to a low of -8.2% in 1981.Trends in real interest rates over the 1980s changedsignificantly-from 1982 to 1984and from 1987onward represent the only periods in more than two decadesthat agriculturefaced positive real interestrates. Total debt levels beganto shrinkin the latter part of the 1980s as these real interest rates began to rise.

6.27 Ihe real interestrate represents an importantelement of the cost of capital. The lower the rate, the greater the incentivefor capital-intensiveinvestment-also including investments in land. Tnroughout most of the 1970s low and subsidizedreal long-termrates offered substantialincentives to borrow for purchasesof durable equipment,fixed improvements,and land. This was supplementedby short term off- shore borrowing in the early 1980s, ahead of the sharp fall in the exchangerate.

6.28 Tax legislation. Another importantfactor which reduced the cost of capital substantiallywas a change in tax legislationin 1977, which made farm equipmentpurchases cheaper by permittingfarmers to write off the full cost of machineryand implementsagainst taxable incomein the year of purchase (de Klerk 1991). Ihis situation changed again in 1989 with farmers allowed to depreciate machinery purchasesonly over three years like other businesses-50% in the year of purchase, followedby 30% and 20% in subsequentyears. Estimatesof the value of the preferentialtax treatmentto farmers vary. One source puts the value at RI 17 million in 1983/84. Another sourceindicates that incometax concessions to firmers amountedto 70 % of their theoreticaltax bill in 1981-84(Lamont, 1990, quoted in Vinkand Kassier, 1992).

6.29 Financialassistance. The governmentintervened in the 1980s, and continuesto do so in the early 1990s, by providing "cheap" credit and debt-reliefto insolventor near-insolventfarmers through the Financial Assistance Schemes of the Department of Agricultural Development, managed by the Agricultural Credit Board. In 1993, around 17,000 farmers benefitted from such assistance. If such farmers are also the most inefficient-whichseems a reasonableassumption-such a policy only adds to the financialunsustainability of the sector. Preferentialfinancial assistance included schemes for:

a. Debt consolidation.Farmers who did not qualifyfor commercialcredit could consolidate their debt at a nominalinterest rate of 8% (1991 rate) with a loan period of 15-20 years. In the same year, the Land Bank's long-termnominal interest rate was 16.1%. Annual disbursementsfor consolidatingdebt increasedrapidly from R9.2 million in 1981/2to R106.1 million by 1988/9for a accumulatedtotal of R522.5 million in 1991.

b. Means of crop production. Crop productioncredit to farmers who as a result of adverse farming conditions cannot acquire short-term credit elsewhere increased from R6.9 million in 1981/82to R124.0 million by 1988/9, for an accumulatedtotal of R833.6 million. In 1991, the nominal interest rate was 8%, comparedto a commercialrate of 21%.

ChOaer6: Fam Profiabilty Page 145 Soh dfkan Adtur,e: Struture. Performance& ODionsfor the Future

C. Interet subsidy schemes. There is a specialdrought assistancescheme for producersof winter grain and summercrops. The terms in 1991 were 5.5% for 5 to 10 years. The accumulatedtotal from 1981 to 1991stood at 1,077.8 million.

6.30 AMram. In spite of the generousfinancial assistance offered to insolventfarmers, loan arrears on interest owed mounted in the late 1980s as the farin financial crisis worsened. Interest arrears on long-term loans graned by the Land Bank remained relatively constant from 1970 through 1984, but increasedfrom 1.94% of interestowed in 1984to 6.4% in 1990. The Land Bank's capital arrears have remained relativelyconstant at 1-1.5% from 1970until 1991.

6.31 However, as outlinedabove, insolventor near-insolventfarmers are directedto the Agricultural Credit Board which managesthe FinancialAssistance Schemes. In 1993, the AgriculturalCredit Board reported that 64 % of the 17,000farmers servedby the schemeswere in arrears. Total arrears amounted to R1.4 billion, and most of it was owed by grain farmers. Such arrears were down considerablyfrom pre-1992levels, when the governmentinjected R2.6 billion in debt-reliefto grain farmers. Ihis reflects the government'spolicy to keep farmerson the land by, e.g., rollingover debt, infusingnew capital, and delaying foreclosureson property.

6.32 Se. Insolvenciesin agriculture remained relativelyconstant during the 1960s and 1970s (between40 and 118 per annum). In 1985 the total rose to 141, in 1986 to 243, and in 1987to 317, before decliningto 261 in 1988,201 in 1989, and 267 in 1990 (SouthAfrican Statistics 1990). The relatively constant level of sequestrationsin the 1980s, combinedwith the increase in short-term debt since 1983, seems to suggestthat the financial sector is essentiallypursuing a strategy aimed at keeping farmers on the land. Sucha strat may be fuelledby the fear that additionalliquidations would fiurher depress real estate prices; but It also seemsreasonable, given expectationsthat governmentwill continue to bail out insolventfarms.

FinancialAssistance to Farmers

6.33 The fundamentallack of profitability in many parts of the farm sector would have produced substantia declines in farm incomes,particularly during the 1980s, had it not been for massivestate aid carried out under the Financial AssistanceSchemes of the Department of Agricultural Development (1991)74.Moreover, massivedrought relief support in 1992 (a total of R3.4 billion) providedthe sector with a substantialfinancial injection. This assistancewas given because of the goverunent guarantee receivedby agriculturalfinancial institutions during the late 1980s.The 1992recapitalization concentrated governmentfinancial supportof agriculture in one year, but shouldbe seen as continuingover a much longer period.

6.34 During the 1980s, the state granted financialassistance in one form or another to some 27,000 (about one-third) of the country's white farmers (1991 Year Book).

'm b_nfo is povied in Anne=4.1 on the numberof loans, ata &sbursmnaitsof loan, and subsidis ad to ocial farmsk d e fdnanciaas asan for dheRSA, and for fams in the desiVated s (Land Bank data thlough30 Noavmber1991).

Chzapa6: Farm Projkabty Page 146 South Aftan AeLdcutr=: Strur. Perfo_me & Odions for theEuure a. Direct finmcialassistance to farmersover the period 1981182to 1991/92(Nov. 30) amountedto R1,728.1million; and b. Subsidiesto farmsover the sameperiod amounted to R2,353.6million. 6.35 Giventhe mid-periodestimate of the numberof all farms in the sector(65,880 in 1985),the averageloan assistanceamounted to R26,231per farm; and subsidiesaveraged R35,726 per farm (1981-91assistance). These programs, among other things,provide loans (at subsidizedinterest rates, e.g., 8% in 1991),as wellas subsidiesfor the following: a. interestof R1,077.8million;

b. meansof cropproduction-22,472 loans for R852.4million; c. consolidationof debt-8,104 loans totallingR558.2 millionand subsidiesof R99.6 million;

d. disasterand droughtrelief-3,397 loans for R88.1 millionand subsidiesof R540.6 million; e. floodassistance-1,163 loans for R1O.9million and subsidiesof R271.3million; f conversionof marginallands-subsidies of R133.7million; g. purchaseof agriculturalands-845 loans for R104.6million; h. emergencydrought schemes-subsidies of R104.6million;

L interestsubsidies on LandBank loans in designatedareas-R46.8 million; j. allocationsof stateland-925 loans for R41.0million;

k safeguardingof residents-subsidiesof R36.6million; 1. water works-555loans for R25.6million; m. soil conservationworks-805 loans for R18.1million; n. waterquota subsidies-R15.9 million; o. purchaseof livestock-440loans for R1O.4million; p. productioncost subsidies-R6.4 million; q. improvements-153loans for R3.8 million;

aupter 6: FarmPraft Page 14 SouhhAfrican Aariculture: Sture. Perfonmance& Ovtionsfor the Future r. purchaseof implementsand vehicles-121loans for 1.5 million;and s. sinkingof boreholes-150 loans for RO.4million.

6.36 The 1992/93'drought relief" pronm. 75 The structuraldecline of farmprofitability since the early 1980swas accompanied by a worseningdebt burden (real total debt withrespect to the real value of capitalassets) and an increasein therole of short-termdebt, both in absolute,real terms and in relation to net profits (Table6.7). An importantcomponent of such short-termcredit fell under a carry-over schemefor farmdebt whichwas guaranteedby the government.This programwas initially introduced as a temporarymeasure after the 1982/83drought, but becamea permanentfeature and escalatedafter the 1991/92drought when the guaranteerequired from the governmentrose froman initialR800 million in 1983to R1.4 billionin 1991,and finallyto R2.4 billionin 1992.

6.37 Thecarry-over debt program applied to grainproducers in the summerand winterrainfall areas. An investigationby the Departmentof Agricultureinto the debtposition of farmersestimated that the 33 cooperativesresponsible for 90% of grain productionin the countryheld a substantialproportion of members'total debtin the formof carry-overdebt. The studyalso foundthat beforethe drought28% of the membershad debt burdensexceeding the 50% limit abovewhich cooperativesdo not extend productioncredit to farmers,this ratiohad risento 45% after the drought.

Table 6.10: Farmers' Debts at the 33 CooverativesProducine 90% of al Grain in the Republicof South Africa- Before and After the 1991/92Drout

Before the drought After the drought Debt of members R bilioon R/ha R billion R/ha Carry-overdebt 1.4 218 2.4 375 Total debt 3.6 561 4.9 766 Share of carry-over 39% 49% debt

Source: Riner, 1993, citing Ministryof Agricultureand of Agricukual Development(1992:4).

6.38 The total financialaid packagewas based on the abovestudy and amountedto R3.4 billion, consistingof R2.4billion of debt-reliefand Rl.0 billionof droughtrelief. For the 6.4 millionhectares coveredunder the scheme,the debt-reliefprogram included R375/hectare to farmers(of which R175 was a subsidyto be appliedagainst debt carry-over and R100,on a sliding-scale,to cropdamage) and R640 millionto cooperatives,also as a subsidyto be appliedagainst debt carry-over. Thehomelands were includedin the remainingRI billion-R543million for droughtrelief, nutritional aid, overdraftfacilities, and other financialassistance. Not all of thesefunds, however, have been disbursed. 6.39 In effect,the 1991/92drought relief program constituted a substantialrecapitalization of the least

7' his seutionis basedon MartinRimner, 'Debt rief and the SouthAfrican drought relief programme: an overviw', Land nd Agriculture Policy Cate, Policy Paper 1, 1993.

waapter6: Farm Projltabily Page 148 Soea*AA*= Arkulue: Srcture. Pe,fon,we & Ontlonsfor the Future efficie sub-seto of the agricul sector. h seems clear that land prices have been shored up tempoaily and that the financialposition of the LandBank and the agriculturalcooperatives has been given temporaryrelief. At the sametime, however,the droughtprogram helped to keepsome of et labor employedin the sector. Ihe recapitalizationdoes not, however,influence the structr inefficienciesof significantparts of the agriculuralsector, in particularthe grainproducing areas; and it is likelythat a similar,but possiblyshorter, cycle of debtaccumulation and subsequentcrisis may now set in.

Chapte 6 Fat Proft. Page 149 Soxth Aftan Ayg*ayUre:Strmure. Parformawce& Optionsfor the Future

CHAFFER7: DIRECrIONS FOR THE FUTURE

Overview

7.1 'Ibis report has analyzedthe current conditionof the agriculture sector in South Africa from a number of different perspectives-its role in the macroeconomy,structure and endowments,historical development,marketing performance and policies,various aspectsof sectoralefficiency, and the funancial profitability. In this chapter, a short reviewof the main themesemerging from these chaptersis presented followedby an assessmentof some strategic options for the future.

Macroenomy

7.2 Over the last three decades, the macroeconomicperformance of South Africa has been marked by declininggrowth. At present, the economyis characterizedby capital-intensiveproduction patterns, however,this capital intensityhas been unable to generateeconomic growth and is unlikelyto be a source of growth in the future. Rather, this capital intensitywas associatedwith and contributedto declining growth, rising unemployment,and relatively insignificantgrowth in productivity. Apartheid, and the segregatedlabor markets it produced, created a situation in which modest growth of wages for the employedcould co-exist with increasingunemployment.

7.3 The inward-lookingpolicy framework adopted by the governmentis partly to blame for this disappointingperformance. Public sector policy strongly supporteddomesic producers throughtariffs, quotas and other barriers. As a result, many producers-including those in agriculture-have been insulatedfrom internationalcompetition and the need to increaseproductivity and re-allocateresources to their most efficientuse. Public sector investments-seeminglyin an effort to compensatefor sluggish private sector investments-wereoften undertakenwithout regard to the social rates of retum.

In this contex, It is not surprsing that agriculture's growth path has also been distorted-characterizd by a decliningshare of GDP, increasingcapital-intensity, and declining employment. Agricutural policies were aimedat assuringnational self-sufficiencyin basic agriculturalcommodities, while ensuringa price structure that aimedto guatee the profitabilityof the large-scalefarming enterprise. These policies were supportedby public expendituresin agriculturethat resulted in significantsubsidies to large-scale agriculture in order to ensure its financialsuccess.

Structure and History

7.4 Productionin the agriculturesector is dominatedby comparativelylarge, capital intensivefarms. These farms, which are mainly white-ownedand account for more than 86% of the agriculturalland, make the economy self-sufficientin basic agriculturalcommodities and generate income levels for its owners that are comparableto the urban sectors. As part of the effort to achieve this structure, the homelandswere used as sources of labor while black commercialagriculture was suppressedin order to mdnimizecompetition and increasethe supplyof inexpensivelabor.

7.5 The current conditionsof agricultureand the rural economyof SouthAfrica cannot be assessed without an understandingof the historical forces that contributedto them. After the initial period of

Chaper 7: Directonr &eFWare Page I" Sout Afric Aykubyg: Strucune.Parfomanw A Om fotthe Fur

European settlement,the profitabilityof whitefarming was graduallyincreased through a series of policy interventionsthat spanned more than a century. Initially,the main objective of these interventionswas to solve the difficultiesthat manyEuropean settlers had in attractingAfrican labor by reducingthe relative profitabilityof independentAfrican farming. This process involvedinterventions in several agricultural markets (and sales and renas, capital, labor, inputs, outputs, research and extension), and gradually transformedblack farmers from independentproducers into wagelaborers on white farmsand in the non- agriculturl sectorsof the economy. Thosewithout the possibilityor wherewithalto find employmentin the white-controlledsectors of the economyfound themselvesjoining rapidly expandinglabor reserves.

7.6 In contrastto the incomelevels earnedby white, large-farmowners, the standardsof livingamong blacks living in rua South Africa are very low. A recent estimateputs 45% of the overall population below the poverty line with the worst poverty found in the rural areas. There is no apparentdifference between the incidenceof poverty in the rural areas of the Republicof South Africa and the incidenceof poverty in the homelands. In both areas, abouttwo-thirds of blackslive in absolutepoverty. In the urban areas, the percentageof blacks living in poverty is about one-third.

Marketing

7.7 The agricultural marketing system-which has been an important element in defining the agricultura policy environmentin SouthAfrica-has been largelydominated by producerinterests. It has been cdhar zed by a complexnetwork of marketingregulations, arguably in an effortto accommodate specificproducer interests almoston a crop-by-cropbasis. An important characteristicof the existing marketing environment is the high degree of concentrationand regulation found in marketing and processing.

7.8 The marketingenvironment is also characterizedby a prominentdomestic orientation in the case of foodcrops and livestockproducts where domesticproducers have benefitted from a wide variety of protectionistpolicies mostly in the form of quantitativeimport controls and protective tariffs. Further, the combinationof controlledinput and outputprices along withsingle-channel marketing systems for the majority of agriculturalcommodities have restricted competitionand provided large-scalefarmers with a incomeguantee. An exceptionshoula be made here for certain export-orientedcrops suchias can be found in the horticulturalsb-sector. In general, the impactof the domesticorientation of marketing and processing aTrangementshas been to benefit producersat the expenseof consumers. The welfare implicationsof the agriculturalmarketing environment for both the urban and the rural poor have tended to be negative.

7.9 Sincethe early 1980s, the governmenthas attemptedto introducemarket oriented reforms in the marketingarrangements with the result that a considerable,but unevenamount of deregulationhas been achieved. As a consequence,agricultural producer prices of most commoditieshave fallen in real terms, yet consumer prices have increased. The resulting increase in marketing margins is, at least, partly caused by remahilng distortions in agriculturalmarkets and possibly the result of deregulatedoutput markets without concurrentderegulation in input, processing,and distributionmarkets.

Cwpter 7: DrecMorsar the Futwe Page 151 SouthAfrican Anricuiture: Sttcture. Perfornance & Ovijonsfathe Future

Pouction Effcig=c

7.10 Given the protection that agriculture received from internationalmarkets and the considerable influenceproducers had in the determinationof producer prices, it Is often difficultto analyzewhether producers allocatedtheir resources in the most efficient way. Considerableempirical evidenceexists, however, to suggest that-in response to the Drevailini structre of policy incentives-South African farmers made less-than-optmal use of labor and employed capital beyond its economic optimum. Moreover, given the inabilityof the sector to adapt its input bundle to the cost-price squeezethat has characteized the sector now for two decades, one could argue that there has becn a gradual deterioration of price efficiencydue mainly to increasedemphasis on non-priceincentives by the government.

7.11 Inionally, agriclture is generally characterizedby constant or declining returns to scale. In SouthAfrica, however, the selar trend towardslarger and fewerfarms seemsto suggestthe existence of Increasingreturns to scale. In large measure, this trend is not causedby genuine economiesof scale, but is the result of a policy environmentthat favoredthe establishmentand viabilityof farms that were large enoughto sustain a fill-tme white farmer at incomelevels comparableto those in the urban sector. This is essentiallya social policy objectivethat may carry high economiccost, since the availableevidence for SouthAfrica does not indicatethat increasingreturns to scale exist. Compellinginternational evidence on the efficiencyof small-scalefarming, combinedwith the existingempirical evidence for SouthAfrica, moreover, suggests that importantefflciency gains can be obtainedfrom a down-sizingof farm size in significantparts of the agriculturalsector.

7.12 Undoubtedly,South African agriculturehas demonstratedimprovements in technicalefficiency tha on average are modest by internationalstandards, but compare favorably to other sectors in the economy. These gains, however, have been uneven within the sector with some sub-sectorsachieving sgnificant gains in productivitywhile others have been disappointing. Moreover, the combinationof increasedtechnical efficiency and capital intensityhas exacerbatedthe slow growth of employmentlevels in the economy.

FinancialProfitabily

7.13 Since the early 1970s, agiculture in SouthAfrica has been exposedto a cost-pricesqueeze that has reduced farm profitability. This was compoundedduring the 1980s by: Ci)unfavorable weather conditions;(Qi) a reductionof farm subsidies;(iii) liberalizationof marketingpolicies that often resulted in lower real producer prices; and (iv) higher capital costs due in part to higher interest costs. Due to the prevailing stucture of non-price incentives many farms did not respond to the changing price environmentwith the result that they continuedto rely on the established,capital-intensive production methods. Given the variable agro-climaticconditions to which South African farming is exposed, the coninued reliance on mechanized,monocropped farming also entailed a bigher exposureto climaticrisk and increasedthe volatilityof farm profits.

7.14 As a result, the debt burden carried by many farmersescalated to unsustainablelevels during the 1980s,in particularfor farms in the summerrainfall areas (i.e. wheat and maize giowers). The financial position of many farmers has continuedto deteriorate,as illustratedby a continuingincrease in the debt- asset ratio. At the same time land prices have declined. Higher capital costs, tighter credit standards,

Ch0pr 7: Dedowjfbor she Fstre Page 152

.,

I

4~ ~ r So*h Afri Adgicukure:SSIurce. Performanwe & O&i_onsfor the Future fewer farm subsidies,and the growing unsustainabilityof the debt burden shouldhave positiveeffects on labor employment in the sector and force less efficient farmers into foreclosure. However, the governmentcontinued its attemptsto aid the sector financially,and farmers who could no longer obtain credit at commercialrates have had access to publicly-financedsoft credit programs and debt-relief measures. For example,in responseto the 1992drought, substantialamounts of debt accumulatedduring the 1980s, was written off through a drought relief program totaling R3.4 billion.

Efficiencyand Equity Issues in Agriculture

7.15 In summary, it is clear that the main theme to emerge from this report is that the agricultural policies of the past several decadeshave created a sector that is characterizedby numerousdistortions. These distortionshave led to inefficienciesin many parts of the sector. The policy reforms of recent years have exposedmany of the more inefficientfarmers to market pressures with the result that many farms presently face bankruptcy. At the same time, it is clear that many parts of the agriculture sector operate very efficiendy and that many of the large-scale farmers are competent managers who deal exceedinglywell with an uncertain agro-climaticenvironment, but who are faced with an increasingly harsh sectoral policy environment. Hence, the picture of large-scale farming that emerges from this report is an uneven one in terms of efficiencyand performancein recentyears. This has led to a situation in which the future of the sector is uncertainin terms of what constitutesan appropriateset of objectives and, in turn, policy environment.

7.16 Althoughthe questionof the efficiencyof large-scalefarming is an importantone for the fiture of the sector, it is not the only factor that needs to be consideredwhen formulatingstrategic options for the future. Even if it could be demonstratedthat all parts of the sector were highly efficient, there remainsthe undeniableneed for a land reform programas a matterof addressingas issueof socialjustice, specificallythe skewed distributionof land and asset ownership in the rural economy. International experiencein this regard indicatesthat the present highly inequitabledistribution of resourcesin South Africa-most notablyland-is not sustainable. Althoughinternational experience is not an infallibleguide to the future, it clearly suggests that economieswith a distributionof land similar to SouthAfrica's are prone to a patternof civil disorderand violence.7' Such prolongedperiods of conflicttypically result in capital flight, economicstagnation or decline, and possiblysocial and political disintegration. Hence, when examiningstrategic optionsfor the growth and developmentof the rural sector, one is confronted with three loomingrealities: (i) a large-scaleagricultural sector with a mixedhistory of performanceand which is going through a difficultstructural adjustment;(ii) the need for a meaningfulredistribution of rural assets-the most importantof which is land; and Ciii)the imminentchange in government.

7.17 Hence the broader issue is to design a strategy for South African agriculture and the rural economythat will achieve greaterdistributional equity, improvethe operationalefficiency of the sector,

16 In Zimbabwea fairly subtatl, but poody designedland refom programhas madeonly limtd conutins to oup and employmentgrowth, and left the land isue a an unresolvedpolitical and economsciility. Most other dualisic countries have fared nuch wrese Chile, Colomba, El Salvador,Guatemala, Nicaragua, Mozambique, Peru and Angolahave suffered prolongedpeiods of up to forty yearsof uneand civil war. Strongpeasant support for revolutionaryforces indicates tiat this was closelyrelabtd to thesecountries' delay or inabiliy to effectivelyredistibut landto a large numberof beneficiaries.

Chapter7: Directionsjfr ite Future Page 153 SoUshAfrican Arltufre: Sracture.Performace & Ontionsfor the Fzuure increase the level of employmentand foster growth that is both economicallyand envirommentally sustainable?

Options for Rural Restructurine

7.18 The combinationof raciallybased distortions in the old rural strategy, inequitiesin the distribution of and access to resources,and imminentpolitical change require that choices be made about the nature of a new strategy in order to ensurethat it will serve the needsof a muchbroader constituency. In light of international experience and the present circumstancesin rural South Africa, the conclusion that emerges from this report is that there are few strategic options for agriculture and the rural economy availableto a new Governmentin South Africa.

7.19 One option is to continueliberalizing the present policy environment,while leavingthe present structure of agricultureproduction units largelyunchanged. The process of ongoingliberalization would include: (i) further reform of the input and output marketing system, (ii) efforts to reduce the concentration in the agro-processingsector, (iii) revision of land sub-division guidelines, and (iv) restructuringof the present agriculturalcredit system. Continuedliberalization along these lines will likly result in more bankruptcies among large-scale farms, the expansion of small-scale farming (especiallynear urban areas), the further expansionof the horticulturalsector, and contractionof cereal and livestockproduction. Althoughthe appealof a more efficientagricultural sector through continued lealization is clear, the drawbackto this option is that changesin the ownershippattern of the farming sector-and especiallythe large farm subsector-are likely to be very slow. Ihe main reasons for this are that: (J) in an environmentof uncertaintyover land ownership, few large-scalefarmers will be willingto make the investmentsnecessary to respondto the new policy environment;and (ii) few of thosewho were disfranchised under apartheid will have the resources needed to gain access-either through lease or purciase-to land currently held in the large-scale farm sector. Hence, while this option offers some increased efficiency,it is only a partial solutionto the problem at hand because it essentiallyaccepts a large-farm structure which-as this report demonstrates-is inefficient in many critical subsectors and bighy inequitablein the distributionof assets.

7.20 A second strategic option would be to expand the focus of the first option to include a rural developmentstrategy for the homelandsthat wouldfeature upgrading of agriculturalsupport services and Investingin improvedphysical and social infrtructure. This has the advantageof continuingthe process of policy liberalizationand of concentratingpublic sector resourceson some of the most obviousvictims of apartheid. Such an option would be likely to contributeto a significantexpansion of agricultural production in the homelands,but one that is limitedto those comparativelyfew areas in the homelands with good agriculturalpotential. An obviousdrawback to this approachis that-given the fragile condition of many of these areas-any effort to intensifyagricultural production in the homelandsis likely to have undesirableenvironmental consequences. A further problem with this option is that while it seeks to address the poverty in the homelands,the efficiencygains are incompleteand the problem of distributional Inequity remains largely unaddressed. In addition,this optionsdoes nothingfor the significantnumber of disenfanchised citizens that reside and/or are employedin the large-farmsector.

7.21 A third strategic option builds on elementsof the two optionsdescribed above-continuedpolicy

Chqpter7. Directionsfor the Future Page 154 South Afrcan derkuluure. Strueu-re.Perormancw A Opttonsfor the Future reform and investmentin the homelands-and in additionsupports a redistributionof agriculturalland In the large-farm sector. A clearly formulatedpolicy of redistributionof agriculturalland would achieve three criticalobjectives: (i) reduce uncertaintyexperienced by current owners thereby encouragingthose who continuefarming to invest; (ii) address the present inequitabledistribution of land access; and (di) encourage,if the redistributionmechanism is properly designed,those with the greatest interest in land use to gain access to and use land efficiendy. This latter objective would lead to a more dynamicrural economy and to greater employmentand income creation among low-incomegroups than would either of the first two strategicoptions. Hence, it is the third strategicoption that merits further elaboration, especiallyits redistributiveelements.

Overviewof Land ReformOptions

7.22 The startingpoint for developinga set of land reform options is an examinationof experiencein this area. Among others things, the lessons of internationaland SouthAfrican experiencedemonstrate that a successfulland reform program needs to: (i) be implementedexpeditiously in order to avoid a combinationof bureaucraticinertia, legal challenges,and the power of present landowners(generally the elite) that are likely to render the program ineffective;(ii) be politicallyacceptable and legitimatein order to forestall efforts to reverse the reform in the future; (iii) have a clear role for the public and private sectors defined in such a way that beneficiariescan exerciseas much choice as possible-programsthat have relied entirely on the public sector in the belief that the sector is the only one capableof maintaining integrity, delivering services, determiningneeds, and managingthe process have been failures in most instances;and (iv) be part of a clearly articulatedbroader strategy of economicgrowth.

7.23 At the outset, it needsto be recognizedthat wherepressure existsfor a land reformprogram there are three basic methods for landless people to gain access to land in any country: (i) invasion; (ii) restoration; and (iii) redistribution. Where either restoration or redistributionfail to provide land, invasion is often seen as the only option availableto the landless population. A continuationof land invasionswould present a new SouthAfrican government with an exceedinglydifficult problem. It would mean that the credibilityof existinglegal institutionswill have been underminedbecause of their inability to implementeffectively the land restitutionand redistributionprocesses. Consequently,people whohave invadedland are unlikelyto respondto courtorders to vacateand forced evictionsmay become necessary. Given the recent history of SouthAfrica, howevar, it is highly unlikely that a new governmentwill be able to carry out such evictionswithout causing massive political alienation. In order to avoid large-scale land invasions,an eilfectiveprogram of land restorationand redistributionis necessary.

7.24 The land restoration(or land claims)process involvesthe return-by means of an administrative or adjudicativeprocess-of specific parcels of land to individualsor communitieswho were unjustly removed in pursuance of racially based land legislationor policies. The land redistribution process involvesproviding access to land for selectedgroups or individualsand supplyingthem withthe necessary support services to use that land effectively.

7.25 Assumingthat a equitableand effectiveland restorationprocess (e.g., land claims court) will be implemented,the remainingIssue is the design and implementationof a redistributionprocess. One of the central tensions in designingthe desired land redistributionmodel is between the desire to address welfare and asset transfer objectives through the redistributionof land and the need to promote the

Chapter7. Directionsfor the Ftue Page 155 SouthAfiJM A&,g&jLcrA-:Strwure. Perfonnaws & Otflonsfor the Fulr productiveuse of agriculturalland. It is often the casethat thoseindividuals who qualifyfor landor assistanceunder welfare objectives of a landredistribution program, have little experience in agriculture or relatedland use activities.In contrast,the mostexperienced and well-qualifiedfarmers typically do not qualifyto receiveland under welfare objectives. Hence, designing a programto addressone of these objectiveswill oftencompromise the other objective. 7.26 Althoughboth of theseobjectives can be accommodatedwithin a singleprogram, in orderto do so it is essentialthat there be agreementon the precise objectivesand methodologyof the land redistibutionprogram. Amongthe salientelements that characterizea successfulprogram are that:

a. market-assistedland redistributionprograms tend to perform better than those administeredand operatedby the publicsector; b. the roleof the publicsector in a landredistribution program centers on ensuringadequate suppliesof land in the marketand monitoringthe overalloperation of the program; C. criteriafor participationare necessaryand mustbe discussedand agreedin advanced; d. welfareobjectives can be metby includinga grantcomponent in the program; e. a matchinggrant schemethat forcesparticipants to use someof their own resourcs in order to gain accessto land will help to assist in self-selectionof participantsand encouragethe productiveuse of land; f. the grantelements of the programare essentialin orderto accomplisha redistributionof assetsand to ensurethat beneficiariesemerge from the programwith a net increasein their asset positionand low debt/assetratio as a means of ensuringviability and sustainabilityof their enterprises; g. in additionto addressingthe fundamentalissue of socialiustice, these options are likely to significandyincrease M ruralemployment and ensurethat the cost of the progrm is veryreasonable;

h. a redistributionprogram will not be able to provideland for everyoneand the program willneed to be complemer.cedby a rural safetynet and by programsfor urbangroups, and i. althoughthe discussionconcenaes on agricultureand smallfarmers these are not the limis of the program. It is envisionedthat the programwill be a vehiclefor supporting a wide range of land use activities,including trading activities,and small-scale enterprises.

Choir 7: Direcionsjfr dh Fsture Page 156 Soth Atfrian Aerkubure:Strmture. etb,MM A Optionfor the Fuure

Conclusio

7.27 As stated at the outset, the purpose of this report is to analyzethe forces that have contributed to the current state of the agricultur sector In South Africaand to beginthe process of identifyingsome strategic options for the future. 'Me major conclusionof the report is that significantchanges iti the structure, ownership, and operationof the agriculturesector are necessaryin order to address the serious problemsof efficiencyand equity. the analysisand options set forth in this report are intendedto serve as the basis for discussionand debate within South Africa on the lessons of history and policiesfor the future. In South Africa, this debate is essential for the success of any strategy, because as has been stressed many South Africans, the process is as importantas the product.

Chapter7: Direcibsfor th Ftr Page 157 Sa Afica Agpbre Smoture.hrfon,Mace & Ontionsfor the Funtre

REFEREDNCES

Ardington, AJ. "Private sector and other possible initiatives in promoting nrual development.' Devlopment SouthernAfrica 7 (Specialissue) (October, 1990), pp. 603-614

Ball, V.E. 'Output, Input andProductivity Measurement in US Agriculture,1948-79." AmerlcanJournal ofAgriculural Economnics,Vol. 67, No. 3, (1985) pp. 475-86.

Ball, J. 'Small is BeautifulThe Key Concept for SuccessfulIrrigation Development.' Development SouthernAfrica, Vol. 4, No. 2 (May, 1986).

Beaumont,W.H., S.W. Burger, W.R. Collins,W.E. Stanford,and C.H. Wessels. Rejort of the Natives Land Commision. Vol 1. (CapeTown: Cape Times Ltd., GovernmentPrinters, 1916)

Beinat, W. 'The PoliticalEconomy of Pondoland, 1860-1930.' (CambridgeUniversity, 1983).

Bekker, S., C. Cross, and N. Bromberger. "Rural Poverty in South Africa: A 1992 study using secondarysources. (Durban: Universityof Natal, 1992).

Bell T. and V. Padayachee. 'Unemploymentin SouthAfrica: Trends, Causes and Cures.' Development Sou5ewn4frIa 1(3& 4) (November,1984).

Bembridge,TJ. 'Car tics of small-scalefarmers in the Transkei'. Social Dynamics, Vol. 2, No. 2 (1986a).

. 'An Overviewof agriculturaland rural developmentin less developedareas of SouthernAfrica,' Dewlpment SouthernAfrica, Vol. 3, No. 1 (1986b).

. 'The Practice of AgriculturalExtension: A Training Manual.* (HalfwayHouse: Development Bank of SouthernAfrica, 1991).

-. 'Practical Guidelinesfor Agricultura ExtensionWorkers: A Field Manual.' (HalfwayHouse: DevelopmentBank of SouthernAfrica, 1991).

'.Crop Farming System Constraints in Transkei: Implicationsfor Research and Extension.' DeuelopmentSouthern 4frlca, Vol. 4, No. 1 (February, 1991), pp. 67-152 .

Berry, R.A. and W.R. Cline, Agrarian structureand producdvity In developing countries. (Geneva: Intrational Labor Organization,1979).

Biggp, F.P. 'Aspects of CombiningCapital and Labour in South Africa, with SpecialReference to the Modem Agriculture.' Unit for Future Research,Report No. 82/4 (Stellenbosch: Universityof Stellenbosch,1982)

Rdermenu5 Page 158 SowhAfta AgfrltSurw:Str. Performae & Onionsfor the Future

Binswanger,H.P., K. Deininger, and G. Feder, "Power, distortions, revolt and reform in agricultural land relations,"in T.N. Srinivasanand J. Behrman(Eds.), HandbookofDevelopmentEconomics, Vol. IIf, forthcoming.

Binswanger,Hans and K. Deininger, 'South Africa Land Policy: The Legacyof History and Current Options," WorldDevelopment, Vol. 21, No. 9 (Oxford: September,1993).

Brand, S. S., N.T. Christodoulou,CJ. Van Rooyen and N. Vink 'Agriculture and Redistbution: A Growth with Equity Approach." (Halfway House: DevelopmentBank of Southern Africa, Corporate CommunicationsDivision., 1992)

Brand, S. S. and F.R. Tomlinson "Die plek van die landbouin die Suid-AfrikaanseVolkshuishouding." South AfricanJournal of Economics,Vol. 34 (1966)pp. 2649.

Brand, S. S. 'A Critical Evaluationof the ExistingDevelopment Agencies and Iheir Role During the Reconstructionand TransitionPeriod: Is There a Need for Reconstruction?"(Durban: Annual Conventionof Associationof Black Accountantsof SouthernAfrica, 1966)

-. *The Conributions of Agricultureto the EconomicDevelopment of South Africa Since 1910" (D.Sc. dissertation.Pretoria University, 1969).

Brotherton, I.A. and J.A. Geoenewald. "OptimalOrganisation of DevelopedIrrigation Farms in the Malelane-KomatipoortArea." DevelopmentSouthern AfWca,Vol. 7, No. 1 (1982).

Bundy, C. ^The Rise and Fall of the SouthAfrican Peasantry." (Berkeleyand Los Angeles:University of CaliforniaPress, 1979).

Bureau of Market Research. "PersonalIncome of the RSA-TNVCCountries by PopulationGroup and Magieri District." (Pretoria, 1989).Paper preparedfor the Finance2000 Workshop,May 27- 28, 1993. Photocopy,(Madison: Universityof Wisconsin).

Cairns, R.I., and ID. Lee. 1990. DevelopmentSouthern Afrca, Vol. 7, No. 1 (February, 1990), pp. 77-104.

Carter, M.R., . "Does it Take More than Liberalizationto GenerateBroadly Based Growth? Financial Markets and the Microdynamicsof Agrarian Growth Transformation." Photocopy. (Madison, WI: Departmentof AgriculturalEconomics, University of Wisconsin, 1993).

Carter, M.R. and J. Kalfayan, "A GeneralEquilibrium Exploration of the AgrarianQuestion," photocopy (Madison, WI: Departmentof AgriculturalEconomics, University of Wisconsin, 1989).

Chavas, J.P. and Aliber, Michael. "An Analysis of Economic Efficiency in Agriculture: A NonparametricApproach." Photocopy(Madison, Wisconsin: Collegeof Agriculturaland Life Sciences, 1993)

Rqfereces Page159 Soud AfricanAegricuure: Structure. PAffonnance & Ovtons for the Future Chavas, J.P. and J. van Zyl. 1993. forthcoming

Caristodoulou,N.T. and N. Vink. 1990. -The Potential for Black SmallholderFarmers Participation in the SouthAfrican AgricultureEconomy." Paper presentedat the Conferenceon Land Reform and AgriculturalDevelopment, Nowick Park Initiative, United Kingdom.

Cobbett,M. "The Land Questionin SouthAfrica: A PreliminaryAssessment." Presentedat Workshop on the SouthAfrican Agrarian Question. (Witwatersrand:University of Witwatersrand:May 24- 26, 1987).

Cross, C. 'Land Tenure in Black Rural Areas: Social and Political Underpinnings."Paper presented to Idasa Workshopon Rural Land Tenure, Botsrivier,Cape (March 1990).

Data Research Africa. "Micro-Enterprisesin Rural Areas of South Africa." Results of a Survey on Farm and Non-farmEnterprises in Bophuthatswana,Kwazulu, and Transkei." Report prepared by Leanne Lickin, Mike Lyne and Julian May. de Kiewiet, C.W. A Hstory of SouthAfrica-Socal and Economic. (London: Oxford UniversityPress, 1942). de Kterk, M. J. (ed.) MTheAccumulation Crisis in Agriculture." In South Africa's Economc Criss, Stephen Gelb, (Cape Town: David Philip, 1991). de Klerk, MichaelJ. (ed.) A Hwvest of Discontent: lhe Land Queston In South Africa (CapeTown: Institute for DemocraticAlternatives in South Africa, 1991).

"The Labour Process in Agriculture: Changes in Maize Farming in South Africa During the 1970's." SocidalDynamics, Vol. 11, No. 1 (1985) pp. 7-31.

Technologial Change and Employment In South African Agrculture: The Case of Maize Haesting In the Westernlhnsvaal, 19681981. UnpublishedM.A. (Economics)dissertation. (Capetown: Universityof Capetown, 1983)pp. 195-234.

DevelopmentBank of SouthernAfrica, . "An Inter-RegionalProfile." (HalfwayHouse, 1991).

- A RegionalProfitle of The SouthernAfrican Population." (HalfwayHouse, 1990).

-. "An Inter-RegionalProfile." (HalfwayHouse, 1990). de Wet, C. "LandTenure andRurat Development:Some Issues Relating to the Transkei/CiskeiRegion," Dew copenSouthern Afica, Vol. 4, No. 3 (1987).

Duncan, A. and Jones, S., "AgriculturalMarketing and Pricing Refbrm: A Review of Experience," Worl Deweopment,Vol. 21, No. 9 (Oxford: September, 1993).

Rfermnc Page 10 South Afi Agrdcubre: Structure. Performance& ODdonsfor JheFuture Dusmnanitch,V.Y. and M.A.G.Darroch. 'An EconomicAnalysis of the Impactsof MonetaryPolicy on SouthAfrican Agriculture." Agrekon, Vol. 29, No. 4 (1990)pp. 269-283 Elliott, M.B. 1992. Personalcommunication.

Fallon,P.R. 'An Analysisof Employmentand WageBehavior in SouthAfrica." Informaldiscussion papers on aspectsof the economyof SouthAfrica; Paper no. 3. (Washington,D.C.: Southern AfricaDepartment, World Bank, 1992). Feder, 0. Ihe Relationbetween Farm Size and Farm Productivity:the Role of Family Labor, Supervision,and Credit Constraints,"Journal of DevelopmentEconomics, Vol. 18, 1985, pp. 297-313. Fenyes,Tom and Johanvan Rooyen,1985. "SouthAfrican Agriculture and Migrant Labor." PhilipL. Martin(ed.) MigrantLabor in Agriculture:An InternationalComparison (Washington, D.C.: GianniniFoundation of AgriculturalEconomics, University of Califomiaand the German MarshallFund of the UnitedStates). Food and AgricultureOrganization (FAO), ProductionYearbook, (Rome,1989). FoodStudies Group. "Agricultural Marketing and Pricing in South Africa." Draft. (Oxford: International DevelopmentCentre, University of Oxford,1992) Fuler, K.G., and M.A.G.Darroch. "Factors Associated with CreditUse by SummerCrop Producers in the Transvaaland OrangeFree State." Proceedings:Annual Conference of the Agricultural EconomiesAssociation of SouthernAfrica (1988) pp. 336-344. (roenewald, J.A. "Retuns to Size and Structureof Agriculture:A SuggestedInterpretation. DeweapmentSothern Alffca, Vol. 8, No. 3 (August,1991) pp. 329-342.

. "Reviewof the Reportof the Committeeof Enquiryinto Alternative Marketing Arrangements for Maize." Dewlopment SouthernAJiWca, Vol. 6, No. 4 (November,1989).

Groenewald,J.A. and G.F. Liebenberg. "Die RSA-Landbouruilvoet."Agrekon, Vol. 29, No. 3 (Pretoria,1990) pp. 178-184. Groenewald,J.A. and J.P.F. du Tolt. "SomeFeatures of LivestockFarming in Bophuthatswana." Devlopment SouthernAfrica, Vol. 4, No. 2 (1983) pp. 225-242.

Halbadh,AJ. 'The SouthAfrican Homeland Policy and Its Consequences:An Evaluationof Separate Development." DevelopmentSouthern 4frica, Vol. 5, No. 4 (1988) pp. 508-526.

Hattngh, H.S. "SkeweInkomeverdeling in die Landboutse Uitdagingaan Landboubelaid,"Agrekon (Pretoria:1986). Translatedby Viljoen,M.F..

Rgfereca Page 161 Soh AfrkcanAricuture: Structure.Performance & Ovtionsfor the Future

Hatting4 H.S., and A. Herzberg. 'Ownership or Leasing of Agricultural Land: Production Economic Aspects." Agrekn, Vol. 19, No. 2 (Pretoria:1980). Howeroft,J. R. 1992?. "An EconomicAnalysis of AlternativeMarket Scenarios for Wheatin South Africa." Photocopy. Jaffee,S. "AgriculturalDiversification and Tradein SADCCCountries." Photocopy.(Washington, D.C.: SouthernAfrica Department, World Bank, 1991). Janse van Rensburg,B.D.T. and J.A. Groenewald."The Distributionof FinancialResults Between FarmersDuring a Period of AgriculturalSetbacks: Grain Farmers in the WesternTransvaal, 1981/82."Agrekon, Vol. 26, No. 1 (Pretoria: 1987)pp. 13-19. Joubert,J. andJ.A. Groenewald.1974. 'Land TenureSystems in WhiteSouth African Agriculture: Ihe Importanceof Different Systems." Agrekon, Vol. 13, No. 3 (Pretoria: 1974) pp. 3-9.

Kahn, B., Senhadji, A., and Walton, M. "South Africa: Macro-economic Issues for the Transition," mimeo (2nd draft), January, 1992.

Kassier, W.E. et al., "Report of the Committee of Inquiry into the Marketing Act," known as lhe Kasser Report, (Stellenbosch: 1992)

Kan, T. Rural transformations in industrialkiing South Africa: The Southern Highveld to 1914. (Johannesburg: Ravan Press, 1986).

Lacey, M. "Working for Boroko: The Origins of a Coercive Labour System in Southern Africa." (Johannesburg: Raven Press, 1982).

Levy, B. "How Can South African Manufacturing Efficiently Create Employment? An Analysis of the Impact of Trade and Industrial Policy." (Washington, D.C.: World Bank: Southern Africa Department, 1992).

Lenta, G. and 0. Maasdorp, "Limit Factors on Food Production in the Homelands," in C.R. Cross and R.J. Haines (eds.), Towards freehold? Options for land and development in South Africa's black rural ar. (Cape Town: Juta and Co., 1988).

LIebenberg, G.F. and J.A. Groenwald. "Die RSA-Landbouruilvoet." Agrekon, Vol. 29, No. 3 (Pretoria: 1990) pp. 178-184.

Lipton, Merle. "South Africa: Two Agricultures?" in F. Wilson et al. (eds.) Farm Labour In Sowh Africa (Cape Town: David Philip, 1977).

Capitalism and Apartheid (Totowa, NJ: Rowman & Allanheld, 1985).

Rerewce Page162 Sout Afia Arrcubeur: Stbuly. Parfonawe 0 vons fort h Future Lipton,M. and M. Lipton. "SomePolicy and InstitutionalFactors Affecting the Impactof Agriculture on the Poor: Lessonsfor SouthAfrica from Experience Elsewhere." World Developmen, Vol. 21, No. 9 (Oxford:September, 1993).

Lipton.Michael. WhyPeople Stay Poor. (London:Temple Smith, 1977). Lodge,T. 'Perceptionsof AgrarianIssues in BlackPolitics." DevelopmnentSouthern Africa, Vol. 7 (Specialissue) (October, 1990) pp. 483494.

Louw, A.R. "enPerspektief op die aanbodvan voedselin Suid-Afrikamet spesialeverwysing na voedselsekerheid.Unpublished M.Sc.(agric) dissertation, (Pretoria: University ofPretoria, 1990) Louw, A.H. and J. Van Zyl. "'n Perspeltiefop voedselselfvoorsienendheidas 'n determinantvan voedselsekerheid."Agrekon, Vol 30, No. 3 (Pretoria: 199)pp. 129-138. Lubbe,W.F. "NormalisingCompetition in AgriculturalMarketing-the Case for Red Meat in South Africa." Agrelon,Vol. 30, No. 4 (December,1991) pp. 253-259. "Alternativesfor Privatizationand Deregulation."Report to Sumyslde Groqp. (mimeo). (1990). TheDecomposition of Price lime SeriesComponents of the Red Meat Industryfor Efficient Policyand Marketing Strategies." Agrekon, Vol. 29, No. 4 (Pretoria:December, 1990) pp. 296- 304.

Maddison,A. lhe World Economy In the 20th Centwy, (Paris, 1989) pp. 81, 91.

MaizeBoard. 1991. AnnualReport. Annexes. (No publishinginformation). Marcus,T. 'Restrucuringin CommercialAgriculture in South Africa, Modernising Super-Exploitation." Dr. GovanMbeki Fund (Amsterdam: University of Amsterdam,1986). McKenzie,C.C., D. Weiner,and N. Vink. "LandUse, AgriculturalProductivity and Farming Systems in SouthernAfrica." Unpublishedresearch report. DevelopmentBank of SouthernAfrica (HalfwayHouse: 1989)

Moll,P.C. "Modelsof LandReform." Development Southern Africa, Vol. 7 (SpecialIssue) (October, 1991)pp. 441-50.

-. "TheRationality of Farm Size Growth:An Examplefrom 'White' SouthAfrica. European Review of Agrutural Economies,Vol. 16, No. 3 (1989) pp. 345-357. "Economiesof Size in "White"South African Agriculture: Implications for Land Reform." Unpublishedpaper (Cape Town: University of CapeTown, 1988).

Rfreces Page 163 SouthAfrin ArkICuue: Structure. Performance*& Orions for theFLtture

Moll, T. "Output and ProductivityTrends in South Africa: Apartheidand Economic Growth." PhD thesis, (Cambridge: CambridgeUniversity, 1990) pp. 110-114.

_. "Did the ApartheidEconomy Fail?" Journalof SouthernAfrican Studies, Vol. 17, No. 2 (June, 1991)pp. 271-291.

Moolman,J.H. 'Historical Perspective- the Process of Occupation.",and "Consolidation.*(Chapters I and 4 of unidentifiedms.).

Morris, M.L. "The Developmentof Capitalismin South Africa Agriculture: Class Struggle in the Countryside." Economy and Society, Vol 5, No. 3 (1976) pp. 292-343

Nicholson, C.A., and T.J. Bembridge. "Characteristicsof Black CommercialFarmers in Gazankulu.' South frlca Journmlof AgriculturalExension, Vol. 7, No.17. ( 1991).

Nieuwoudt,W.L. 1990. 'Efficiency of Land Use." Agrekon, Vol. 29, No. 4 (Pretoria: December, 1990)pp. 210-215.

"Farm Household-Economiesand IncreasedEarnings from TraditionalAgriculture: Implications to SouthernAfrica." (mimeo)(Unpublished paper, 1988).

- An Economic Analysis of Demand and Policies in the Beef Industry." Agrekon: Biannua Journal on Agricultural Economics, Vol. 24, No. 2 (Pretoria: Department of Agriculture, October, 1985).

"An EconomicAnalysis of demand and supplyfunctions for wheat(bread) in SouthAfrica, 1948- 1981." Agrekon: Biannual Journal on Agricultural Economics, Vol. 22, No. 2 (Pretoria: Departmentof Agriculture,October, 1983).

Nieuwoudt,W. L., and N. Vink. "The Effects of Increased Earnings from TraditionalAgriculture in SouthernAfrica." South Aftican Journa of Economcs, Vol. 57, No. 3 (1989) pp. 257-269.

Ortmann, G.F. "Impactof AvailablePolicy Optionson ConsumerWelfare." Agrekon. Four-Mondty Journal on AgriculturalEconomics, Vol. 25, No. 2 (Pretoria:Department of Agricuture, June, 1986) pp. 42-44.

Payne, N. L., J. van Zyl, and H. J. Sartoriusvon Bach. "Labour-RelatedStructural Trends in South African CommercialGrain Production: A ComparisonBetween the Summerand Winter Rainfall Areas, 1945-1987." Agrekon, Vol. 29, No. 4 (Pretoria: 1990) pp. 407 416.

Plaatje, Sol. Native Life in Souh Africa. (London: Longman, 1987).

Prosterman,, Roy L. and Jeffrey M. Riedinger. Land Reforn and Democratc Development.(Baltimore and London: The Johns Hopkins UniversityPress, 1987).

Roftreces Page 164 Souh Aft Au : Strucurs. Ermnc. OAD=t Abe Fuure

Rosenzweig, Mi. and H.P. Binswanger, 'Wealth, Weather Risk and the Composition and Profitablity of Agriculral Investments," Economic Journd (forthcoming).

Roth, M., H. Dolay, and K. Wiebe. 'Employment, Efficiency and Land Markets in South Afica's AgrAcuLtl Sector: Opportnites for Land Policy Reform' draft (Madison: Land Temnre Center, May 21, 1992).

Roukens de Lange, A., 1990. 'Use of the Social Accounting Matrix in Investigating the Informal Sector of the South African Economy.' Institute for Futures Research, University of Stellenbosch, mimeo.

Sartorius von Bach, H. J. and J. van Zyl. 'Comment: Returns to Size and Structure of Agriculture - A Suggested anpretation. DevelopmentSouth Africa, Vol. 9, No. 1 (February,1992) pp. 75-79.

Sartorius von Bach, H. J., J. van Zyl, and B. H. Koch. "Managerial Ability and Economies of Size in South African Agriculture: Empirical Evidence for Structural Adjustment.' Paper presented at the annual conference of the South African Society of Agricultural Extension, Bloemfontein, May 5-7, 1992.

Schoeman, J.L. and D.M. Scotney. 'Agricultural Potential as Determined by Soil, Terrain and Climate.' SouthAfia Jouwn of Science,Vol 86 (1987)pp. 395402.

Simkins, C. 'Financing Rural Development.' Development Southem Africa 7(Special issue) (October, 1990) pp. 591-601.

-. 'Black Population, Employment and Incomes on Fams Outside the Homelands Revisited.' Paper presented to IDAKSA Rural Workshop, (Capetown: March, 1990).

'What Has Been Happening to Inoome Distribution and Poverty in the Homelands.' Development Southemn ca, Vol. 1, No. 2 (1984).

South African Agricultura Union (SAAU). Die Finansiele Posisie van Boere in die R.S.A. (Pretoria: SAAU, 1984).

South Africa (Republic of). 'Labour Statisdcs: Employment and Salaries and Wages Wholesale, Retail and Motor Trade and Hotels: September 1991.' Statistical News Release, February 11, 1992. rewtoria: Central Statistical Service, 1992).

Abstract ofAgiculurl Sadstics (Pretoria: Central StatisticalService, 1993)

South Afrcan Stattcs, 1992 (Pretoria: Centra Statistical Service, 1993).

Abstract of Agrcultural Stttcs (Pretoria: Central StatisticalService, 1992).

Re dec Pase 16S Su* A*at Aricukure:SrItaurfF PerfonnanceA On onsfor the Future

'An Investigationint the Price Mechanismin the Food Chain,' Board on Tariff and Trade, Report No. 3273 (Pretoria: Departmentof Agriculture, 1992)

DirectorateofAgrcultural EconomicTrends (Pretoria: Departmentof Agriculture, 1992)

h"Ivestigationinto RestrictivePractices that Exist or May Arise in the Supply and Distribution of Fertilizer to BonaFide Farmers"and "Whetheran AgreementBetween Indian OceanFerdlizer (PTY) LTD, AECI Limited, AECl OpencastServices (PTY) LTD and Ktnoch Fertilizer LTD Gave Rise to an Acquisitionand, if so, Whether Such Acquisitionis Justified in the Public Interest,' CompetitionBoard, Report No. 31. (1992).

SouthAfrican Stadscs, 1992 (Pretoria: Central StatisticalService, 1993).

Annual Report. Pretoria: Departmentof DevelopmentAid, 1990).

Directorateof AgricidturalTrends. (Pretoria:Department of DevelopmentAid, 1990).

Annual Report. Pretoria: Departmentof DevelopmentAid, 1989).

Census of Leaing to Own Fred Property,1986, Report No. 83-02-01 (Pretoria: Central StatisticalService, 1989).

Commission- of Enquby Into the Red Meat PriceDiscovery Process. (Pretoria: Report of GovernmentPrinter, 1989)

Agriadtural Survey (Pretoria: Cental StatisticalService, 1988).

CensusofAgriculture (Pretoria: Central StatisticalService, 1983).

Cesus of Agriculture(Pretoria: Central StatisticalService, 1972).

CensusofAgriculture (Pretoria Central StatisticalService, 1963).

Census ofAgriculture (Pretoria: Central StatisticalService, 1953).

Departmentof Bantu Administrationand Development.Central Statistical Service. Pretoria. (unpublisheddata)

SouthAfrica (Union of). 'Land Survey, Tenure and Occupation,' in OfficialYear Book of the Union. No. 22 (Pretoria: The GovernmentPrinting and StationaryOffice, 1956) pp. 493-514.

-Land Survey, Tenure and Occupation,' in OfficialYear Bookof the Union, No. 22 (Pretoria: Ihe GovernmentPrinting and StationaryOffice, 1941) pp. 623-644.

Page 166 SoudhAfrican Arkuturw: Sncturw. Perfonnaw A Ovdonsfor d*eFuture

*Land Survey, Tenure and Occupation,"in Official Year Book of the Union, No. I (Pretoria: The GovernmentPriting and StationaryOffice, 1916) pp. 357-376.

South African Reserve Bank. SouthAfrica's NationalAccounts. (Pretoria: June, 1991).

QuarterlyBulletin. Various years/issues. Pretoria.

South African Sugar Association.

Stadler, J.J., J.P. van der Heever, and J.A. Lombard. The Costs, Advantages and Financing of Protectionin South Africa: the Influenceof Protectionon Agriculture." First interim report; in Afrikaans. Research Report no. 17. (Pretoria: University of Pretoria, Bureau for Economic Policy and Analysis, 1983)

Stewart, D.A. and M.C. Lyne. 'Socio-EconomicCharacteristics of the Rural Populationin Gcumisa Ward, KwaZulu." DevelopmentSouthern AfWca, Vol. 5, No. 2 (May, 1988) pp. 186-195.

SunnysideGroup. 1991. 'The Red Meat Industry: Assessmentand Recommendations.'Photocopy.

Tapson, D. R. "TheAgricultural Potential of the Homelands: Problemsand Prospects," in H. Giliomee and L. Schlemmere(eds.), Up Againstthe Fence (1985).

Thirtle, C. and P. Bottomley, Total Factor Productivity in UK Agriculture, 1967-90, Journal of Agrkulural Economics, Vol. 43, Nol, 3 (1992) pp. 381-400.

ITirde, C. H. Sartoriousvon Bach, and J. van Zyl. Exlaining Total Factor Productivit Growth in South African CommercialAgriculture. 1947-91,Working Paper (93/02) (Pretoria: University of Reading, Departent of AgriculturalEconomics and Management,March 1993).

"ExplainingTotal Pactor ProductivityGrowth in SouthAfrica CommercialAgriculture, 1947- 91." Working Paper, (Reading: Universityof Reading, Departmentof AgriculturalEconomics and Management,1993).

Trapido, S. "Landlord and Tenant in a ColonialEconomy: The Transvaal. 1880-1910." Journal of SouthernAfian Studies, Vol 5, No. 1 (1978) pp. 26-58.

United Nations. Centre AgainstApartheid, Department of Politicaland Security CouncilAffairs. Lhi Tenure Conditionsin SouthAfrica. (Rome: 1976)

United States Departmentof Agriculture. 1990. EconomicResearch Service, internal data. van Heerden, W.R. and J. van Zyl. "The AggregateMeasure of Support for Maize in South Africa, 1985/86to 1990/1." Photocopy. (Pretoria: Universityof Pretoria, 1992).

Refereue Paeg167 SozahAfrc Aewcakue: SOwtwr. Prfommanc Opilonsfor the t van Rooyen, CJ. 'Agricultural Restrucuring in SouthernAfrica: Ihe Contributionof the Developing Agricultua Sector.' Agrekon,Vol. 29, No. 1 (Pretoria: 1990) pp. 3-10.

'Using Results from ExperimentalFarms in Planning Irrigation Schemes for the Fams at Eartebeestpoorm' unpublished M.Sc. (Agriculture)dissertation. (Pretoria: University of Pretoria, 1973) (in Afrikaans). van Rooyen,CJ. and J. van Zyl. "AgriculturalRestructuring in SouthAfrica: Changesin the Production Structure and Some ConsiderationsOn the Food Question.' Paper prepared for Inational Conferenceon AgriculturalRestructuig in SouthernAfrica. Sakopmund,Namibia: International Agricultura EconomicAssociation. (1990).

'Agricultural Production in South Africa: An Overview.' Unpublishedpaper, presented at IDASA 'Rural Land' Workshop (HoushoekInn, 9-11 March, 1990). van Rooyen, CJ., van Zyl, J., and T.I. Fenyes. 'A Comparisonof the Contribution and Relative Performanceof Agriculturein SouthernAfrica.' DevelopmentSouwhern Africa, Vol. 4, No. 2 (1987) pp. 183-198. van Seventer, F. and van Zyl. 'An Input-OutputAnalysis of Agribusinessin South Africa.' Agrekon, Vol. 31, No. 1 (Pretoria: March, 1992)pp. 12-21. van Schalkwyk,H.D., and J.A. Groenewald. 'An Estimateof Aggregate Supply Response in South African Agriculture." (unpublishedpaper, 1992).

'Agricutural Land Price and Quality.' (unpublishedpaper).

'Price of AgriculturalLand and Quality.' DevelopmentSoahern Africa (forthcoming).

'Regional Analyses of South African AgriculturalResource Use and Productivity.* Agrekon, Vol. 31, No. 3 (Poria: September, 1992) pp. 116-127. van Schalkwyk,H.D., and J. van Zyl. 'Ihe SouthAfrican Land Market: An Analysisof Land Prices.' forthcoming. van Wyk, S.P. 'Trends in Land Valuesin SouthAfrica.' Agrekon, Vol. 6, No. 1 (Pretoria: 1967) pp. 23-30.

'Detennination of an EconomicFarming Unit.' (7) pp. 18-29. van Zyl, Johan. 'Practical Approachesto Land Reform - 1: Values - Economic Reform in SA." Fo?mer'sWeekly (July 26, 1991)pp. 41-43.

Modelsfor Land Use.' Famer's Weekly (Augus 2, 1991) pp. 32-33.

Rfrnc Pase 168 Souh 4*ican Agriculture:Structure. Performane Odtionsfor the Future

*Agriculture'sPotential for Growth." Farmer's Weekly (Marct. 15, 1991)pp. 34-35.

-Productivity and the Bottom Line - 3: The Grain Industry." Fanner's Weekly. (August 17, 1990)pp. 19, 21.

. 'interrltonships in Maize Markets in Southern Africa I: Structural Effects of the Farmer Support Programme.' DevelopmentSouthern Africa, Vol. 6, No. 2 (May, 1989)pp. 202-270.

-Te Effect of Inflationon AgriculturalProduction under Conditionsof Risk." Agrekon, Vol. 25, No. 3 (Pretoria: 1986)pp. 56-63.

Reportingon Herman van Schalkwyk's M. Com. dissertation at the University of Pretoria. Farmer's Weekly (September11 & 18, date). van Zyl, J. and J.A. Groenewald. 'Flexibility in Input Substitution: A Case Study of South African Agriculture," DevelopmentSouthern Africa, Vol. 5, No. 1 (1988)pp. 2-13. van Zyl, J., A. van der Vyver, and J.A. Groenewald. 'The Influenceof Drought and General Economic Effects on Agriculture: A Macro-Analysis.' Agrekon, Vol. 26, No. 1 (Pretoria: 1987) van Zyl, J., A. van der Vyver, and C.W. Mostert. "'The Effect of Debt Burden, Interest Rates and nflation on the Survival of Farming Enterprises: A Case Studyin the Western Transvaaland the North-WesternBushveld.' Agrekon, Vol. 26, No. I (Pretoria: 1987) pp. 1-7. van Zyl, L., HJ.G. Nel, and J.A. Groenewald. "Agriculture's Contribution to the South African Economy.' Agrekon, Vol. 27, No. 2 (Pretoria: 1987)pp. 1-9. van Zyl, J. and J. van Rooyen. 'Agricultural Restructuringin South Africa: Changesin the Production Structure and Some Consideratio-. On the Food Question." Paper prepared for International Conferenceon AgriculturalRestructuring in SouthernAfrica (Sakopmund, Namibia: International Agriculturd EconomicAssociation, 1990).

. 'Agricultural Productionin SouthAfrica: An Overview.' Paperpresented at an IDASAworkshop on Rural Land, 1990b. van Zyl, J. and N. Vink. 'Effects of Farmer SupportProgrammes on Consumptionand Investment.' DevelopmentSouten Africa, Vol. 9, No. 4 (November, 1992).

Vink, N. and W.E. Kassier. 'Agricultural Policy and the South African State," in A Harves of Discontent: the Land Ouestion in South Africa, M. de Klerk ed. (Cape Town: Institute for a Democratic Alternativefor South Africa, 1992)

Viljoen, P. and J.A. Groenewald. 'An approach to farming efficiencyanalysis as applied in Ruens, Agrekon, Vol. 16, No. 4 (1977) pp. 6-8.

R*,rene Pace 169 SouthAfrican Agnculture: Structure.Performance & ODdonsfor the Future

VolkskasBank. "CapitalEfficiency in Agriculture," EconomicSpotlight (February, 1991).

_Relating Perceptions and AssociatedEconomic Criteria to EconomicSurvival In Commercial Dryland Farming In South Africa." Draft of unpublishedpaper.

Weiner, J. "LandUse, AgriculturalProductivity and FarmingSystems in SouthAfrica." Draft Report, 1989.

Wheeler, M.W. & O.F. Ortmann. "Socio-economicFactors Determiningthe SuccessAchieved Among Cotton-AdoptingHouseholds in Two MagisterialDistricts of KwaZulu." DevelopmentSouthern Africa, Vol. 7, No. 3 (August, 1990)pp. 323-333.

Wilson, F. 'Farming, 1866-1966.' OxfordHistory of SouthAfrica, M. Wilsonand L. Thompson,eds., Vol. 11 (1971) pp. 104-71.

Wilson, F. and M. Ramphele.Uprooting Poverty: The South African Challenge (Cape Town: David Philip, 1989).

World Bank. "An Economic Perspective on South Africa.' (Washington, D.C.: Southern Africa Department,World Bank, 1993).

"Venezuela: Land Markets, Land Reform, and Rural Land Ownership." (Washington,D.C.: Latin Americanand the CaribbeanDepartment, World Bank, 1992).

- World DevelopmentReport 1992: Developmentand the Environment. Table 33: (Oxford, New York: Oxford UniversityPress, 1992) pp. 282-83.

-_ "Outlineof Work Briefsfor ExternalConsultants." ZimbabweForestry Sector ReviewMission. (Washington,D.C.: World Bank, 1991).

-. "Uruguay: Agricultural Sector Review." Vol. 1, Main Report. (Washington,D.C.: Latin Americaand the CaribbeanRegion, World Bank, 1988).

.Jordan: Towards an AgricultureSector Strategy." (Washington,D.C.: CountryDepartment m, AgriculturalOperations Division, Europe, Middle East and North Africa Region,World Bank, 1990)

. "Kenya: Growth and StructuralChange." Basic EconomicReport. (Washington,D.C.: World Bank, 1983)

Wright, Peter David. A Study in PoliticalEconomy: The South Afi.can Maizeand Dairy Industry. Ph.D. Dissertation (Pietermaritzburg: University of Natal, Departnent of AgriculturalEconomics, 1992).

Rferecwn Page 170 &AndArI A&p*ur: &ractur. Perf,ce A Olioonsfor the Futar#

ANNEX I

Table 1.1: UnemploymentAmong the EconomicallyActive Population

RegisteredUnemploymaent Coloreds Asians Blacks (%) ' (%) (%) (%) Whites Coloreds Asians Male Female Male Fnemale Male Female

A B C D E P 1975 0.2 0.6 0.9

1976 0.3 0.7 1.0

1977 0.6 1.6 2.1

V78 0.7 1.7 2.2 6.7 17.7 1979 0.6 1.4 1.9 6.6 12.4 6.0 15.6 1980 0.5 1.1 1.S 4.1 9.8 5.3 14.9 1981 0.3 0.7 1.0 3.6 5.9 5.0 13.8 1982 0.3 0.9 1.0 4.5 6.0 5.1 13.7 1983 0.7 1.8 2.7 7.0 8.5 6.4 9.6 6.0 14.1

1984 0.7 1.1 1.3 5.2 7.8 5.5 9.2 5.4 13.5

198S 1.3 2.6 3.8 7.9 8.5 6.7 10.4 5.4 12.7

1986 1.6 2.9 4.2 9.2 12.9 8.0 15.0 15.4 27.2

1987 1.4 2.5 3.3 13.0 15.8 10.7 16.1 11.6 23.8

1988 1.1 2.0 3.2 9.7 11.5 8.5 11.3 10.6 18.9

1989 7.3 9.0 6.1 7.3 8.3 16.3 a. Colms A-C are regisred unemployed. b. Data in colmns D-F are the unemploymentrate based on the current populaion survey (whits excluded). Sounao South Afican Saistis 1990,pp. 7.34 and 7.41.

Anna I Page 171 Souwh fdc Akricu.re: Structwtre.Perorawe A Onons for the Fwure

Table 1.2: lhends in oexhaneoras (US S 2ew mand:and indices:gcgod averaues.A

Ma*ue NominalEtrectvc Real EffecdveEBchange Commerci Rate ExchangeRate Rate (1985 100) (SIR) (1985=100) 1978 1.15 - 1979 1.18 - 1980 1.29 - 1981 1.15 - 1982 0.92 - 1983 0.90 - 1984 0.70 147 132 1985 0.46 100 100 1986 0.44 79 92 1987 0.49 78 104 1988 0.44 68 98 1989 0.38 62 99 1990 0.39 58 102 1 3rd Q 0.35 54 106

buri bntenationalFinaenl Sistics, InternationalMonetary Fund, WashingtonDC. Various issues.

Annex)IPage 172 So)h Aftan Asrkuiczt,re:&KIcre. Prftnormancea Optionsfor the Future

Table 1.3: Shares and Rate of Growthin General GovermmentExpenditures'

1978-82 1983-87 1988-92 1991 1992 EIIc_tase Shtaresof GovernmentExpenditures Recurrentexpenditure 70.3 80.2 86.3 83.3 89.0 Remunerationof employees 26.1 30.7 31.42 32.7 63.6' Other goods and services 23.9 2S.S 28.73 30.0 Interest payments 8.3 10.5 13.0 10.9 11.9 Subsidies 4.S 6.2 S.6 5.0 4.3 Other current transfer 7.2 7.2 7.0 5.9 7.6 Capitalexpenditure 20.1 18.2 12.4 14.0 9.9 Net Lending 9.6 1.6 1.3 2.7 1.9 Toal expenditure 100.0 100.0 100.0 100.0 100.0 GC)Pshare of governmentepnitur 31.3 33.0 35.7 37.2 37.0

Real Rates_ of wt (constan 1=_ urioc Current expenditure 3.7 6.1 3.7 1.2 5.6 Remunerationof employees 4.1 7.1 S.22 11.4 4.8' Other goods and serices 3.2 5.S 5.21 -1.4 Interest payment 11.7 8.9 6.1 -4.8 7.4 Subsidies 3.7 9.S -4.4 -9.7 -14.3 Other current transfer 3.4 2.8 3.8 -12.2 27.0 capital epenitu -1.1 3.8 -7.3 17.9 -3S.0 Total expenditure 2.3 3.1 2.6 8.3 -1.0 GDP 4.0 0.2 1.4 -0.7 -0.6

~gI~: Calculationsbased on SARBdata. Fscal year endingMarch 31.

Notes: i. Dat are in fiscal years, e.g. fiscal year 1978 refers to fte year precedingMarch 31, 1978. 2. 1987/88 - 1990191 3. Refer. to 1991/92 total for remuneation of employeesand other goods and services.

Annec.1 Page 173 South Africn A rulure: Structure. PerformanceA Optionsfor the Future

Table 1.4: MacroeconomicIndicators

ExchangeRates Annual Short to GovemmentBond Land BankLong-Term Inflation MediumTen Yiold Interest Rate LendingRate ($/R) SDRslR) (CPI) (%) Nominal Real Nominal Real (%) (%) (%) (%) A B C D E F=E-C 0 H=G-C 1990 0.39 0.29 14.4 21.0 16.2 1.8 16.1 1.7 1989 0.38 0.30 14.7 19.8 16.9 2.2 14.6 -0.1 1988 0.44 0.33 12.9 15.3 16.4 3.5 13.7 0.8 1987 0.49 0.38 16.1 12.5 15.3 -0.8 13.7 -2.4 1986 0.44 0.38 18.6 14.3 16.4 -2.2 13.9 -4.7 1985 0.46 0.45 16.3 21.5 16.8 0.5 13.3 -3.0 1984 0.70 0.68 11.5 22.3 15.2 3.7 10.0 -1.5 1983 0.90 0.84 12.4 16.7 12.7 0.3 10.0 -2.4 1982 0.92 0.84 14.7 19.3 13.5 -1.2 10.0 -4.7 1981 1.15 0.97 15.2 14.0 13.0 -2.2 7.0 -8.2 1980 1.29 0.99 13.8 9.5 10.1 -3.7 7.0 -6.8 1979 1.19 0.92 13.2 10.0 9.3 -3.9 7.0 -6.2 1978 1.15 0.92 11.0 12.1 10.4 -0.6 7.0 -4.0 t977 1.15 0.99 11.0 12.5 11.0 0.0 7.0 -4.0 1976 1.1S 1.00 11.2 12.3 10.4 -0.8 7.0 -4.2 1975 1.37 1.13 13.5 11.8 9.7 -3.8 7.0 -6s 1974 1.47 1.22 11.6 10.2 9.0 -2.6 6.0 -5.6 1973 1.44 1.21 9.4 8.0 7.8 -1.6 6.0 -3.4 1972 1.30 1.20 6.5 8.8 8.4 1.9 6.0 -0.5 1971 1.40 1.39 6.4 8.8 8.4 2.0 6.0 -0.4 1970 1.40 1.40 5.1 8.2 7.2 2.1 6.0 0.9 1969 1.39 1.40 2.9 8.0 6.5 3.6 6.0 3.1 1968 1.39 1.40 1.8 8.3 6.5 4.7 6.0 4.2 1967 1.39 1.40 3.0 8.3 6.5 3.5 6.0 3.0 1966 1.39 1.40 3.8 7.5 6.3 2.5 6.0 2.2 196W 1.40 1.40 3.9 6.9 5.6 1.7 6.0 2.1

a. Periodaveag mak exchange ae of U.S.S per Rand (IMF Statistics). b. Piod average SDR valueof the Rand (IMF Saistics). Prior to July 1974, theSDR wasvalued at thepar value of the dollar. BeginningJuly, the value of the SDR is determineddaily on the basis of a baskt of currencies.

Annx I Page 174 SouthAfi&M A,rkuhare: Struaure. PerformanceA Optionsfor the Future

Table 1.S: Inflation. Real Interest Rates and Real Price Changes. 1953.1991

Year Inflation Land Bank Co-ops Commercial Weightedtotal Real Realiput Real land banks producer prics price A A,B,C pices B C (%) (%) (%) (X) (%) (%) (%) (%) chan_ cange chng 1955 1.6 3.90 6.40 4.40 4.96 -0.40 0.50 -0.20 1960 1.4 4.33 6.88 5.08 5.50 -0.72 4.59 9.11 1965 3.9 2.08 4.58 3.08 3.31 -0.74 -3.14 -0.3S 1970 5.1 0.94 3.44 3.44 2.70 -1.63 -0.68 -13.39 1971 6.4 -0.42 2.08 2.58 1.52 -3.10 -1.52 5.70 1972 6.5 -0.53 1.97 1.97 1.23 7.91 0.80 -1.13 1973 9.4 -3.43 -0.93 -1.07 -1.72 16.27 1.13 3.39 1974 11.6 -5.64 -2.64 -0.56 -2.80 1.00 5.78 4.27 1975 13.5 -6.51 -5.01 -1.66 -4.29 -4.93 8.02 13.94 1976 11.2 -4.22 -1.97 1.16 -1.55 -1.80 4.13 -8.15 1977 11.0 -4.01 -1.76 1.49 -1.29 -1.56 1.62 3.92 1978 11.0 -4.02 -0.77 -0.16 -1.52 -4.93 1.71 -16.21 1979 13.2 -6.15 -3.65 -3.55 -4.36 S.99 7.55 -6.30 1980 13.8 -6.82 -4.24 -3.8S -4.86 4.86 4.00 14.39 1981 15.2 -8.22 -1.87 0.48 -2.93 -3.21 -4.07 9.78 1982 14.7 -7.72 1.55 3.88 -0.38 -3.54 1.61 -0.32 1983 12.4 -2.3 2.02 6.73 2.36 0.92 1.50 -0.50 1984 11.5 2.46 4.33 12.49 6.62 0.69 -3.71 -0.92 1985 16.3 -2.28 -4.26 1.00 -1.84 -9.44 2.77 -8.37 1986 18.6 -4.60 -7.46 -5.61 -5.97 -4.90 0.70 -13.36 1987 16.1 -2.10 -2.25 -2.89 -2.43 0.08 -6.80 -10.13 1988 12.9 1.15 4.79 4.83 3.72 -0.29 0.11 -4.87 1989 14.7 0.83 4.08 6.05 3.80 -9.16 5.29 -2.93 1990 14.4 1.63 4.38 6.38 4.27 -6.02 -2.42 2.75 1991 15.3 -0.31 3.38 4.80 2.78 -6.31 -4.89 -15.64

Aam I Page 175 Souh AfricanAgricklture: Structure. Performance A OvIlonsfor the Future

ANNEX 2

Table 2.1: Number and Area of Earms

1953 1963V7 1972 1983 1988 Number of Farming Units Cape 43,576 38,156 30,767 23,468 25,160 Natal 11,514 9,645 8,202 6,044 6,770 Transvaal 39,024 34,874 28,363 19,511 23,260 Orage Pree State 25,084 18,712 14,603 10,937 11,820 Totl 119,198 101,387 81,935 59,960 67,010 Total Area of Faming Units ('000 ha)3 Cape 55,970 58,448 56,978 55,251 55,209 Natal 4,488 4,622 4,388 4,291 4,124 Transvaal 15,727 16,293 14,687 14,631 14,634 Orange Frec State 11,823 12,008 11,863 11,842 11,801 Total 88,008 91,371 87,916 86,015 85,768 Owned Area ('000 ha) Cape - 51,664 47,691 45,235 45,641 Natal - 4,113 3,764 3,460 3,475 Transvaas - 13,957 11,684 11,150 11,290 Orngi Free State - 9,645 8,780 8,499 8,622 Total - 79,379 71,919 68,344 69,028 Rented or Leased Area ('000 ha) Cape - 6,78471 8,58S 9,721 8,946 Natal - 508 577 803 601 Tranavas - 2,335 2,560 3,281 3,025 Oange Frec State - 2,362 2,650 3,123 2,896 TOal - 11,989 14,372 16,928 15,468 SharecroppedAa ('000 be) Cape - - 703 295 623 Natal - - 46 27 47 Tanvaai - - 443 201 319 Orange Free State - - 431 220 284 Total - - 1,623 743 1,272

Source: 1953, 1963, 1972, and 1983 Censuaof Agrculture; 1988 AgriculturalSurvey.

Originaldata in morgen(1 norg - .8566 ha).

^ Total area = owned + retd + sharecroppedarea.

7 Fues for 1963 for Rentedor Leased Area include leased and sharecroppedland.

Annex2 Page 176 South AI*c Azrkuure: Sruchwe. Pe_formnnae& Optwoayforthe Fbre

Table 2.2: Size (Number' Distrbution of Fams

Hectas. 19S0 1962 Hectae 1978 1983 1988 Classifiation (percent) (ponent) Clasdfic (perce*) (pert) (pert)

< - 4 ha 8.40 S.28 < - 2 ha 0.44 0.31 0.18 S - 17 ha 9.40 11.88 2 - 4 ha - 1.99 1.67

18 - 86ha 11.90 14.41 5 - 9 ha 7.39 2.79 2.64 87- 428 ha 32.10 28.55 10- l9 ha - 3.29 3.00 20 - 49 ha 13.14 7.61 7.98 SO - 99 ha 7.17 6.50 6.57 100 - 199 ha 8.67 8.26 9.67 200 - 299 ha 6.96 6.69 6.92 429 - 857 ha 17.50 15.97 300 - 499 ha 11.10 10.81 11.21 858 - 1713 ha 11.00 11.48 500 - 999 ha 16.53 16.88 17.04 1714 - 2570 ha 3.80 8.33 1000 - 1999 ha 13.04 15.70 14.36 2571 - 4283 ha 3.10 2000 - 4999 ha 10.52 12.88 12.33 4284 - 8566 ha 2.10 2.96 5000 - 9999 ha 3.61 4.46 4.453 - 8566 ha 0.70 1.14 > - 10000 ha 1.43 1.82 1.98 Tota 100.00 100.00 Total 1.00 1.00 1.00

Note: Ori classifiation in 1950 and 1962 was in morgens (1 morgen = .8566 ha).

-Soure Roth, et aL, 1992.

Annex 2 Page 177 SoWhAfrc Aak*bm Arucue. POrfonnanceA Optios &l FUu*re

Table 2.3: Si Aal Ditnbution of Farms

Hoctare 1950 1962 Heotae 1978 1983 988 Classifica pent (prcent) Classificton (peren) (peren) (ecnt) < - 4h - - - 2 ha 0.00 0.00 O.O0 - - 2 - 4 ha - 0.00 0.01 1 - 171h - - 5- 9 h 0.03 0.02 0.02

- - 10 - 19 ha - 0.03 0.05 18 - 86 h 0.90 0.81 20 - 491h 0.28 0.17 0.22 50 - 99 ba 0.43 0.33 0.47 87- 428ha 10.44 8.68 100 -199 ha 1.06 0.84 1.15 200 - 299 ha 1.42 1.14 1.35 429 - 857 ha 14.76 11.11 300 - 499 ha 3.63 2.96 3.36 858 - 1713 ha 18.07 15.65 500 - 999 ha 9.90 8.51 9.16 1714 - 2s70 ha 10.94 25.02 1000 - 1999 ha 15.33 15.51 15.03 2571 - 4283 h 14.26 - 2000 - 4999 a 27.21 27.93 27.22 4284 - 8566 ha 16.67 1.96 s500 - 9999 ha 20.78 21.25 20.20 - 8566 ha 13.96 18.77 > - 10000 ha 19.92 21.31 21.77 taol 100.00 100.00 Tota 100.00 100.00 100.00

~: Roth, et aL, 12.

Ann 2 Page 178 o Afa Agrlre: Suure. Performace & O&Aonsfor the Fture

Table 2.4: Distributionof Arable Land Holdings in OcumisaWard. KwaZulu

KwaZulu KwaZulu Nata Dry Bushveld KwaZulu KwaZulu Mistbelt Area Newspaper Nqunquma o 0.8 3.3 0.01 - 0.25 ha 10.6 11.5 0.26 - 0.50 ha 18.9 22.7 0.51 - 0.75 ha 22.7 19.7 0.76 - 1.00 ha 11.4 18.0 1.01 - 1.2S ha 11.4 3.3 1.26 - 1.50 ha 6.8 1.6 > - 1.50 ha 17.4 16.4 < - 1.00 ha 27.7 7.5 1.00 - 1.90 ha 18.1 30.0 2.00 - 2.90 ha 28.6 6.0 3.00 - 3.90 ha 12.4 3.0 4.00 - 4.90 ha 8.6 43.0 5.00 - 5.90 ha - 4.5 6.00 - 6.90 ha 1.8 1.5 4.00 - 7.90 ha 1.8 1.5 > 8.00 ha 1.0 3.0 Total 100.0 100.0

~gg~: Stewartand Lyne, 1988, p. 191; and Cobbett, 1984, p. p. 371.

Annex2 Page 179 South African Agrlcadure: &trucuru.Performance & Opions for the Future

Table 2.5: Trade in Selected AericulturalCommodities-lmpoTts

Imports ('000 t) 1990 Import Value Import (RMn) 1970 1980 1970 Value 1990 1980

Mae 200.1 4.7 3.4 - 3.2 5.8 Wst 118.3 4.8 584.4 - 0.8 228.8 Barley 14.0 0.0 0.0 1.2 0.3 0.6 Sugar 12.1 32.9 30.1 1.4 6.2 28.5 Beef 24.4 0.9 6.6 10.9 1.3 7.1 Sheep 0.0 0.0 0.9 0.0 0.0 1.6 Pork 2.0 0.5 1.3 0.9 0.5 8.4 Poultry 0.0 3.4 20.1 0.0 3.7 44.5 Wool 3.5 1.4 2.0 2.9 4.4 15.8 Hides/Skin 5.5 2.0 0.3 1.7 2.3 2.3 Mobair 0.0 0.1 0.0 0.2 1.0 0.0 Coffee 15.1 14.8 17.1 - - - Tea 17.4 18.6 14.0 - - -

Annex 2 Page 180 South -AficanArricukure; Structure, Perfonnance& Optionsfor the Future

Table 2.6: Trade in SelectedAericultural Commodities - Exports

Export Volume (OOOMT) Export Value (Rand Mn) 1970 1980 1990 1970 1980 1990 Maize 1218 3663 2001 54 421 607 Wheat 0 120 146 0 19 58 Sorghum 80 149 10 3 18 2 Barley 0 26 1 0 4 1 Sugar 790 976 1014 48 397 729 Apples 122 179 202 20 64 201 Grapes 28 42 79 4 18 133 Apricots 0 0 2 -- 0 5 Pears 28 42 79 4 18 133 Peaches 1 1 2 0 1 7 Plums 2 6 11 1 6 28 Oranges 200 377 305 24 91 245 Grapefruit 28 81 62 S 24 59 Lemons 4 0 32 1 0 29 Narges 0 0 3 0 0 3 Avocadoes 2 13 30 1 9 44 Mangoes 0 1 3 0 1 5 Beef 28 34 19 13 57 104 Sheepmeat 1 0 0 0 0 0 Pork 5 2 2 3 3 8 Poultry 0 22 2 0 IS 18 wool 93 - 56 64 159 538 Hides/idns 51 0 40 22 61 247 Ostich 0 0 0 1 4 0 Mohair S 0 3 S 26 39

Source Deparztent of Agriutr

Annev 2 Page 181 South Afrkan Agriculture: St.ucture. PerformanceA Options for the Future

Table 2.7: Percentaueof CultivatedArea Under PrlncipalField Crous. 1955/56-1990191 ( percent of area cultivated)

Year Maize Wheat Oats Sunflower Sugar- Coton Gtain Barley C000ha) Ae Seed came Sorghum Cultivated (percentof farn area) 1990/91 43.7 22.4 10.1 3 5.A 1.8 1.7 1.6 6917 1989/90 45.8 24.1 9.1 6.9 4.9 2.2 1.8 13 7582 9 1988/89 47.7 25.1 7.2 5.2 4.8 2.6 2.3 1.0 7917 9.0 1987/88 47.6 22.5 6.5 6.0 S.0 2.7 3.4 1.3 7686 9.4 1986/87 49.9 23.8 5.6 4.7 5.0 2.0 3.9 1.1 8077 9.4 1985/86 50.1 24.1 S.4 4.0 5.1 1.6 3.8 1.2 8093 - 1984185 49.3 24.2 6.2 3.9 5.1 1.4 4.0 1.1 ml 9.2 1983/84 51.0 22.9 5.7 4.1 5.2 1.3 3.6 1.0 7892 9.1 1982/83 51.0 24.7 S.8 4.1 5.1 1.3 2.3 0.8 7990 9.0 1981/82 S3.6 22.8 5.0 3.9 5.0 1.4 2.2 0.9 7829 9.6 1980/81 S4.4 20.7 4.6 4.9 4.9 1.S 2.S 0.9 784S 9.6 1979/80 51.8 22.8 5.1 4.0 45 1.4 2.9 1.3 8337 9 5 9.7 1978/79 52.7 23.2 4.9 4.S 4.4 1.2 2.5 1.2 8168 9.s 1977/78 52.7 21.9 5.1 6.6 4.4 0.0 3.2 1.1 8174 9.S 1976/77 S3.6 23.4 5.9 4.7 4.1 0.0 2.7 1.1 8306 9.7 1975/76 55.9 22.6 S.7 3.S 4.2 0.3 2.6 0.9 8139 8.5 1974/75 S4.7 22.7 S.1 2.9 4.1 0.6 3.1 0.8 8203 9 1973/74 S2.S 23.8 S.1 2.8 3.8 0.0 4.0 0.7 849S 93 1972/73 48.6 27.1 S.6 4.7 43 0.0 2.9 0.7 7433 9.2 1971/72 54.8 24.1 4.5 23 3.8 0.0 3.9 0.6 883S 93 1970/71 S3.S 23.S 4.5 2.2 4.0 0.3 4.6 0.5 8226 - 8.7 1969/70 S1.1 22.4 7.6 2.3 4.0 0.0 4.0 0.7 8247 8.S 1968/69 S43 20.7 6.9 2.0 4.1 0.0 3.2 0.8 8074 9.1 1967/68 S8.9 16.2 7.8 2.0 4.2 0.0 33 1.0 =026 - 1966167 S7.8 14.3 6.9 1.8 43 0.0 8.1 0.8 7934 - 1965/66 553 17.7 7.5 23 43 0.0 6.0 0.9 7663 8.2 1964/65 SS.S 17.1 7.2 1.8 3.8 O.S 6.2 0.8 773S - 1963/64 S4.3 18.6 93 2.0 3.1 0.0 4.5 0.9 8171 - 196V63 S3.8 17.0 8.3 2.2 3.0 0.4 6.2 0.8 8067 - 1961/62 SS.8 18.1 9.3 2.0 3.3 0.3 3.4 0.7 7778 6.7 1960/61 54.6 16.6 9.0 2.5 3.4 0.2 4A 0.7 7538 19S9 56.8 IS.9 7.4 2.3 3.4 0.2 43 0.6 7452 1958159 S7.9 17.2 6.2 3.1 4.0 0.3 3.0 0.6 6287 1957/58 S5.6 21.8 6.2 2.2 3.9 03 3.0 05 6113 1956157 54.9 20.8 7.2 2.1 3.7 03 2.7 0.6 6312 19SS/56 58.0 193 6.9 1.8 3.9 0.4 2.1 0.6 5851

Area cultivatedrefers to the aea unde' principa field crops rpoted inlthe Absta of Agricultul Staiics, 1992. Source: Abstact of Agculturl Sti'stics, 1992.

Annex 2 Page 182 Sottth .tganARrkgltUr_: Structure. Performance & Opgtonsfor the Future

Table 2.8: Growth Rates in AcriculturalCommodity Areas. Productionand Yields. 1I970-989

70-79 Area Production Yield 80W89 70-89 70-79 80-89 7089 70-79 80-89 7089 Republicof South Africa Maizm 0.18 -1.86 -0.93 3.46 0.82 -0.22 3.28 2.73 0.71 Wheat -0.78 0.83 -0.27 2.87 4.68 2.64 3.68 3.82 2.92 Oats 1.25 5.47 1.71 -4.50 -8.09 -5.86 -5.69 -12.86 -7.44 Sunflower 8.70 4.38 3.25 11.25 6.46 3.63 2.34 1.99 0.37 Sugarcane 1.77 -0.45 1.31 3.62 2.31 1.33 1.82 2.78 0.03 Barley 11.07 4.36 3.19 21.07 13.75 11.77 9.00 9.00 8.32 Cotton 8.09 16.83 7.45 6.36 -0.59 Sorghum -4.43 -0.28 -1.64 1.32 2.03 -0.39 6.02 2.31 1.27 Soya Reans 18.29 9.89 8.62 38.07 17.93 15.75 16.73 7.32 6.57 Groundauts -3.8. -8.94 -3.58 -4.42 -0.37 -5.54 -0.62 9.41 -2.04 Dry Deans 1.00 1.25 -0.86 1.33 6.37 1.40 0.33 5.06 2.28 Rye -1.10 1.17 1.16 -6.15 -14.41 -5.43 -5.11 -15.39 -6.51 Tobacco 1.43 -3.24 -2.99 3.93 -1.14 -0.18 4.10 2.17 2.59 Cowpeas -20.18 2.53 -11.29 -5.26 -20.40 -11.86 18.69 -9.71 3.65 Chicory 9.21 -4.05 0.84 0.87 -1.65 1.60 -2.22 2.50 0.70 Buckwhea 0.15 -3.46 -19.07 -16.97 -3.61 Hay 5.64 -0.57 3.00 Dry Pexs 4.46 2.06 1.15 Lentlus 10.21 -0.72 -1.44 Hoaticurc 2.12 2.54 2.75 states,a Mawe -9.63 7.19 0.37 Dry Beans 8.01 Sorghum 7.01 Wheat 3.35 Groundnuts 9.85

Esiaswerederived *om lea egresson. Not al estmates are sautstay smiicanL RSAbuckwbh dudonct and cowpe area and yield figurs thugh 1987 only; chicoryarea and yield figure ecxlude1979;, bacco rea and yield figu fiom 1972 only. NationalStates exclude Tandki from 1976, Bophutetwata from 1977, Vendafrom 1979,and Ciskei fiom 1981; wheat productionfigure exchue 1983.

u Roth, et aL, 1992.

Aw= 2 Page 183 South AfricandgrkIuImLre. Srucrturt. Performance& Optlonsfor the FuPture

Table 2.9: Maize Production.Area Planted. Yield and Value. 1980/81 to 1992/93

Season Production Area Yield Value White Yellow Total 1980181 3847 6633 10480 4322 2.42 1.28 1981/82 6702 7721 14423 4338 3.32 1.93 1982/83 4269 3993 8262 4278 1.93 1.28 1983/84 2030 1974 4004 4065 0.98 0.68 1984J85 1492 2817 4309 3953 1.09 0.93 1985/86 3611 4298 7909 3887 2.03 1.73 1986/87 3455 4471 7926 4044 1.96 2.19 1987/88 3579 3489 7068 4014 1.76 1.9 1988/89 3780 2951 6731 3656 1.84 1.79 1989/90 6553 4999 11552 3778 3.06 3.02 1990/91 4365 3977 8342 3548 2.35 2.52 1991/92 3745 3954 7699 3025 2.55 2.54 Trend -2.60 -5.76 -4.30 -2.72 -1.59 4.66 Growth*

* Definedas the annual compoundgrowth rate between 1980-82and 1990-92

Source Food Studies Group, 1992

Annex2 Page 184 SouhAfra AgrcuIture:&Wrucur. Perfodman & Oion foAthe Futurn

Tablb2.10: Mai Produgtion.Consumpion. Imoros and ExPOrts

Season Produoion Consumpdion Impots Exports (a) (b) 197011 6133 5104 73 1302 1971/72 8600 4824 16 283S 1972/73 9483 S164 0 3917 1973/74 4160 S157 0 607 1974/7S 11037 5807 0 3698 197S/76 9098 S856 0 3769 1976/77 7472 S888 0 2087 1977/78 9714 5688 0 3170 1978J79 10056 5490 0 3843 1979/80 8332 S877 10 3141 1980/81 10762 5887 23 4256 1981/82 14656 63S4 62 5924 1982/83 8359 6960 89 5096 1983/84 4083 6806 2389 833 1984/8S 440S S727 2076 712 1985/86 7909 5479 224 1204 1986/87 7926 5206 34 3538 1987/88 7068 5573 31 2357 1988/89 673: 567S 57 1390 1989/90 11552 6362 3 4909 990/91 8342 6673 0 1963

Source: Maize Board (1991)

Notes: (a) Departmentof Agriulture Esfimates (b) Includingsales to BLNS, TBVC

Anner 2 Page 185 South Afikan Aorkau: Struture. Performan & Orins for the Future

Table 2.10: Maz Prduction. Consumdtion.Imofts and Exoorts

Season Productoa Consumption Imports Exports (a) (b) 1970171 6133 5104 73 1302 1971/72 8600 4824 16 2835 1972/73 9483 5164 0 3917 1973/74 4160 5157 0 607 1974175 11037 5807 0 3698 1975/76 9098 5856 0 3769 1976/77 7472 5888 0 2087 1977/78 9714 5688 0 3170 1978/79 10056 5490 0 3843 1979/80 8332 5877 10 3141 1980/81 10762 5887 23 4256 1981182 14656 6354 62 5924 1982183 8359 6960 89 5096 1983/84 4083 6806 2389 833 1984/8S 4405 5727 2076 712 1985/86 7909 5479 224 1204 1986/87 7926 5206 34 3538 1987188 7068 5573 31 2357 1988/89 6731 5675 57 1390 1989/90 11552 6362 3 4909 1990/91 8342 6673 0 1963

Sou Maizo Board (1991)

Notes: (a) Depament of Agricuure Esimates (b) Includingsales to BLNS,TBVC

Annex 2 Page 186 Soth Afian Asrlcukure:Snruture. Perfonnance & Otpis for the Future

Table 2.11: Consumptionof LivestockProducts. 1955/56to 1990/91

Population Milc Consumption Beef Consumption White Moat Red Meat Miflon Total Kg/Cap Total Kg/Cap Total Kg/Cap Total Kg/Cap 'Q00MT '000 MT '000 MT '000 MT 1956 15.3 976 64.0 429 28.1 29 1.9 612 40.1 1957 15.8 997 63.0 479 30.3 29 1.8 667 42.1 1958 16.2 1018 63.0 474 29.3 31 1.9 661 40.9 1959 16.4 1131 69.0 492 30.0 40 2.4 680 41.5 1960 17.0 1139 67.0 486 28.6 41 2.4 685 40.3 1961 17.6 1108 63.0 498 28.3 42 2.4 692 39.3 1962 18.0 1118 62.0 487 27.0 43 2.4 691 38.3 1963 18.6 1093 58.9 501 27.0 44 2.4 697 37.6 1964 19.1 1104 57.9 537 28.2 48 2.5 734 38.5 1965 19.6 1075 54.8 508 25.9 52 2.7 706 36.0 1966 20.2 1107 54.9 505 25.0 61 3.0 727 36.1 1967 20.7 1075 51.9 512 24.7 65 3.1 722 34.9 1968 21.3 1110 52.1 508 23.8 77 3.6 746 35.0 1969 21.9 1101 50.3 496 22.7 95 4.3 762 34.8 1970 22.0 960 43.6 525 23.8 111 5.0 809 36.7 1971 22.6 992 43.8 553 24.4 120 5.3 84u 37.4 1972 23.5 965 41.1 588 25.0 159 6.8 866 36.9 1973 23.9 1015 42.4 600 25.1 191 8.0 840 35.1 1974 24.6 1022 41.5 567 23.0 230 9.3 813 33.0 1975 25.3 1026 40.5 S15 20.3 266 10.5 763 30.1 1976 26.0 1096 42.1 548 21.1 290 11.1 797 30.6 1977 27.1 1096 40.5 613 22.7 290 10.7 859 31.7 1978 27.9 1096 39.3 624 22.4 299 10.7 881 31.6 1979 28.7 1135 39.6 661 23.1 291 10.2 921 32.1 1980 29.5 1112 37.7 742 25.2 312 10.6 1010 34.2 1981 30.3 1181 39.0 640 21.1 339 11.2 907 30.0 1982 31.0 1167 37.6 654 21.1 379 12.2 944 30.4 1983 32.0 1179 36.9 667 20.9 438 13.7 994 31.1 1984 32.8 1181 36.0 681 20.8 460 14.0 1009 30.8 1985 33.8 1223 36.2 674 19.9 485 14.4 1009 29.9 1986 34.5 1130 32.8 621 18.0 498 14.5 929 27.0 1987 35.3 1114 31.6 644 18.3 525 14.9 936 26.6 1988 36.0 1253 34.8 648 18.0 545 15.1 938 26.1 1989 36.7 1245 33.9 639 17.4 545 14.8 927 25.2 1990 37.6 1244 33.1 638 17.0 548 14.6 948 25.2 1991 38.4 1262 32.9 693 18.1 554 14.4 1028 26.8

Note: The populationseries used is the implicittotal from Table 71 of the above publication Source: Abstract of AgriculturalStatistcs

Annex 2 Page 187 Soguthhfi*n AgruIre: Structure.Perfonmattce & Opions foirthe Euture

ANNEX 3

1. Makn. Maize production fell by 9.6% per year during the 1970s, partly because of the change in statistical base resulting from the exclusion of the TBVC countries (beginning with Transkei in 1976f77, Bophuthatawana 1977n8, Venda 1979180, and Ciskei 1981/82). Production levels fluctuated from 200,000 tons in 1972/73 to 427,000 tons the following year. During the 1980s production levels recovered, growing at an annual rate of 7.2 percent, from 184,000 tons/yr in 1980/81 to 324,000 tons in 1989/90. Approximately 299,000 tons were produced in 1990/91.

2. D Beans. Production of dry beans fell from 23,000 tons in 1978/79 to 4000 tons in 1981/82, following the exclusion of the TBVC countries, but recovered to an average of 15,125 tons/yr over the remainder of the decade. Roughly 16,000 tons were produced in 1989/90 and 14,000 tons in 1990/91.

3. Sgrghum. Sorghum production fell from 24,000 tons in 1980/81 to 9000 tons the following year, peaked again at 26,000 tons in 1986/87, and fell to 12,000 tons in 1989/90 and 11,000 tons in 1990/91.

4. Whbea. Wheat production peaked at 19,000 tons in 1977/78, but fluctuated widely throughout the period. Crops of 14,000 tons and 11,000 tons were harvested in 1982/83 and 1987/88 respectively, but production fell below 5000 tons in each of the decade's remaining eight years. In four of those years production fell to 2000 tons or less. 3000 tons were produced in 1989/90 and again in 1991/92.

5. Groundnuts. Production of groundnuts fell from 9000 tons in 1978/79 to 1000 tons per year between 1981/82 and 1984/85. About 5000 tons were produced in 1987/88, 2000 tons in 1989/90, and 3000 tons in 1990/91.

6. Sunflowe Seed. Production of sunflower seed grew from less than 500 tons prior to 1985/86 to 2000 tons in 1989/90 and 1990/91.

7. Livestock. Livestock is an important type of investment for the black population. Although the homelands are heavily overstocked and average turnover is 2 percent, some districts in the homelands have cattle turnover rates exceeding 6% (Groenewald and du Toit, 1983). Numbers of cattle and pigs appear to be falling in the homelands, while numbers of sheep and goats are on the increase, reflecting perhaps more favorable mutton prices and/or the substitution of sheep and goats for cattle because of the shortage of land and the decline of quality pasture in homeland areas. Growth in poultry production has been mixed, increasing in Venda, Bophuthatswana, and Lebowa, but declining in the Transkei and the Ciskei (Table 3.1).

Annex3 Page 188 South Afircan Agricuture: Structure. Performance& ODrionsfor tIaeFuture

Table 3.1: Gowth Rates in Livestock Production. 1970-1989

Number Slaughtered Meat 1970-79 80-89 70-89 1970-79 80-89 70-89 1970-79 80-89 70-89 (% per year) Republo of South Africa:

Cattle 1.86 -0.19 -0.14 3.60 -1.74 0.81 4.34 -0.80 1.42 Pigs 0.32 2.29 1.52 0.68 3.94 1.77 0.86 2.68 1.97 Sheep & Goab 0.51 -1.01 -0.44 -1.33 -0.79 0.66 -1.01 -1.80 0.31 Poultry 11.48 4.38 7.20 Transkok

Cattle -0.88 Pigs -11.89 Sheep & Goats 1.84 Poultry -11.26 Venda:

cattle -2.89 Pigs -7.57 Sheep & Goats 6.50 Poultry 15.50 Bophutbhawana:

cattle -3.84 pigs -7.66 Sheep & Goats 4.34 Poultry 2.54 Ciskei

Cattle -6.52 pigs -3.05 Sheep & Goaft 1.53 Poultry -4.33 Lebowa

Cattl 2.76 pigs -4.86 Sheep & Goats -8.18 Poultry 33.05

Sourcc Appendix Tables B.9, B.12 - B.16.

Livestockin the TBVC Countries

8. Catl. Cattle numbersremained steady in Transkei over the 1980s, averaging 1.5 million animals. In Bophuthatswanacattle numbersdeclined at an averagerate of 3.8% per year, from 602,560 in 1981 to 467,354 in 1988. In Vendathe mnmberof cattlefell from 169,365 in 1980 to 83,485 in 1985, before recoveringto 136,339 by 1989. Cattle numbersin Ciskei have followed a similar pattern, from

Annex 3 Page 189 South African Africubure: Structure. Performance & Optionsfor the Future

184,677 in 1980 to 69,242 in 1985, and then back up to 124,329in 1989.

9. Pigs. The numberof pigs reported in Transkei fell at a rate averaging11.9% per year over the 1980s, from 424,744 in 1983 to 181,815 by 1988; no pigs were reported in 1989 and 1990. In Bophuthatswanapig numbers fell from an averageof 13,128animals for 1981-83to an average of 7953 animalsbetween 1984 and 1988. From a total of 5141 pigs in Venda in 1980 the number grew to 12,635 in 1985 before falling to 3512 in 1989. Pig numbersheld steady in Ciskei over the 1980s, at an average of 15,323.

10. Shee and goats. Transkei sheepnumbers grew from 2.1 million in 1980to 2.7 millionby 1990. Goats averaged 1.7 million over the decade, with 1.7 million animalsreported in 1990. The number of sheep in Bophuthatswanarose from 194,002 in 1981 to 263,328 in 1988, while goats increased from 425,489 in 1981 to 530,430 by 1988. Consideredtogether, sheep and goat numbers grew at an average annual rate of 4.3% during the period. Sheepnumbered an averageof only 2291 animalsin Venda, less than a tenth the goat total, which grew from 45,687 in 1980to 77,960 by 1989. Sheepand goats together increased at an annual rate of 6.5% in Venda. The number of sheep and goats held steady in C.skei, averagingabout 498,000 animals,equally divided.

11. Eq__. The number of fowl in Transkei has fallen from an averageof 1.5 million for 1980-83 to an averageof about 668,000for 1984-88,a declineaveraging 11.3% per year over the 1980sas a whole. Numbersin Bophuthatswanahave held steadyaround an averageof about201,000 per year. In Venda,fowl numbershave increasedat an average rate of 15.5% per year, from 90,044 in 1980 to 336,366 in 1989. Fowl numbers in Ciskei have fallen slightly, from an average of 112,000for 1980-82to an average of 86,105 for 1983-87.

AgriculturalProduction in Lebowa

12. Maize, sorghum, millet, and sisal together accountedfor 94% of the total area under principal field crops in Lebowain 1989, as indicatedin Table 3.2. Data on area, production,and yields of these and other major crops (Table 3.3), as well as livestock, are availablefor Lebowa between 1981 and 1989. These data are discussedby commoditybelow:

13. Maim. The area devotedto maizefell steadilyduring the 1980s. In 1981 and 1982more than 100,000hectares wereplanted to maize; but in 1983 this figure had fallento 67,014 hectares, and by 1989 only 54,623 hectares of maize were planted. Nevertheless,maize remains the single most importantcrop in Lebowain terms of area, accountingfor 49% of the land under principalfield crops in 1989. Production levels fell less abruptly than cultivatedarea, from 52,014 tons in 1981 to 36,429 tons in 1989. Yields improved modestly,from 482 kilograms/hectarein 1981to 667 kilograms/hectarein 1989.

14. Soiulm. The area devotedto the productionof sorghumdipped from 31,363hectares in 1981 to 10,216hectares in 1984 before climbingback to 30,708 hectaresin 1989, representing27% of fieldcrop area. Productiondropped an even more dramatically,from 13,472tons in 1981 to just 1142tons in 1984, before climbingto 16,909 tons in 1989. Yieldsaveraged 382 kilogramsper hectare across the decade, but fell to 112 kilograms/hectarein 1984. About551 kilogramslhectarewere produced in 1989.

Anne 3 Page 190 Sowh African Agricultre: Strruture. Perfonnance& Optionsfor the Future

15. Millet. Millet Is the third most importantcrop in Lebowain terms of area allocated,after maize and sorghum, representing 13% of field crop area. A total of 14,054 hectares were planted in 1989, up from a low of 7057 hectares in 1984, but down from 23,453 hectares in 1981. Productionlevels ranged from 1425tons in 1984,when yields wereonly 202 kilograms/hectare,to 18,152tons in 1989,when yields reached 1,292 kilograms/hectare. The increase in yields over the period representedan average annual growth rate of 14.7 ton/hectare

16. Sisa The area devotedto sisal grew slowlybut steadilyin Lebowaduring the 1980s,from 4858 hectaresin 1981to 5912 hectaresin 1989, representingan averageannual increaseof 2.5 ton/hectare Sisal accountedfor more than 5% of field crop area in Lebowa in 1989. Productionfell over the same period, from 1565tons in 1981 to 1170tons in 1989. Yields of sisal declined at an averagerate of 4.9% per year over the decade, from 322 kilogramslhectarein 1981to 198 kilograms/hectarein 1989.

17. Wheat, The area plantedto wheat in Lebowaaveraged 1557 hectares In the early 1980sand then fluctuated widely over the remainder of the decade, exceeding2200 hectares in four years, but fallingto 200 hectares or less in the other two. 2445 hectares were planted in 1989. Production fluctuated accordingly, totaling 8557 tons in 1989. Yields grew frorr. 917 kilograms/hectare in 1981 to 3.5 tons/hectare in 1989.

18. Cweas. The area devoted to cowpeashas fallen very sharply over the last decade. From 15,470 hectares in 1981, only 1809 hectares were planted in 1989, representing an average decline of 22.8% per year. Productionfell from 6339 tons in 1981to 780 tons in 1989, while yields fluctuatedaround an averageof 448 kilograms/hectare. Yields were 431 kilograms/hectarein 1989.

19. Cotton. Cotton has increasedrapidly in terms of area plantedover the past decade. A total of 1402 hectares were allocatedto cotton in 1989,up from 138hectares in 1981, representinga growth rate of 33.4% per year. Productiongrew at an even faster rate, averaging36.7% per year, from 312 tons in 1981to 3869 tons in 1989. Yields averaged 1.9 tons/hectare,but reached2.8 tons/hectare in 1989.

Annex3 Page 191 South African Aericulture:Structur,. Perfornance& Optionsfor the Future

Table 3.2: Percentaceof CultivatedArea by Crop. Lebowa. 1981-1989

1981 1982 1983 1984 1985 1986 1987 1988 1989 Maie 55.8 55.3 54.6 54.1 54.8 56.0 53.0 50.4 48.6 Sorghum 16.2 15.9 13.1 16.5 17.4 17.4 21.9 28.4 27.3 Millei 12.1 12.5 17.9 11.4 14.7 11.8 12.9 8.5 12.5 Sisal 2.5 2.7 4.3 8.2 5.4 5.2 5.5 4.4 5.3 Wheat 0.7 1.0 1.3 0.3 2.6 2.6 0.1 1.7 2.2 Cowpeas 8.0 7.6 4.2 5.4 2.4 2.0 2.1 2.2 1.6 Cotton 0.1 0.2 0.5 0.5 0.5 2.0 1.4 1.4 1.2 Dry Beans 2.5 2.6 2.4 0.9 1.2 1.8 2.2 2.0 0.9 Groundauts 1.9 1.8 1.3 1.9 0.6 0.5 0.3 0.9 0.2 Sugar-cauie 0.1 0.1 0.1 0.3 0.2 0.2 0.2 0.0 0.2 Sunflower 0.0 0.3 0.1 0.3 0.1 0.5 0.2 0.0 0.0 Tobacco 0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.0 Lucerne 0.0 0.1 0.1 0.1 0.0 0.0 0.1 0.0 0.0 Total ('000 ha) 193 183 123 62 98 107 107 129 112

ource: AppendixTable B.11.

20. Dry B t The area devoted to dry beans varied between 564 hectares in 1984 and 4795 hectares in 1981, with 971 acres planted in 1989. Productionlevels varied widely as well, from 58 tons in 1984 to 2274 tons in 1987. Some 660 tons were produced in 1989. Yields ranged from 103 kilograms/hectarein 1984 to 961 kilograms/hectarethree years later and averaged680 kilograms/hectare in 1989.

21. Groundnuts. Groundnutsdeclined sharply in importancein Lebowa over the 1980s, fallingby 24.8% per year in area and by 23.2% per year in production. While 1670 tons were produced on 3692 hectaresin 1981,only 125 tons were producedon less than 277 hectaresin 1989. Yieldsfluctuated around an averageof 545 kilogramns/hectare,and averaged451 kilograms/hectarein 1989.

22. Sug . The area devotedto sugarcanehas remainedrelatively stable over the courseof the last decade. Roughly 181 hectareshave been reported under sugarcaneeach year with the exceptionof 1988,when only 44 hectareswere reported. Totalproduction fluctuated between 277 tons in 1988 and 1983 tons in 1989, for an average of 720 t/yr. Yields increasedat an averageannual rate of 14 percent, from 3.1 tons/hectarein 1981 to 10.7 tons/hectarein 1989.

23. Sunflower. More than 530 hectareswere devotedto sunflowerin Lebowain 1982 and again in 1986; the total in the rest of the 1980swas less than half that level, and fell as low as 49 hectares in 1989. 1986 was an exceptionalyear in terms of both productionand yield, averaging2.7 tons/hectarefor a total production of 1418 tons. A total of 471 tons were produced in 1982, at an average yield of 885 kilograms/hectare;production in other years remainedbelow 100 tons in total, and yields in other years never exceeded400 kilograms/hectare.

24. Tobac. The area devotedto tobaccoproduction in Lebowafell from a peak of 49 hectares in

Annex 3 Page 192 SouthAfcan Atricukra: Siacture.Perfo_rance & Otlons for theFuture

1983 to just 9 hectares in 1989. The figures for total production in those years were 37 tons and 16 tons respectively. Yields varied between 405 kilograms/hectarein 1984 and 1778 kilograms/hectarein 1989, and averaged743 kllograms/hectareduring the period.

25. Lucerne. The area devoted to the production of lucerne has fallen by 22.9% per year since 1981, from 90 hectaresor more in the early 1980sto only 4 hectares in 1989. Productionfell from a peak of 2234 tons in 1983 to just 66 tons in 1989. Yieldsdipped to 0.8 tonslhectare in 1985, but exceeded7.0 tons/hectare in other years, and reached 16.5 tons/hectarein 1989.

Tablo3.3: Lebowa: GrowthRates in Asricultural CommodityAreas. Productionand Yields. 1970-1989

Area Production Yield 1981-89 1981-89 1981-89 Lebow: M(%per year) Maize -6.15 -0.64 5.87 Sorghum 3.31 6.42 3.01 Mlet -7.38 6.21 14.68 Sisal 2.51 -2.46 -4.85 Wheat 1.34 14.87 13.35 Cowpeae -22.83 -17.75 6.57 cotton 33.36 36.65 2.47 Dry Beans -11.68 0.65 13.96 Groundnuts -24.83 -23.23 2.14 Sugarcane -6.46 6.63 13.99 Sunflower -20.08 -28.74 -10.84 Tobacoo -12.40 -7.29 5.85 Lucne -22.87 -24.96 -2.71

Soure: Roth etA, 1992.

26. Cat&. The number of cattle slaughteredin Lebowaaveraged about 16,000 animalsduring the early 1980s and increasedto about 24,000 for 1986-88. A total of 14,039 cattle were slaughteredin 1989, the last year for which data are available.

27. M. The numberof pigs slaughteredvaried widelyduring the 1980s,from 978 in 1981to 8382 in 1987. Only 33 animalswere reported slaughteredin 1989.

28. Shee and Coats. An averageof 2476 sheepand goats wereslaughtered each year between 1981 and 1989. A total of 493 sheep and 204 goats were slaughteredin 1989.

29. Fowl. Approximately3.7 millionfowl were slaughteredin Lebowain 1989,up sevenfoldfrom 1981, when 532,112 fowls were slaughteredand nearly nine times the number (63,439) in 1985.

Amnex3 Page 193