This item has the following downloads:
1 THREE ESSAYS O N THE ECONOMICS OF THE SOUTH AFRICAN CITRUS INDUSTRY By JEAN PAUL BALDWIN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGRE E OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013
2 2013 Jean Paul Baldwin
3 To my family and loved ones
4 ACKNOWLEDGMENTS I would like to thank my advisor, Dr. Thomas H. Spreen, for h is mentoring and encouragement in accomplishing my research objectives. His wealth of knowledge in citrus and economics has instilled much confidence in me while throughout my doctorate degree. I am also grateful to other members of my supervisory commit tee: Dr. Zhifeng Gao, Dr. James Sterns, Dr. Richard Weldon and Prof. Linda Young. I would like to give special mention to Dr. James Sterns for his excellent mentoring in helping me understand institutional economics, as well as Dr. Zhifeng Gao who provide d a great deal of guidance in building import demand models. I would also like to thank Mr. Pet er Turner from Biogold for making my research in South Africa possible. His assistance in setting up interviews with producers, researcher and exporters within the South African citrus industry proved invaluable during this study Additionally, I would like to thank the Food and Resource Economics Department at the University of Florida for providing me with an excel lent educational experience. Any future succ ess as an econometrician is directly related to this learning environment. Lastly, I would like to thank my parents who have always encouraged me to follo w my dreams, and Barbara for the love and patience she provided during my educational tenure at the Un iversity of Florida.
5 TABLE OF CONTENTS page ACKNO WLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIA TIONS ................................ ................................ ........................... 11 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 The South African Citrus Industry ................................ ................................ ........... 15 Aim and Scope ................................ ................................ ................................ ....... 18 2 THE SOUTH AFRICAN CITRUS INDUSTRY PRE AND POSTDEREGULATION: A STRUCTURE, CONDUCT, AND PER FORMANCE ANALYSIS ................................ ................................ ................................ .............. 20 Introductory Statements ................................ ................................ .......................... 20 Structure ................................ ................................ ................................ ................. 20 Exp ort Channel and Organizations ................................ ................................ ... 20 Growers, Nurseries, Packers and Shippers ................................ ...................... 24 Hectares Production and Growing Regions ................................ ...................... 25 Packing Systems ................................ ................................ .............................. 26 Technology ................................ ................................ ................................ ....... 27 Funding ................................ ................................ ................................ ............ 28 Natural Endowments ................................ ................................ ........................ 29 Business Climate ................................ ................................ .............................. 30 Barriers to Trade ................................ ................................ .............................. 30 Conduct ................................ ................................ ................................ .................. 33 Domestic Consumption ................................ ................................ .................... 33 Pricing and Marketing ................................ ................................ ....................... 33 Labor and Land Issues ................................ ................................ ..................... 35 Shipping and Logistics ................................ ................................ ..................... 37 New Plantings ................................ ................................ ................................ .. 39 Performance ................................ ................................ ................................ ........... 40 Market Share ................................ ................................ ................................ .... 40 Major Export Markets by Volume ................................ ................................ ..... 41 Production ................................ ................................ ................................ ........ 42 Gross Value ................................ ................................ ................................ ...... 43 Export Percentage of Total Production ................................ ............................. 43 Concluding Statements ................................ ................................ .................... 44
6 3 THE EXPORT MARKET FOR FRESH SOUTH AFRICAN CITRUS ....................... 63 Introdu ctory Statements ................................ ................................ .......................... 63 Literature Review ................................ ................................ ................................ .... 64 Model Specification ................................ ................................ ................................ 66 Results ................................ ................................ ................................ .................... 71 U.S Dollar ................................ ................................ ................................ ......... 71 Oranges ................................ ................................ ................................ ..... 72 Mandarins ................................ ................................ ................................ .. 72 Grapefruit ................................ ................................ ................................ ... 73 Lemons ................................ ................................ ................................ ...... 74 European Euro ................................ ................................ ................................ 74 Europe ................................ ................................ ................................ ....... 75 Oranges ................................ ................................ ................................ ..... 75 Mandarins ................................ ................................ ................................ .. 76 Grap efruit ................................ ................................ ................................ ... 76 Lemons ................................ ................................ ................................ ...... 77 Optimal Allocation across Export Markets ................................ .............................. 77 Oranges ................................ ................................ ................................ ............ 80 Mandarins ................................ ................................ ................................ ......... 81 Grapefruit ................................ ................................ ................................ ......... 81 Lemons and Limes ................................ ................................ ........................... 81 Concluding Statements ................................ ................................ ........................... 82 4 VERTICAL INTERGRATION OF FOREIGN RETAILERS WITHIN THE SOUTH AFRICAN CITRUS INDUSTRY ................................ ................................ .............. 90 Introductory Statements ................................ ................................ .......................... 90 Problem Statement ................................ ................................ ................................ 90 Objectives ................................ ................................ ................................ ............... 91 New Institutional Economics ................................ ................................ ................... 92 UK Retailers ................................ ................................ ................................ ............ 96 Major Supply Channels ................................ ................................ ........................... 98 The Industry Responds ................................ ................................ ......................... 102 Concluding Statements ................................ ................................ ......................... 108 5 CONCLUSIONS ................................ ................................ ................................ ... 115 Limitations of the Research ................................ ................................ .................. 118 Future Area of Research ................................ ................................ ....................... 119 APPENDIX A R EGRESSION R CODE ................................ ................................ ....................... 121 B OPTIMAL ALLOCATION GAMS CODE ................................ ................................ 135 C QUESTIONNAIRES ................................ ................................ .............................. 149
7 LIST OF REFERENCES ................................ ................................ ............................. 160 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 167
8 LIST OF TABLES Table P age 2 1 Total hectares of production per variety ................................ ............................. 61 2 2 Comparison of pre and post deregulated citrus producing land areas ................ 61 2 3 Per capita consumption of citrus in South Africa, Kg per year ............................ 61 2 4 Minimum wages for farm workers in South Africa 2009 2013 .......................... 62 2 5 New tree plantings per variety in South Africa ................................ .................... 62 4 1 Summary of producer drivers ................................ ................................ ........... 114 4 2 Summary of retailer drivers ................................ ................................ ............... 114
9 LIST OF FIGURES Figure page 2 1 Citrus producing regions within South Africa ................................ ...................... 46 2 2 Orange exports comparison, South Africa vs. northern hemisphere competitors. ................................ ................................ ................................ ........ 47 2 3 Mandarin exports comparison, South Africa vs. northern hemisphere competitors. ................................ ................................ ................................ ........ 48 2 4 Grapefruit exports comparison, South Africa vs. northern hemisphere Competitors. ................................ ................................ ................................ ....... 49 2 5 Lemon exports comparison, South Africa vs. northern hem isphere competitors. ................................ ................................ ................................ ........ 50 2 6 Comparison of shipping techniques in the South African citrus industry, millions of cartons ................................ ................................ ............................... 51 2 7 New citrus tree plantings in South Africa ................................ ............................ 51 2 8 Market share of orange exports amongst five largest exporters. ........................ 52 2 9 Market share of mandarin exports amongst five largest exporters. .................... 53 2 10 Market share of grapefruit exports amongst five largest exporters. .................... 54 2 11 Market share of lemon exports amongst five largest exporters. ......................... 55 2 12 Major ma rket orange export destinations ................................ ........................... 56 2 13 Major market mandarin export destinations ................................ ........................ 56 2 14 Major market grapef ruit export destinations ................................ ....................... 57 2 15 Major m arket lem on export destinations ................................ ............................. 57 2 16 South African production quantities of citrus varieties ................................ ........ 58 2 17 South African gross value (R100 0) of citrus varieties ................................ ......... 59 2 18 South African export percentage of citrus varieties ................................ ............ 60 3 1 World fresh citrus exports thousan d tons ................................ ........................ 84 3 2 Price, income, GDP, deflator elasticities estimators (For countries operating under U.S dollar during trade) ................................ ................................ ............ 85
10 3 3 Price, income, GDP and deflator elasticities estimators (For the United Kingdom and countries currently operating under euro (domestic currency pre euro) ................................ ................................ ................................ ............. 86 3 4 Actual vs. estimated optim al orange quantities ................................ ................... 87 3 5 Actual vs. es timated optimal oranges prices ................................ ...................... 88 3 6 Comparison between actual and estimated optima l revenue ............................. 89 4 1 Summary of interviews ................................ ................................ ..................... 112 4 2 Five major supply channels ................................ ................................ .............. 113
11 LIST OF ABBREVI ATIONS ANC African National Congress C O Cochrane Orcutt CBS Citrus Black Spot CGA Citrus Growers Association CLAM CRI Citrus Research Institute D W Durbin Watson EU European Union GDP Gross Domestic Prod uct GFS Group Food Sourcing IMF International Monetary Fund IPL International Produce Limited OCC Outspan Citrus Center OI Outspan International OLS Ordinary Least Squares ML AR Market Assisted Agrarian R eform NIE New Institutional Economics PPECB Perishabl e Product Export Control Board SACCE South African Cooperative Citrus Exchange SAPA South African Press Association SHAFFE Southern Hemisphere Association of Fresh Fruit Exporters SME Small Medium Enterprise SPS Sanitary and P hytosanitary standards TCE T ransaction Cost Economics
12 UK United Kingdom USDA United States Department of Agriculture
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Do ctor of Philosophy THREE ESSAYS ON THE ECONOMICS OF THE SOUTH AFRICAN CITRUS INDUSTRY By Jean Paul Baldwin August 2013 Chair: Thomas H. Spreen Major: Food and Resource Economics The citrus industry has continued to establish itself as one of South Afri primary agricultural sectors With a national unemployment rate of over 25% the South African citrus industry has played a significant role in providing a source of economic activity for large ly rural population. Additionally, the agr icul tural sector continues attract ing immigrants from neighboring countries seek ing additional employment opportunities Primarily focusing on its transition from a regulated to deregulated industry ( brought upon by the removal of apartheid and, in turn, t political party ) the citrus industry has encountered numerous transformative periods. Previou s res earch focuses predominantly on the initial implications caused by the deregulation process. Comparatively, little research ha s analyzed the efficiency or more recent economic performance of the citrus industry, nor the relations between instit utional arrangements from a pre and post deregulated period This research aims to contribute to the limited existing literature by qualit a tively and quantitatively analyzing these understudied components. W e present a n industrial
14 demand model, and a case study based upon interviews and responses from Sou th African farmers and exporters. Attempting to provide insight into the current institutional arrangement of direct trading adopted by mass retailers, f armers and exporters discuss its viability Additionally, we examine potential short and long term consequences brought upon by this new channel.
15 CHAPTER 1 INTRODUCTION The South African Citrus Industry As trade route to and from the East was quickly recognized by the Dutch East India Company. European sailors found the approximate year long sea journey, as well as the inaccessibility to proper nutrients, highly treacherous. The result was the establishment of a permanent supply base at the Cape of Good Hope which overtime created a market for agricultural and food products grown in southern Africa. During the 1800s, the fresh fruit industry found itself unable to function purely on domestic demand. New routes were sought in order to maintain the very survival of t he industry. to the United Kingdom in 1892 (Fresh Produce Export Forum, 2010 ). The evolution of the South African citrus indust ry has certainly not been predictable over the past one hundred years; however, its recorded success cannot go unnoticed. Although many export industries can attest to the difficulties associated with limited technology or resources during the 18 th and 19 th century, little can be compared to adversities the South African industry inherited from its domestic struggles and visitations during the late 20 th and 21 st century. These challenges continue to transform the industry, most notability within the land a nd labor sectors. The 1940s proved to be a compelling period in defining not only the structure of the citrus industry, but also the country itself. During this period the government adopted the South African Citrus Cooperative Exchange (SAC C E) as the sing le citrus exporting entity within the industry
16 -while also pursuing complete, country wide, racial segregation. Known today as apartheid, the corresponding policies include the Population Registration Act, Group Areas Act, Prohibition of Mixed Marriages A ct, and the Immorality Act, all of which account for some illegal or criminal offense pertaining to your heritage, particularly indigenous black citizens. Perhaps most troubling, apartheid would be enacted by a government comprised of whites -a race that was, by far, a minority of the population. The industry flourished during its earlier years under the regime of the Citrus Exchange. South Africa had soon become a respectable and established industry for fresh citrus which boasted many international cli ents. Yet their business model of supplying large quantities of fruit on spot markets was soon to be outdated. Internationally, the industry was facing difficulties in various forms throughout the 1980s. Retailers were expecting higher quality, niche prod ucts, which the current infrastructure of Sout h Africa was unable to fulfill (Mather and Greenberg, 2003). Furthermore, the industry faced additional pressure from newly emerging southern hemisphere citrus growing competitors. Domestically, Sout h Africa be ca me increasingly unstable through the petitioning and rioting of its citizens for the removal of the government and its apartheid regime. Both the citrus industry and country needed desperately to meet international demands in terms of its fruit and gover nance manner. From the inevitable unsustainable future and international pressure, apartheid was abolished in a series of negotiations from 1990 to 1993. This event laid the foundation for South Africa's first democratic election in 1994 in which the black governed party, ANC led by Nelson Mandela, won. The new government brought about many changes, one of which was the removal of a regulated single operated
17 export industry body of all fruits. This decision was the catalyst for the registration of over 20 0 new citrus e xporting agents in the country (de Beer, Paterson and Olivier, 2003). The development of a deregulated industry, under a newly democratic government, painted a very bleak and uncertain future for many growers. Additionally, in owned farm land to either up coming black farmers or black families removed from their home under the Group Areas Act. The transition from one to over 200 citrus exporters catapulted the industry into turmoil. Institutions once required to oversee the management and quality control of one export body now oversaw hundreds. Newly registered exporters with no or limited export knowledge were unable to adjust to deviations in prior contrac tual agreements. This put many farmers at risk due to their product being disposed, or sold severely under market value due to the limited alternatives of the inexperienced exporter. Today, the ANC remains South Africa's elected political party. Government intervention has since been limited to quality assurance for exports and market reform initiative, aimed at transferring land from white controlled farmers to black owners This has proven to be a failed experiment, and has forced the government to look towards a financial co mpensation for the white farmer (SAPA, 2013). has also caused some distress for farmers who predict similar land evasion practices as witnessed in Zimbabwe.
18 Through this unrest, South Africa has transformed itself into the second largest exporter of fresh citrus in the world. The industry co ntinues to be tested from various avenues; nevertheless, it proceeds to adapt to externalities not commonly found in other countries. More recently, the industry faces a new battle from international retailers. The practice of ethical trade, one that shoul d benefit all transacting parties, is being challenged by the citrus industry. Retailers continue pushing for greater quality and specifying business practices for growers; however, they have themselves continued acquiring rebates from shipments, debrandin g (retailers removing producer labels and branding fruit under own trade mark) and rejecting (satisfactory) fruit when fruit has struggled to sell. Through these practices, as well as the purchasing and ownership of their own farms, retailers have acquire d greater returns which leave many citrus export players in uncertain territory. The sustainability of the traditional exporter is proposed to be most at risk, as these tasks and duties are slowly being acquired by the retailer -either contracted out or attained through a division of their own corporation. In terms of development, trade and its institutions, the evolution of South Africa and its citrus industry has emerged as an exciting research agenda for agricultural economists. The three essays prese nted herein aim at shedding some insight on the efficiency of and future outlook for the industry Aim and Scope The goal of this study is to analyze quantitatively and qualitatively various dimensions within the South African citrus industry. Although pre vious studies have examined the impacts of market liberalization within the industry, most notably by Mather et al., no research has analyzed the efficiency of South African citrus exports,
19 nor potential consequences to the structure of the industry throug h vertical integration by foreign retailers. In accomplishing these topics, the study is organized as follows: Chapter 2 reviews the industry through a structure conduct performance outlook, comparing these dimensions during a post and per regulated period Chapter 3 empirically analyzes and compares current South African export quantities and revenue to a proposed optimal maximizing profit allocation This is achieved through the estimation of elasticities obtained from import demand models and the constru ction of a quantity constrained programming model. Chapter 4 investigates institutional changes and how South African producers and exporters have adapted to this external movement. Through literature proposed by New Institutional Economics, export players provide insight for this institutional environment change, as well as short and long term consequences of it. Finally, Chapter 5 concludes with key findings as well as discussing potential future work. South Africa was identified as an exciting sector to conduct our research primary through its rich history, limited research, as well as being a global leader in citrus exports. Each chapter provides some awareness as to the direction and operation of the industry. Spec ifically, Chapter 4 introduces vertical i ntegration -a recent innovation adopted by retailers that has yet to fully materialize. While its consequences have not been empirically estimated, vertical integration could radically change the structure and performance of the South African citrus ind ustry. Given this uncertain future, this
20 CHAPTER 2 THE SOUTH AFRICAN CITRUS INDUSTRY PRE AND POSTDEREGULATION: A STRUCTURE, CONDUCT, AND PERFORMANCE ANALYSIS Introductory Statements The Sout h African citrus industry has a history spanning over 300 years, which includes the first orange and lemon trees brought in by the Dutch East India Company to the Cape (Western Cape). This history includes many periods of transition encompassing numerou s governance transitions, most notably through the two World Wars, apartheid, and its operation under regulated and, more recently, deregulated export channels. Through its transformation, the industry now ranks as the 2nd largest exporter of fresh citrus in the world and celebrated over 100 years of fresh citrus exports in 2007 ( Citrus Grow ers Association of South Africa, 2007 ). This chapter looks at the industry from a two fold Structure Conduct Performance (SCP) perspective Market structure during the regulated (pre 1997) export channel and the current, deregulated export channel (post 1997) are c ompared. Consequently, how these structures defined the conduct and performance of the market during these periods are examined. Structure Export Channel an d Organizations The South African Cooperative Citrus Exchange ( SACCE ) and the Perishable Prod ucts Export Control Board (PPECB ) were two integral bodies within the citrus industry during the regulated/ single export channel period. The SACCE operated as So PPECB oversaw all fruit assessment at cooperative and private pack houses. The PPECB also
21 later expanded their responsibilities to include the assessment of the fruit within cold stor e fac ilities at South African ports ( Mather ) The SACCE was established initially in the mid 1920s to assist in improving the quality of exported fruit, as well as to attain additional overseas markets. A marketing solution was sought following difficultie s in fruit exportation during the Second World War to Britain. In response, through the Marketing Act of 1937, a Citrus Board was created which held responsibility in setting domestic prices and possessed authority in appointing a solitary exporting body Although separate bodies, the SACCE and the PPECB had overlapping directors serving in both organizations. The SACCE was consequently appointed by the PPECB (Mather and Greenberg 2003 ). The SACCE gained a r eputation as a supplie r of high quality citrus to various markets during the 1960s; however, they faced increasing pressure from southern hemisphere competitors during the 1980s. Their ineffective pooling system of fruit, as domestic citrus growers. The SACCE responded by creating Outspan International (OI), a new operational and marketing organization Domestically, South Africa was soon to appoint its first dem ocratically elected president. In a series of negotiations, the previous apartheid regime was in the process of being dism antled. This regime specifically disallowed the privileges of black South Africa ns -privileges protected for their fellow white citi zens The removal of apartheid effectively reinstated racial equality and rights among all South Africans, including voting rights. It was to no surprise that the traditional ly black empowered political party:
22 the ANC (African National Congress), who wer e also at the forefront in the removal of apartheid, was elected as the controlling political party of the country during its first democratic election As part of their goals t he ANC looked to promote a more transparent economy, e.g., one that allowed a ny entrepreneur to enter. The removal of all regulated markets within South Africa was closely aligned to this vision. The citrus industry of South Africa was effectively facing an indefinite and uncertain future following the removal of Outspan as the s ole exporter of South African citrus. to a single channel marketing system. Almost 90% of growers found the single channel advantageous, while 8% (controlling 5.3% of the fruit v olume) were against the controlled ( Mather). Outspan and Unifruco, the single export agent of deciduous fruit in South Africa, (OI later merged with Unifruco to f orm Capespan) Under deregulation, the PPECB faced a number of administration issues with the entrance of over two hundred new clients. Many bankruptcies resulted due to price competition that producers created in establishing new export markets. The PPE CB responded by visiting European importers to ensure they were maintaining their quality controls and standards. This was obtained through investing in various cool ing and transportation methods (de Beer, Paterson and Olivier, 2003). Deregulation was in my view the single most important event in recent years to create a shift in responding to market demands, as we know it today. Never will the exports of fresh produce again be controlled from the production side market conditions and consumer preference s have now become the dominant driving forces behind fresh produce trade...
23 ... I believe we have the capacity, the will and the drive to become part of a new solution to international trade in food supply, and that the PPECB wil l p role in the years to come Mr Neels Hubinger, C hief Executive, PPECB, 1997 (de Beer, Paterson and Olivier, 2003). A number of other new citrus organizations emerged after deregulation. The FPEF registered in 1998 as a non pro fit organization. Their aim was to create a disciplined fruit exporter sector while providing leadership and service to its foreign buying communit y and its own members Deregulation of the citrus industry also s functions, most particularly, that of funding the Citrus Research Institute. After a difficult season in the new millennium with declining profits, the Citrus Growers Association (CGA) was created in an effort to tions. The CGA provided a forum where government, exporters, research institutions, and supplier interests are represented in the industry (Philip, 2006). In its evolving and emerging presence within the market, the CGA acquired, and enhanced, the functio ns of the outdated Outspan Citrus Center (OCC). In its previous form, the OCC attempted to address issues of marketing for a limited number of cultivators. With its new functions, the CGA consequently renamed the OCC as the Citrus Research Institute (CR I). The institute continues to receive funding through levies acquired by the CGA on all citrus exports (de Beer Paterson and Olivier, 2003). More specifically, they define their mission statement: To maximize the long term global competitiveness of the southern African citrus growers through the development, support, co ordination and provision of Research and Technical services by combining strengths of all CRI Group partners (Citrus Research International 2012 ) The CRI in South Africa is now separa ted into three portfolios, that being horticulture, disease management, and integrated pest management. Their horticulture
24 research is primarily focused on the pre and post harvest management of fruit condition and quality. The research conducted under di sease management focuses on the pre and post harvest diseases of citrus (Citrus Research International, 2012) More recently, the CRI has played an influential role in the regulations pertaining to citrus black spot and the movement of citrus. Through ex tensive surveys on the greening disease and citrus black spot, the CRI is able to provide insight into the distribution of these diseases. Such research continues to protect unaffected areas and maintain trade agreements from t hese regions to foreign mark ets South Africa, 2012). Growers, Nurseries, Packers and Shippers South Africa had approximately 3,500 commercial citrus farmers operating during the latter years of the regulated export period. Of these, 1,200 were involve d in exporting to approximately 40 countries. The Western and Eastern Cape accounted for over 35% of exports while the Northern Transvaal (now Limpopo region) exported 20% of South African citrus. Approximately 200 export pack houses were established acr oss the country Mather estimated this number to be 174 cooperative and estate pack houses throughout the country. Of these, 52% of packed fruit was exported through the port of Durban, 21% from Cape Town, 18% Port Elizabeth, and 9% Maputo. All exports w ere conducted through OI, and later Capespan, during the single operated export channel (USDA, 1995 ). During deregulation, in particular the late 2000s, it was estimated there were approximately 1,400 export citrus growers. Some of these farmers struggle d to adapt to the new protocols implemented during the change, while others attempted to establish their own marketing companies in an effort to cut additional costs (Mather, 2008 ). The
25 number of local small farmers (100 trees or less) remained similar to the regulated period of approximately 2,200. The number of p ack houses dropped significantly during deregulation to around 75, the smallest of which was capable of handling about 400,000 cartons (U.S International Trade Commission, 2006). Deregulation br ought approximately 180 new export agents. Of that figure, 20 exporters controlled 85% of total volume (Philip, 2006). Mather and Greenberg (2003) estimated that approximately 240 export agents were registered during the governance change to a deregulated export channel. Capespan lost approximately 20% of its share in exports due to increased competitors in the market; however, they were still able to secure almost all of the fruit within the Western and Eastern Cape due to their strong historical ties t o the cooperatives in the region (Mather). The deregulation of the industry saw a drastic decline in nurseries. Seventeen registered nurseries, down from the initial 160 pre deregulation, were operating in the early 2000s. In 2011, most South African exp orts are still shipped through Durban: 51% (675,981 pallets), 18% (240,477 pallets) through Port Elizabeth, and 18% (235,615) from Cape Town. Exports through Maputo have decreased to 4% (issues relating to the Maputo port are discussed elsewhere ), while S approximately 9% of all citrus (CGA, 2012). Hectares Production and Growing Regions Approximately one third of all citrus commercial farms are export oriented. The Western Cape orchards average in s izes between 30 40 ha with one of the largest producing citrus crops on approximately 200 ha. The Western Cape is also most populated in terms of orchards (687 operations), approximately 47 percent more than
26 the Eastern Cape, and 128 percent more than Lim popo. The Eastern Cape traditionally has larger orchards, the largest being approximately 4,500 ha in size (U.S International Trade Commission, 2006). area has occurred in Limpopo and the Easter n Cape. These two provinces have now established themselves as the two largest growers of citrus in the country. The province of Mpumalanga has had the largest decrease of citrus producing area sinc e deregulation, while Swaziland ( landlocked country with in South Africa) has also experienced a decrease. Current production per land area and a comparison of pre and post deregulation production areas is illustrated in Table 2 1 and Table 2 2 respectively. Packing Systems The payment scheme for farmers in the late 1980s encouraged farmers to produce large volumes of fruit. The single channel market was exposed in terms of adjusting to a transforming global market, including oversupply and demand for higher quality and new varieties. As a consequence, the p acking system was soon tested on its ability to meet new global demands. The Western and Eastern Cape contained some of the biggest pack houses in the wo rld, handling millions of 15 kg 1 boxes between six and seven months of the year. Boxes were then tran sported onto pallets that stored up to 70 boxes individually. This infrastructure struggled to adjust to world markets demand -1 Approximately 33 pounds
27 more complex as they were more susceptibl e to damage, while the standard 15 kg boxes were also insufficient in holding them (Mather and Greenberg 2003 ). Many of the pack houses, after the removal of a single operated export channel, were sold or abandoned due to the dissolution of its correspon ding cooperative facility. The pack houses in the Western and Eastern Cape were significantly larger due to the high membership of growers in cooperatives. After liberalization, large numbers of established high quality citrus growers distanced themselve s from the cooperative system and established private pac king facilities (Mather, 2008) This was initiated through an amendment to the Cooperative Act that allowed for privatization of new pack houses after the deregulation of the industry. These pack h ouses varied in size, ranging from small, 20 to 30 workers, to larger sizes that acquired up to 400 workers during peak season (Mather and Greenberg 2003 ). Today, the actual packaging of the fruit requires products present their design to the PPECB befor e any carton orders can be made. Some of the information required includes class I or II fruit, fruit type, country of origin, name of carton, and food safety accreditation number s (Agricultural, Forestry and Fisheries Department of South Africa, 2011 ). T echnology During the 1970s, integral containers for fruit transport were produced that, at the time, were very costly to manufacture. Some traditional methods such as refrigerated ships and porthole containers were proving to be inefficient for some of th e more delicate varieties Through the research and assistance of the PPECB, these integral containers became more financially feasible and had an influential role in increasing fruit quality during the 1990s. Through them, fruit was now able to be pack
28 specification. More specifically, fruit was now able to be transported to markets without any handling that was previously required (de Beer, Paterson and Olivier, 2003). Today, modern technology continues to transform the export produc tion chain into an even more efficient unit. Some of the technology advances used within the industry today includes computerized fertigation and digital access to production records on a per field basis to remotely apply fertigation. Computerized sensors are also utilized within pack houses that sort f ruit by certain characteristics (U.S International Trade Commission, 2006). for greater phytosanitary conditions have also assisted in advancing technology to preserve and maintain fr uit quality. Wireless technology and bar codes are used to monitor locations and temperatures of export pallets. This becomes essential for fruit that is pre cooled or packed directly into cold storage, as some allow little to no margin of error of temp erature for optimal fruit quality (U.S International Trade Commission, 2006). Funding The SACCE implemented a statutory levy for each box packed and sent for export. The funds obtained from the levies were instrumental in creating infrastructure developmen t, research, and transportation within the industry. The Exchange also vertically integrated and acquired a nursery system for producing disease free tre e s. Furthermore, they assisted growers by establishing extension offices and laboratories in high citr us production areas. These offices were primarily constructed to assist with fruit quality and diseases in the area. Lastly, the levy also aided the industry in upgrading loading and cooling facilities within ports, along with purchasing dedicated vessel s for citrus exports (Mather and Greenberg 2003 ).
29 After market liberalization, the CGA was entrusted to obtain funds from a similar mandatory statutory levy mandated on export cartons. This statutory levy, collected each year on all citrus exports, is us ed for research, market access, consumer insurance, logistics, market development, and administration. Currently, the statutory level is 41cents 2 per 15kg carton. An additional levy is also imposed on grapefruit exports to fund market development and con sumer education i n the United Kingdom and Japan ( Citrus Growers' Association of South Africa, 2012). Moreover this money also fund s the CRI, whose staff include professors from both the Universities of Stellenbosch and Pretoria. As these staff members a re government salaried, this is one method in which the government indirectly supports the industry (U.S International Trade Commission, 2006). Natural Endowments (Figure 2 1) Through regulated and deregulated channels, the industry maintains a vigorous counter seasonal advantage over their competition in the northern hemisphere. Navel oranges and lemons are produced predominantly in the Western and Eastern Cape due to the coo l moist win t ers associated with Mediterranean climate ( Mather). Essentially a frost free area, the Western Cape also has an advantage in the production of deciduous fruit -thus utilizing their packing pl ants for most of the year. As g rapefruit and Valen cias are more inclined to grow in warmer, subtropical climates, they are mostly found in Mpumalanga, Limpopo, and KwaZul u Natal. 2 Approximately U.S $0.052 (2012)
30 Business Climate Mr. J. Chadwick from the CGA prepared a study examining currency fluctuations He demonstrated South Africa wa s the only country, from its southern hemisphere competitors, whose domestic currency depreciated against the US Dollar between 2006 and June 2008. Indeed, South African currency is traditionally one of the more volatile against the U.S. dollar. Currency e xchange has a variety of vital roles, including the cost of supplies, interest rates, and exports within the industry. The South African Rand consistently fluctuated within 3 4 Rand to 1 U.S Dollar from the early to mid nineties ( OANDA) However, much of the volatility occurred around the turn of the new millennium, with the Rand reaching over R12 per $1 in 2002. Three year s later, this exchange decreased to approximately R6 per $1 (U.S International Trade Commission, 2006). Currently the South African r and trades between R8 9 per U.S dollar. Barriers to Trade Tariff and non tariff barriers are two mechanisms restricting market access for South African exporters. Many northern hemisphere markets operate under an entry price system which adopts multiple t ariff rates during different seasons. These tariff rates are typically higher when domestic producers are attempting to sell their product, and lower during their off season. This tariff rate can also change under the current market price. If the market is deemed to be too competitive (low season price), the tariff rate is consequentl y increased ( Agricultural, Forestry and Fisheries Department of South Africa, 2011). In many cases, competing citrus exporting countries have negotiated Free Trade Agreemen ts to make their product more desirable to foreign importers (Citrus Growers' Association of South Africa, 2012).
31 It is mainly non tariff barriers that are seen as the most restrictive measure for citrus exporters. These barriers include: sanitary and phy tosanitary standards (SPS), food health issues, food safety issues, labeling, packaging, and product certification procedures ( Agricultural, Forestry and Fisheries Department of South Africa, 2010). Domestic regulations within the United States pertaining to the movement of f ruit from citrus black s pot (CBS) infested areas in the U.S. are also under review, and is an excellent example of a non tariff barrier to South African trade ( Citrus Growers' Association of South Africa, 2012 ). South Africa gained acc ess to the US market post deregulation; however, all production areas outside of the Western Cape were excluded as the US Department of Agriculture (USDA) found th e Western Cape free from CBS disease Exports to the US have risen drastically since libera lization of the market -increasing to half a million cartons in 2002 from its initial thirty thousand in 1997, when it was first allowed access (Mather, 2008 ). The South African industry has been frustrated at their reluctance to accept technically justi fied treatments previously agreed upon (Citrus Growers' Association of South Africa, 2012) In example, the cold steri protocol was amended in 2010 with USDA support. The protocol allowed reduction of the cold steri period from 24 days to 22 days. The implementation of such a change was not agreed upon by South African and United States authorities, and so the protocol remains at 24 days for the control of false coddling moth. Industry officials have reported losses of between 6% and 15% percent that ar e attributable to the implementation of 24 days. This figure is reduced to 3 % when operating under the protocol of 22 days ( USDA, 2011). The U.S. has only recently declared additional
32 districts free of citrus black spot and; therefore, open to trade. Thes e regions (Figure 2 1) are located within the Western Cape, Free State, North West, and Northern Cape Provinces ( USDA, 2010). Additional barriers to trade includ e restrictions imposed by many eastern c ountries. Japan, South Korea, and Taiwan were identi the regulated period due to strict protocols of trade, which were considered far stricter than conditions enforced by other markets. Some conditions included specialized registration and additional quality inspection, whe re, if failed, would result as an expense to the exporter (Mather 2008 ). Thailand and South Korea have only recently provided strong indications to South African exporters regarding future market access. After a decade of attempting to gain access to th e Thai market, export protocols between South Africa and Thailand have now been finalized. This has come after two previous draft protocols failed to reach fruition as a result of regime changes. South Korea has also accepted protocols, and is allowing im ports of lemons and grapefruit from South Africa ( Freshfruitportal, 2010) As the southern hemisphere export leader, overlapping seasons with northern hemisphere competitors is a constant struggle for the industry. During their increased production of ora nges, mandarins, grapefruit, and lemons over the past decade, more common occurrences of overlapping resulted. While not all countries cater to the same market, the increasing occurrence of seaso n overlaps with other leading northern hemisphere importers, specifically Spain, Turkey, USA, and China -as demonstrated in Figure 2 2 to Figure 2 5.
33 Conduct Domestic Consumption Domestically, the per capita consumption of citrus fruit has varied over the past 30 years. With an average consumption of 12.19 kilogr ams 3 per year (Table 2 3) part of the new millennium ( Agricultural, Forestry and Fisheries Department of South Africa, 2011). Fresh oranges are ranked as one of the top consumed fruits in South Africa, selling for a market price around R61 4 per 15 kilograms in 2012. Oranges are seen as having a more desirable taste than grapefruit while also being comparatively cheaper; additionally, oranges purchased through the local m arket are consumed as fresh or processed for juice. Soft citrus is a popular alternative in the local market due to its ease at peeling and seedless characteristics. Aggregate domestic consumption of lemons was estimated to be 11,000 MT in 2012 and is use d for numerous aspects -including food, cleaning, and pharmaceutical industries (USDA, 2011). Pricing and Marketing Due to an oversupplied overseas market in the late 1970s and early 1980s, the South African citrus industry faced a difficult period. The issue of overlapping seasons was quickly becoming a topic of discussion. South African producers were continuing to produce larger quantities and, sometimes, questionable quality, due to the incentive payment structure which awarded for higher volumes. T he SACCE were, until then, highly successful in developing a first 3 26.87 pounds 4 Approximately U.S $7.625 (2012)
34 efforts were outdated and out of touch with the realities of the international trading 2008 ). In an attempt to adapt to new and growing demands for greater citrus variety, the SACCE responded in two ways. First, the pooling system was restructured to encourage higher quality fruits from the growers. This would result in niche fruits now specially branded to distinguish re putable growers from mass merchandisers such as Tesco, Marks & Spencer, and Sainsbury. Second, the SACCE attempted to establish itself as an international fruit trader. Through its marketing scheme, the SACCE was able to meet market demands; nevertheless they received some criticism from growers and pack houses. Certain grow ers were now forced to remodel their form operations, while larger pack houses struggled to meet the new demands for d ifferent varieties and quality (Mather 2008 ). In 1998, after d eregulation, much of the citrus marketing was shifted to the FPEF. Registered as a non operate under expo rters who are voluntarily associated with the FPEF. As part of their vision, t hese members can be expelled if they fail to follow the code of conduct they are required to sign upon becoming participating members. This will exclude them from access to the include: Through screening, an established accredited list of citrus exporters growers is identified that growers are encouraged to work with. Defined contracts in terms of requirements b etween grower and exporter. Provision of specified technical information that relates to the handling and management of produce.
35 Promotion of fruit product in collaboration with the South African government at international trade fairs (Philip, 2006). In a dditional, to assist with marketing, South Africa joined a number of international organizations to promote stronger relationships. These include Comit de Association of Fresh Fruit Exporters (SHAFFE). CLAM was founded to unify and represent the interests of citrus producers in the Mediterranean Basin. As a member, the CGA receives information relating to products and exp orts from CLAM and its members (Citrus Growers' Associat ion of South Africa, 2012). SHAFFE was formed as a trading body to represent the interests of growers and exporters in the southern hemisphere. Their association includes all the major southern hemisphere exporters including: Argentina, Australia, Brazil Chile, New Zealand, Peru, South Africa, and Uruguay. This joint initiative supplies its members with information regarding shipped volumes and export destinations, while attempting to promote free trade and improve access into northern hemisphere market s (Citrus Growers' Association of South Africa, 2012). Labor and Land Issues The SACCE initiated several agricultural projects in the 1980s to assist traditional black farmers. However, these farmers struggled to adhere to the strict protocols of the tr ansforming industry, and lost the financial assistance available under the single exporter regime. Consequently, this shift would end citrus farming in the homeland areas (Mather 2008 deregulatio n of the citrus industry created a smaller, but highly skilled work force, deregulation also resulted in a growing surplus of jobless casual workers. One hundred
36 thousand permanent workers, and an unknown number of seasonal workers, were estimated as empl oyed at the end of the regulated period ( Mather ). Although high unemployment rates were seen during pre and post regulation, producers have still found it difficult to obtain seasonal workers. The U.S International Trade Commission links this anomaly to attributable to its high HIV/AIDS rate of 20 percent, and the undesirable nature of labor intensive, an d low paying, agricultural jobs (2006). Minimum wages for farm workers during the period of March 2009 t o March 2013 is presented in the T able 2 4. Consumer price index (CPI) was used in the adjustment of wages in which wages increases each year are specified as CPI + 1% (Citrus Growers' Association of South Africa, 2012). Deregulation was influenced, and i nitiated, by the new government which ended apartheid. During the apartheid regime, many black South Africans were forcibly removed from their homes. Circumstances such as this loss of personal land, and other situations resulting from apartheid, have caus ed much debate. Attempts have been made, and discussions are still underway, to address and compensate the wrongs of apartheid and those affected. owned by white farmers and then to transfer the ownership to blacks. The government set an aim of transferring 24.5 million hectares by 2014, however only 6.8 million hectares have been distributed by 2011 ( South African Government News Agency, 2012). Th e approximate 30% acquisition of total owned white farms has had mixed reactions. Lahiff explains:
37 While market assisted agrarian reform (MLAR) in South Africa has undoubtedly had some success in terms of transferring land and in not antagonizing landowner s, the complexity of the process, its slow pace and its inability to effectively target the most needy households or the most appropriate land (especially in terms of plot sizes) makes it unlikely that it can ever be a means of large scale redistribution o r poverty alleviation. In Africa is little more than a programme of assi sted purchase, masquerading landowners and a small minority of better off black entrepreneurs (2007 p.1592 ). Land reform, and its potential repercussions, is likely to still be a controversial of land to sufficient beneficiaries, while instilling the financial and educational responsibilities to smallholding (or non) farmers who inherit this l and by the government is of great importance, and incredibly multifaceted. Shipping and Logistics Mather argues that liberalization of the industry inevitably complicated the market. He identifies the single channel operator as a key figure in managing qua lity problems when Sout h African fruit was in transit (2008). Traditionally, fruit that needed to be diverted was sent to less discriminating markets such as Eastern Europe or Russia. European markets were able to take advantage of the new, less experienc ed, exporting agents who were more inclined to sell the fruit at any price than divert to other destinations. This severely drove down the price for South African citrus, frustrating many growers who previously distanced themselves from more established r eputable exporters. The regulated period saw many developments in both shipping techniques and technologies. The Far East, particularly Japan, was proving to be a complex and strenuous market. Additional phytosanitary regulations were imposed on all frui t going
38 to Japan, causing South African fruit to reach its destination over six weeks from its harvest date. In a joint venture by the Department of Agriculture, Unifruco, Outspan and the PPECB, attempts were made to convince Japanese authorities that fru it quality can be improved through quicker and more efficient means without compromising quarantine regulations (de Beer, Paterson and Olivier, 2003). Quarantine tests on board ships proved these results were attainable, allowing produce to reach the Japan ese market three weeks earlier with more desirable attributes. Amber Cherry was the first vessel to hold a consignment of citrus using the new sterilization technology. This consignment was approved unconditionally and ultimately opened additional, and m ore lucrative, trade to Japan and other Far East countries. Today, citrus is exported through a technically advanced, market driven, flexible, and customer focused operation. This gives rise to greater customization, choice, and accountability; however, higher risks and higher costs resulting from poor communication are a risk (Fresh Produce Export Forum, 2010). Transporting citrus is accomplished primarily through ocean and air cargo. Delicacy of the fruit and the cost of transport play two important ro les when determining a specific transport method. Although longer, ocean cargo is the more cost effective method. Before deregulation, only 15% of citrus was shipped by refrigerated container, while the majority was exported through conventional vessels. The transition of shipping techniques (with respect to volumes) occurred in 2006. Currently, the majority of ship ping utilizes the former method (Figure 2 6).
39 Domestic logistics are still far from being considered an efficiently run operation. In 2010, t he CGA attempted to implement a citrus integrated transport management system. The project was initiated to address the issue of trucking into Durban where, since 2007, a major shift from rail to road transportation has occurred. Along with increased cit rus production during this period, Durban now receives over 32,000 trucks annually and costs growers approximately R72 million 5 i n truck delays (Citrus Growers' Association of South Africa, 2012). The application of this project was turned down due to the be pursued in a different area where additional funds can be obtained. The Maputo Port also faced challenges due to citrus being routed to Durban. The Maputo Port is sit uated approximately 450 km closer to citrus production in the north than Durban. However, transport and port costs do not subsidize the movement of fruit through the port due to its high operational costs. The option of operating through Maputo to Russia should be seen as very advantageous for producers in the north; nevertheless, they are forced to divert their product due to insufficient facilities offered by the port, and the overall high cost it necessitates (Citrus Growers' Association of South Afric a, 2012). New Plantings South Africa experienced some of its higher plantings during the latter years of the regulated period. Plantings ranged between three and five million trees per year (Figure 2 7) wi ranging between o ne and two million trees each year. Overall plantings have since decreased, averaging between two and three 5 Approximately $9 million (U.S Dollars) 2012
40 million trees during the past decade. Along with Valencias, navels and mandarin hybrids have become three of the most popular varieties in modern times (Table 2 5) Performance post regulated export period is examined. Along with South Africa, market shares of leading citrus export industries are compared during these tw o eras. A volume export comparison to traditional significant South African citrus markets is also presented. The broad classification of citrus types has been arguably accepted as oranges, mandarins (or soft citrus), lemons and grapefruit. These terms, however, only serve as a branch to an underl ying array of varieties contained within it. For simplicity, these all encompassing definitions are used in presenting results in this section. Market Share Spain has long dominated the majority of citrus expor ts, specifically oranges and mandarins. The growth of South African citrus, particular ly in its orange sector, has been well documented (USDA, 2011) during this modern era. Along with South Africa, Egypt has captured significant export market shares at th e expense of Spain and the United States (Figure 2 8). Still dominant Spain has experienced sharp declines of over 20% in mandarin transformation of 1997 to 2010, this decli presence in t he market. In its emerging citrus exports, South Africa has remained a small scale grower and exporter of mandarins (soft citrus) in comparison to the lar gest markets of this variety. This decline is more attributable to the expansion of citrus
41 production in China Pakistan and Turkey which have all come close to, or exceeded a growth of over twofold (Figure 2 9). The United States has experienced declining grapefruit production and exports more rec ently, attributable to proposed findings of the fruit s interaction with prescribed medication ( Bailey, Dresser and Arnold, 2013). Bailey proposes that a n increased number of medication s are available that when interacting with the fruit, can have advers e effects for the consumer. These findings have aided the decline of grapefruit production within the United States, with consumption dropping from nearly one million metric tons in the 1970 s to around 300,000 in 2012 (IndexMundi, 2013). South Africa, T urkey and specifically China have been some of the primary exporters during this decline. China increase has been from an approximate 1% in 2007 to over 14% of market share in 2012 (Figure 2 10). Continuing the citus variety definitions, lemons and lim es are classified as the fo u rth citrus type While production and exports might differ server ely for these two products when ident ified seperetely, export shares of this inclusive variety is evenly spread across Spain, Turkey and Mexico. South Africa r emains a low sca le exporter of lemons and limes; however, more than doubling their market share from the pre deregulated export period (Figure 2 11). Major Export Markets by Volume The Netherla orange im porters. Emerging markets such as Saudi Arabia, United Arab Emirates and Russia previous p rimary importer of oranges, has declined sharply during this period. The gr owing demand in the Middle East as well as willingness to acquire lower
42 grade fruit, have influenced the emergence of the se previous, less recognizable markets (Figure 2 12). While of smaller volume compared to oranges, mandarins (soft citrus) h ave be come increasingly popular with higher end retailers, especially those within the United Kingdom. The South African industry has experienced increased exports to all traditional major importers of the fruit, including the Netherlands, Canada, United States and the United Arab Emirates (Figure 2 13). Japan and the Netherlands continue to import the majority of South African grapefruit. The transition in adopting Netherland s as the primary European port over Belgium is also demonstrated here (Figure 2 14). A once indistinguishable market, Netherlands, or more precisely Rotterdam, is now the third largest port in the world (after Shanghai and Singapore) and continues to import the majority of South African citrus in serving its surrounding countries s uch as Luxemburg, Germany and France (FINPRO, 2008). A principal entity in Middle Eastern religious events, Sou th African lemons continued to be acquired by most notably the United Arab Emirates and Saudi Arabia. Other notable increases during the post deregulated period have been experienced in the United Kingdom, Netherlands, Hong Kong and Russia (Figure 2 15). Production In comparison to the regulated export period, the industry has achieved an increase in production of three of its four varieties ( Figure 2 16). Oranges, grapefruit, lemon, and limes volumes have increased, while production of mandarins (soft citrus) has reduced. produced variety, have increased in yield s of close to 200% from 1995 to 2011. Grapefruit, S citrus
43 crop, has increased by 317% while lemon and lime production increased by over 350% during this same time period. Soft citrus production, which ranged from 40 to 50 thousand tons during 1995 1996, has decreased to just above 26 thousand tons in 2011. Gross Value Through its amplified production, gross value of total production has increased across all citrus varieties (Figure 2 17) citrus variety, worth approximately R4 257 b illion 6 in 2011. This is an increase of over 400% in comparison to the end of the regulated export period. Gross value of grapefruit and lemon varieties has increased by 631% and 1022% respectively. Soft citrus, even with decreased production volumes, has increased its total value of approximately R47 million 7 in 1995 1996 to R92 million 8 in 2011. Export Percentage of Total Production Export percentages of oranges and lemons have increased gradually during the deregulated export period, while grapefruit e xport percentage yield of total production has remained more stable (Figure 2 18) These three varieties were exported at approximate 54, 52 and 59 percentage volumes of its equivalent total production during the end of the regulated period. Oranges, lemo ns and lime s and grapefruit were, respectively, last reported at 62, 60 and 51 percent during the 2011 season. The 6 Approximately U.S $473 million (2012) 7 Approximately U.S $5.2 million (2012) 8 Approximately U.S $10.2 million (2012)
44 oranges and lemon industr ies are currently experiencing some of their l argest exports percentage of production constantly exporting over 6 0% during the last three years. Concluding Statements Through an analysis of governmental and competitive conditions evolving and emerging organizational bodies within the South African citrus industry are attempting to identify and resolve perceived s hortcomings of citrus exports. By means of the CGA, PFEF, CRI and the South African Department of Agriculture, the industry has created security for producers and exporters Such securities include providing export data information, trading and technica l guidelines, promotion al initiatives, as well as packaging and safety requirements. The CGA also provides a transparent business operation through its annual publication of all received levy funds and corresponding research and project expenditures Prod ucers within continue being consolidated Many of the smaller citrus producers during the regulated period have either dissolved or have been acquired by larger citrus producers. The progressio n of consolidation has; howeve r, declined while some larger producers have created their own export brand and a ttained the responsibilities which follow. Whether exported through its own branch or a stand alone exporter, th e industry continues struggling with expectations enforced by foreign markets. These expectations include the potential ban of imports to the European Union (EU) if five or more shipments are identified as containing traces of CBS within the season This arguably remains tariff barrier t o trade. Domestically the industry continues to struggle with higher operation costs, including electricity and
45 heightened minimum wages for farm laborers. Along with land distribution, th ese elements continue to stunt production and export performance. T he EU is and will remain the predominant market for South African citrus. Nevertheless emerging markets such as India, Pakistan and the Far East are becoming increasingly attractive for South African exporters. This expanded marke t access only bodes w ell for this industry as it faces increasingly stricter trade guidelines from their traditional European markets.
46 Figure 2 1. Citrus producing r egions within South Africa (Source: Citrus Growers' Association of South Africa 2012 Adapted)
47 A B Figure 2 2. Orange exports comparison, South Africa vs. northern h emisph ere c ompetitors A) Comparison from January 1997 to December 1998 B) Comparison from January 2010 to December 2011 ( Source: Global Trade Informati on Services, 2012 )
48 A B Figure 2 3. Mandarin exports comparison, South Africa vs. northern hemisphere c ompetitors A) Comparison from January 1997 to December 1998 B) Comparison from January 2010 to December 2011 ( Source: Global Trade Information Services, 2012)
49 A B Figure 2 4. Grapefruit exports comparison, South Africa vs. northern h emisphere Competitors A) Comparison from January 1997 to December 1998 B) Comparison from January 2010 to December 2011 ( Source: Global Trade Information Services, 2012)
50 A B Figure 2 5. Lemo n exports comparison, South Africa vs. northern hemisphere c ompetitors A) Comparison from January 1997 to December 1998 B) Comparison from January 2010 to December 2011 ( Source: Global Trade Information Services, 2012)
51 Figure 2 6 Comparison of shippi ng techniques in the South African citrus industry, m illions of c artons (Source: Fresh Produce Export Forum 2010 adapted) Figure 2 7 New c itrus t ree p lanting s in South Africa (Source: Fresh Produce Export Forum 2010 adapted)
52 A B Figure 2 8 Market share of orange exports amongst five largest e xporters. A) Year 1997 B) Year 2010 ( Source: Food and Agricultural Organization of the United Nations )
53 A B Figure 2 9 Market s hare of m andarin exports amongst five largest e xporters. A) Year 1 997 B) Year 2010 (Source: Food and Agricultural Organization of the United Nations )
54 A B Figure 2 10 Market s hare of g rapefruit exports amongst five largest e xporters. A) Year 1997 B) Year 2010 (Source: Food and Agricultural Organization of the Unit ed Nations )
55 A B Figure 2 11 Market s hare of l emon exports amongst five l argest e xporters. A) Year 1997 B) Year 2010 (Source: Food and Agricultural Organization of the United Nations )
56 Figure 2 12. Major m arket o range e xport d estinations, 1000KG (Source: Global Trade Information Services ) Figure 2 1 3 Major m arket m andarin export d estinations, 1000KG (Source: Global Trade Information Services )
57 Figure 2 1 4 Major m arket g rapefruit export d estinations, 1000KG (Source: Global Trade Information Services ) Figure 2 15. Major market lemon export d estinations, 1000KG (Source: Global Trade Information Services)
58 A B C D Figure 2 16. South African production q uantities of citrus v arieties A) Oranges, B) Lemons, C) Grapefruit, D) Soft c itr us (Source: Agricultural, Forestry and Fisheries Department of South Africa, 2012)
59 A B C D Figure 2 1 7 South African gross value (R1000) of citrus v arieties A) Oranges, B) Lemons, C) Grapefruit, D) Soft Citrus (Source: Agricultural, Forestry an d Fisheries Department of South Africa, 2012)
60 A B C Figure 2 1 8 South African export p ercentage of citrus v arieties A) Oranges, B) Lemons, C) Grapefruit, Soft c itrus data u navailable (Source: Agricultural, Forestry and Fisheries Departme nt of South Africa, 2012)
61 Table 2 1. Total hectares of production per v ariety (Source: Citrus Growers' Association of South Africa, 2012 Adapted) Region Grapefruit & Pummelos Lemon & Lime Navel Soft Orange Valencia & Midseasons Total Hectares Eastern Cape 234 2021 5185 1631 3438 12509 Free State 0 0 0 0 0 0 Gauteng 0 0 0 0 0 0 Kwa Zulu Natal 1650 335 571 49 801 3406 Limpopo 4048 1422 3945 586 15672 25673 Mpumalanga 2371 209 929 451 2857 6817 North West 0 0 0 0 0 0 Northern Cape 324 94 3 37 64 397 1216 Swaziland 913 57 34 52 719 1775 Western Cape 24 589 3831 2368 2150 8962 Total Hectares 9564 4727 14832 5201 26034 60358 Table 2 2 Comparison of pre and post deregulated citrus producing land a reas (Source: USDA ,1995 and Citrus Grower s' Association of South Africa, 2012 Adapted) Region Regulated (Early 1990's) Deregulated (2012) Change % Eastern Transvaal / Mpumalanga 14008 6817 51.33 Northern Province / Limpopo 12126 25673 111.72 North West / Northen Cape 800 1216 52.00 Wester n Cape 7712 8962 16.21 Eastern Cape 9506 12509 31.59 Natal / KwaZulu Natal 3269 3406 4.19 Swaziland 2119 1775 16.23 Total Hectares 49540 60358 21.84 Table 2 3 Per c apita consumption of citrus in South Africa, Kg per year (Source: Agricultural, For estry and Fisheries Department of South Africa 2012) Year 1995 1996 2002 2003 2010 2011 Citrus 15.48 14.83 18.51 16.68 11.73 10.6
62 Table 2 4 Minimum wages for farm workers in South Africa 2009 2013 (Source: WageIndicator Foundation, 2012 an d Freshfruitportal, 2013 ) Year / Frequency Jan 09 Jan 10 Jan 11 Mar 13 Hourly R6.31 R6.74 R7.04 R11.66 Weekly R284.23 R303.84 R317.51 R525.00 Monthly R1231.70 R1316.69 R1375.94 R2274.82 Table 2 5 New t ree plantings per v ariety in South Afri ca (Source: CGA, 2012 ) Variety 1996 2002 2011 Clementine 182774 51200 86310 Edllendale 5935 0 1400 Grapefruit 540909 224180 58190 Kumquat 30910 3400 10850 Lemon 607742 138120 741933 Lime 2410 36835 25300 Mandarin Hybrid 192872 176720 656503 N avel 722502 857885 533226 Satsuma 101451 94200 218060 Valenica 1626530 683572 604920 Other 69204 17870 8686 Total 4083239 2283982 2945378
63 CHAPTER 3 THE EXPORT MARKET FOR FRESH SOUTH AFRICAN CITRUS Introductory Statements South Africa is ranked as th e 14th largest producer of fresh citrus with approximately 2.2 MMT of citrus produced during the 2010 season (Fresh Produce to land area planted to citrus ( Central I ntelligence Agency, 2009 ). Considering its global position, its ranking on world citrus production is hardly surprising. What is surprising, however, is that South Africa ranks as the second largest exporter of fresh citrus in the world, trailing only Spai n as the largest exporter. Almost 1.5 MMT were exported during the 2010/2011 season, which was comprised mainly of Valencias (42%), navel (26%) approximately R5.8 billion 1 to the eco nomy of South Africa during the 2009 season, the citrus industry is the third largest horticultural industry in South Africa, falling behind the vegetable and deciduous fruit industries respectively ( Agricultural, Forestry and Fisheries Department of Sout h Africa, 2011) Dependent mostly on their exports (approximately 62 percent of citrus production is exported), the South African citrus industry has been regarded as a supplier of high quality fruit in terms of taste, color, and size. They continue to stri ve to meet global demands even through increased competition in global markets (USDA, 2011). The purpose of this chapter is two fold; (1) to estimate appropriate import demand models for citrus commodities produced within South Africa and (2) to use the es timated equations to estimate optimal allocation of fresh citrus fruit across markets. 1 Approximately 800 million US dollars
64 The demand estimation is attempted by adopting previous frameworks used in constructing demand models, while also considering the domestic environment of the importing nation in setting up the model. The estimated own price demand elasticities are used to construct a quadratic programming model whose objective function is industry revenue. Solution of the quadratic programming model using the actual crops produced over the three seasons 2008 through 2010 is compared with observed fruit shipments. Comparison of these values provide s insight into the question of whether the industry is finding the revenue maximizing allocation of fruit and to what extent price discrimina tion will occur in achieving optimal revenue. The paper is organized as follows. Section I is a literature review of import demand estimation while Section II describes the models and data used. Results within commodities and between countries are discus sed in Section III and Section IV. Finally, Section V contains a summary of the paper and concluding remarks. Literature Review There are a considerable number of studies which examine factors influencing the degree of trade for agricultural commodities. While previous studies relating to South African citrus trade are not extensive, past reviews and discussions within the field of agricultural trade allow us to explore the dynamics behind nuances of the industry. Thursby and Thursby comment that the preci se specification of an import demand equation is an empirical issue (1984). This, in part, is due to the little guidance offered by economic theory on what the appropriate functional form should be. In constructing critical economic indicators that affec t demand, Chambers and Just (1981), Lee and Fairchild (1988), Arize, Osang and Slottje (2000) and Orden (2002) examine
65 the effect of exchange rates on U.S agricultural commodity markets, grapefruit foreign demand, and export flows on less developed countri es respectively. In addition to foreign exchange, Bahmani Oskoee (1984), Arize (2001) and Hussein (2009) define predictors. Other notable contributors in estimating simple ag gregate import demand functions include Warner and Kreinin (1983), Thursby (1988), and Tegene (1989). Thursby and Thursby considered different alternatives of import demands consisting of combinations of quantities, prices, GDPs, as well as the addition o f these variables lagged, where the appropriate model was defined as one that generates unbiased and efficient elasticity estimates. This was still viewed as appropriate even if it was an untrue representation of the import demand (1984). An additional co nsideration in constructing import demand models would be t hat of the functional form within the specific trade relationship. Past literature presents the linear and log linear formulations as the two traditional functional forms in modeling these trade relationships (Kreinin, 1967, Houthakker and Magee, 1969, Leamer and study that, based on the Box Cox procedure, empirically proves and confirms previous work done by Khan a nd Ross (1977) in assuming the log linear relationship as the appropriate functional form. Along with their generally superior fit, they are also predominantly used on account of their ease in interpretation (Houthakker and Magee, 1969). Khan and Ross co nfirm this viewpoint -that on grounds of convenience, the log the traditional linear relationship (1977).
66 Single equation modeling has remained a popular method in estim ating demand equations that has previously been implemented by Rosson, Hammig and Jones (1986), Halliburton and Henneberry (1995), Aviphant, Lee and Seale (1990) and Onunkwo and Epperson (2000). Lee, Wang and Kennedy (2007) attribute this to the ability o f independently estimating the demand for the specific commodity with predefined variables that were deemed necessary for the model, the flexibility of the data, as well as the computational ease associated in estimating the model. Additionally, Binkley ( 1981) advocates the use of single equation methods in estimating import demands as it removes the possibility of allowing for simultaneous effects. Crucial economic variables were selected in recognition of this literature to construct import demand models of South African citrus varieties. These variables also took into account the transition of currencies used within each country. Model Specification which include: Belgium, G ermany, United Kingdom, Netherlands, Russia, Japan, China, Saudi Arabia and the United States. Two single equation log linear models were used to represent the import demand relationship for each country. 2 For those countries whose transactions take plac e in U S Dollars (Russia, Japan, China, Saudi Arabia and the United States) in trade of South African citrus, both estimated models represented a relationship between the commodity quantity imported to the import price of that commodity, nominal exchange r ate between their domestic currency and US dollar and nominal Gross Domestic Product (GDP). Additionally, the 2 While most major markets typically import sign ificant quantities all four citrus types (oranges, grapefruit, soft citrus and lemons), some markets only import specific varieties.
67 second log linear model contains a GDP deflator in which the base year is country specific. Monthly indicator variables are included to compare t rade activity between months. The indicator base month was country and commodity specific, and was chosen based on historical figures that indicated the most active month of trade. The two multiple regressions models used to estimate the import demand for each citrus variety for countries trading in US dollars: logQCitrus ij 0 1 logPrice ij 2 logExR j 3 logGDP j 4 Month 1 + 5 Month 2 + 6 Month 3 7 Month 4 8 Month 5 9 Month 6 10 Month 7 11 Month 8 12 Month 9 + 13 Month 10 14 Month 11 i (3 1) l ogQCitrus ij 0 1 logPrice ij 2 logExR j 3 logGDP j 15 logDeflator j + 4 Month 1 5 Month 2 6 Month 3 7 Month 4 8 Month 5 9 Month 6 10 Month 7 + 11 Month 8 12 Month 9 13 Month 10 14 Month 11 i (3 2) where Price ij is the importing price of co mmodity i to country j, ExR j the exchange rate of country j in comparison to the U.S. dollar, Deflator j the country specific deflator variable and eleven Month indicator variables accounting for activity of citrus imports. Countries currently conducting t rade agreements in Euros, but previously operated using domestic currency before the establishment of the Euro (Netherlands, Germany and Belgium), resulted in two alternative models. In such circumstances, the inclusion of an exchange rate variable is dee med unnecessary. From a demand currency is of no relevance to the model as the transaction is operated under one currency. It is only from the supply perspective that an exchange rate should be seen as relevant. This results in the replacement of the exchange rate variable as used in
68 the two models previously with an indicator variable that defines the transition of the cluded to provide statistical evidence of potential trade distortions when operating under different currencies. The two multiple regressions models used to estimate the import demand for each citrus commodity for countries trading in Euros (and previously in their domestic currency): logQCitrus ij 0 1 logPrice ij 16 Indicator j 3 logGDP j 4 Month 1 5 Month 2 6 Month 3 7 Month 4 8 Month 5 9 Month 6 10 Month 7 11 Month 8 12 Month 9 + 13 Month 10 14 Month 11 i (3 3) logQCitrus ij 0 1 logPrice ij 16 Indicator j 3 logGDP j 15 lo gDeflator j + 4 Month 1 5 Month 2 6 Month 3 7 Month 4 8 Month 5 9 Month 6 10 Month 7 + 11 Month 8 12 Month 9 13 Month 10 14 Month 11 i (3 4) Models ( 3 3) and ( 3 4) were used for estimating import demand equations from the United Kingdom, however w ith the removal of the indicator variable. The United Kingdom has consistently traded with South Africa under their domestic currency and thus no variable indicating a transition between currencies is necessary. The signs for 1 and 2 in model ( 3 1) and ( 3 2) are expected to be negative, while the sign for 3 is expected to be positive in all the models This said, importers are expected to purchase less citrus as price of citrus and exchange rate increases; moreover, the quantity imported is expected t o increase when nominal GDP increases. The expected sign of the deflator, 15 is undetermined. While the expected sign of the indicator, 16 is also undetermined, research from the European Committee suggests that there was an increased number of export ers and products traded across borders as
69 a result from introducing the euro (Baldwin et al., 2008) Previous firms who did not engage in trade were seen now exploring these opportunities, attributable to its inherently advantageous features removal of e xchange rate uncertainties and currency transaction costs. All monthly indicator coefficients were expected to be negative, indicating the quantity of trade that decreased compared to the most active month, which is the base month. Serial correlation is o ften a consequence when operating with time series data. The Durbin Watson statistic, a method that tests for the presence of autocorrelation in the residuals, is employed to measure the efficiency of our estimations. As used by Houthakker and Magee (196 9), the occurrence of a significant Durbin Watson statistic may be the result of the exclusion of relevant variables, therefore indicating the model may be too simple to capture the dynamics of demand. Unbiased coefficients are still produced under the pr esence of autocorrelation; however, lower estimated variances of the parameters are produced which could result in estimates appearing more accurate than they actually are. An obvious problem with this test is the occurrence of missing observations within the time series, as seen with seasonal export products. Citrus production generally occurs during the colder months of the year with exports stopping during the warmer months, December and January in South Africa. The drawback in utilizing the Durbin Wa tson technique with non continuous data is easily identified through the composition of the test statistic. In its composition, the test statistic calculates differences between residuals that are assumed to be adjacent to one another within the time seri only with non zero observations or, in other words, operating only with months that have
70 trade activity, adjacent observations include closing and following season export months. Alternate approa ches in estimating Durbin Watson statistics with missing observations have been explored by Doran (1974) and, more notably, Savin and White (1978). Savin and White found that if computing the usual Durbin Watson statistic, ignoring for missing observation s, that these test statistics are still valid, although the power of such a test may be weaker than some alternative tests. An additional consequence of missing data is the inability to utilize a lagged dependent variable wher e, if attempted, will work f airly well in the removal of autocorrelation when operating under OLS (Keele & Kelly, 2006). The Cochrane Orcutt (C O) estimation procedure is a recommended alternate approach in adjusting for serial correlation. In doing so, the C O procedure transforms the model where the sum of the correlation parameter. Although some studies have found similarities in coefficient estimates when operating under a traditional OLS and C O approach (with the added advantage of correcting autocorrelation through the latter,) ours did not produce a similar result. In comparison, when operating with the OLS models producing the strongest possibility of correlation, the C O approach ade quately adjusted for serial correlation, although producing different coefficient estimates. It was then decided that, though not as powerful with missing observations, we would continue to operate under the traditional Durbin Watson statistic to detect p ossible occurrences of correlation. In doing so, we are allowed to report our original coefficient estimates. Data The data used to estimate the import demand for South African citrus varieties are monthly observations from January 1996 to December 20 10 The Global
71 Trade Atlas provided quantities and prices for citrus varieties including oranges, grapefruit, mandarins, and lemons and limes. All prices were initially expressed in US dollars and later converted to domestic countries currency of trade. From January 1996, Belgium, Netherland and Germany prices were adjusted to Francs, Guilders and Marks respectively until the end of the December 2001 period. Dollar prices were then converted to Euros from January 2002 when the currency was first circul ated in the form of coins and banknotes. Deflator and Nominal Gross Domestic Product (GDP) values were collected from the International Monetary Fund (IMF). GDP values are converted to the respective domestic currency in operation during that respective time period. To account for monthly currency fluctuations for converting GDP values, yearly domestic currency exchange averages were calculated to produce constant yearly GDP values. Conversion of currencies from Dollars to Euros and Pounds was calculated with nominal exchange rates obtained from the US Department of Agriculture (USDA) database. Historical nominal pre Euro exchange rates were obtained from the Federal Reserve Bank of St. Louis. These conversions allowed country profiles to inherit identi cal currencies for commodity prices and GDPs during the entire time period. Results U.S Dollar The estimated import demands for countries trading in U.S dollars are shown in Table 3 1 Monthly indicator variables were not included in the table to allow fo r easier analysis. The goodness of fit measure ( i.e. R square ) for the four commodities varied from 0.517 to 0.749 for Japan, 0.569 to 0.577 for China, 0.349 to 0.846 for Saudi Arabia, 0.653 to 0.851 for the United States and 0.590 to 0.734 for Russia. Al though
72 operating under non continuous monthly data, the Durbin Watson (DW) statistic produced satisfactory results for most of the countries. Estimates for Japan, China and Russia performed the best in this regard with most DW p values close to 20%, altho ugh Saudi Arabia did not perform as well in comparison. Oranges Japan, Saudi Arabia, USA and Russia had own price effects that were statistically significant with expected negative signs. Both models yield comparable price elasticity ranging from a low of 0.5985 (Russia) to a high of 1.5169 (Japan). A negative price elasticity was also estimated for China imports, although this was not statistically significant. Exchange rate effect estimates were not significant for any country under both models when estimating orange imports. Importing only 3% of coefficie nts. GDP estimates under Model 3 1 for China and Russia also produced significant GDP coefficients, however this sig nificance was removed when introducing a deflator effect under Model 3 2 All significant variables under both models for orange imports produced anticipated coefficient signs. Mandarins No estimates were calculated for Japan and China imports due to thei r limited number of South African mandarin imports. All price parameters for the other countries were negative; however, only Russian price estimates yielded statistically significant levels under both models, while price was significan t for Saudi Arabia u nder Model 3 2 Significant price values ranged from 0.88333 to 1.0589. Significant exchange rates effects were produced for Russia while Saudi Arabia once again obtained an extreme non significant exchange rate effect. This, occurring under both model s and all
73 commodities for Saudi Arabia, is due to the fixed exchange rate of the Saudi Arabia riyal to U.S dollar, currently 3 trading at 3.75 riyals to 1 U.S dollar. Except for the Russian GDP estimate in Model 3 2 all significant variables obtained anti cipated coefficient serious financial crisis of the late 1990s. Operating under 2008 dollars, huge disparities financial crisis. This is 10 year period of 1998 (12.397) to 2008 (100). Grapefruit Japan, one of the principal importers of South African grapefruit, sees strongly s ignificant price es timates of 1.3291 under Model 3 1 and 1.44935 under Model 3 2 while Saudi Arabia has a slightly less significant price elasticity estimate of around 0.4. South Africa has only recently increased its grapefruit exports to the United States through the USDA assessment of declaring more citrus black spot f ree areas within the country. The Western Cape was previously the only province allowed to export to the United States as a result of their strict phytosanitary restrictions. Similar t o orange imports, exchange rate did not play significant roles in estimating grapefruit imports when operating under a 5% significance level. Nominal GDP displayed strongly significant results for Russia, but this significance is once again removed when i ntroducing a deflator affect. All significant variables produced expected coefficient signs. 3 Year 2012
74 Lemons Saudi Arabia, as part of the Middle East, comprises approximately 17% of South African exports. History has shown that much of these exports are used pri marily to provide households with juice produced from fresh squeezed fruit. Traditionally Saudi its large Muslin population and corresponding religious events (Fresh Pro duce Export Forum, 2010) Price and GDP play significant roles within both models with expected coefficient signs. The only unexpected significant coefficient sign is for Russia, which is addressed earlier. The results suggest that price is the most signif icant attribute when estimating import demands for countries operating in the U.S dollar currency. GDP is seen as the next influential characteristic during these estimations. Most countries attained some level of a significant positive GDP estimate, alt hough this occurrence was less prevalent when adding a deflator variable. Exchange rate was seen as the least influential predicator, with estimates for Russia only experiencing any level of significance. In months of operation were mostly shared between July and August, while January and February often experienced little, and even no, exports. European Euro The estimated import demands for countries operating in Euros are shown in Table 3 2 Monthly indicat or variables parameter estimates were once again not included. The goodness of fit measure for the four commodities varied from 0.506 to 0.808 (Belgium), 0.645 to 0.841 (Netherlands), 0.644 to 0.690 (Germany) and 0.494 to 0.841 for the United Kingdom. Ope rating once more under non continuous monthly
75 data, the Durbin Watson test statistic estimates suggested the presence of serial correlation among the residuals for most of the European countries. Europe The European market can essentially be divided among three regions: the United Kingdom and Northern Europe, Southern Europe, and Eastern Europe. Although being respect to its fruit requirements. UK and Northern Europ e consumers are more willing to pay premium prices for higher quality and new varieties of citrus. Much attention is Europe has been described as more cautious with respect to trade with South Africa. Being relatively new export destinations, Southern European countries have been more inclined to protect their own citrus producers through stric ter import requirements, while Forum, 2010). Oranges Price effect estimates were significant across Belgium, Netherlands and the United Kingdom. These significant price ef fects ranged from a low of 1.0252 (Belgium) to a high of 2.32883 (Netherlands) under Model 3 3 while experiencing a low of 1.1024 (Belgium) and a high of 2.24435 (Netherlands) under Model 3 4 It is important to note the logistical roles of Belgium and Netherlands within the European Union (EU). Both EU partners (Government of Canada, 2012 ). This can be seen as a strong explanation
76 capita GDP, this negative estimated GDP does not allow for the added explanation of the role it serves to the rest of the continent. Surrounding c considered and incorporated into the estimation to capture this true effect. Most surprisingly, Belgium is the only country that attained significant negative currency indicator estimates, not only for oranges but also grapefruit and lemons, going against previous findings of B aldwin et al. (2008). Mandarins Price and GDP once again played significant roles in estimating imports for the Netherlands and the United Kingdom regions. The United Kingdom has typically been the leading i mporting nation of soft citrus from South Africa; however the industry is expanding in other areas through increased opportunities in Russia, U.S.A and the Middle East (Agricultural, Forestry and Fisheries Department of South Africa, 2010). Germany genera lly imports a very limited number of mandarins from South Africa and so no import models were estimated for the country. All significant variables had correctly anticipated coefficient signs. Indicator and deflator variables playe d no significant role in Model 3 4 Grapefruit This is assisted through a duty free agreement for grapefruit trade between the two nations, while Japan imposes a ten percent duty. This is seen as significantly higher duty fee and the US duty free imports of grapefruit (Agricultural, Forestry and Fisheries Department of South Africa, 2011). Price was estimated as s trongly significant for the Netherlands, Germany, and the United Kingdom, while also producing significant
77 positive GDP results. Germany and the Netherlands encountered increased imports gesting otherwise. Lemons All estimates produced strong, to very strong, significant levels for price. The effect of GDP is strongly significant when estimating import demands for the Netherlands and the United Kingdom under Model 3 3 which is then remo ved through the addition of the deflator variable. In agreement with the previous set of estimations, European import quantities are strongly influenced by price. This is further illustrated by 26 of the 30 estimations reporting some level of significance on the price variable, all attaining the anticipated coefficient sign. The occurrence of signifi cant GDP estimates under Model 3 3 was frequently removed by the addition of the deflator variable. This addition also resulted in the removal of significant estimation results. Optimal Allocation across Export Markets A mathematical programming model that maximizes revenues subject to a constraint can be defined as: w here P i and Q i is the price paid and quantity purchased by importing country i allocation of the fruit not exceeding actual total year specific export quantities.
78 The objective is to ca lculate an optimal allocation of fruit within each commodity that maximizes industry revenue. Average weighted season prices and aggregate quantities are required in estimating optimal allocations for each season. This is in contrast to the previous mont hly price and quantity data used when estimating the import demand equations. Quantity depended prices are estimated through inverting and the linearization of the respective import demand equations giving (3 5) where P jct is a weighted price paid by country j for commodity c (oranges, mandarins, grapefruit and lemons) in t ime period t (2010, 2009, 2008). We then define (3 6) where R jct is the revenue from country j importing commodity c during time period t. In obtaining est jct and jct, we replicate our log linear import demands as a function of quantity and price, using the previous estimated elasticities as the price effects. (3 7) (3 8) We substitute ( 3 8) along with average weighted season prices and aggregate jct such that
79 In a standard revenue maximizing model, summing across the different regions with respect to price and quantity forms the objective function. To simplify the exposition, if all demand equations are of the form and there are 9 demand regions, the objective function can be defined as: where Q jct defines the optimal quantity of citrus variety c imported by country j in season t and Q jct the total actual amount exported. The estimated optimal quantities can therefore not exceed the actual quantity exported during that particular year. Under such a scenario, export quantities should be al located such that marginal revenues are optimal quantities within each year and commodity will ther efore be that of the amounts, Q jct, satisfying all first order conditi ons given by: Empirical Results The empirical results of this analysis a re summar ized and displayed in Figure 3 4 and 3 5 Optimal quantities maximizing total revenue are
80 presented in Figure 3 6 These quantities are compared with actual quantitie s exported during the season. A similar comparison for price is presented that compares actual purchasing prices for oranges, mandarins, grapefruit and lemons to optimal maximizing revenue prices across export destinations during 2010. Similar occurrences were observed when comparing quantity and price comparisons during the three seasons across commodities where for application purposes, only 2010 results are presented. Oranges The Netherlands, Russia, and Saudi Arabia have traditionally been major expor t destinations for South African oranges. Other prominent countries include Belgium, China, Japan, United Kingdom, and the U.S.A. Germany, another significant destination, was excluded from the estimation procedure due to their insignificant positive estim ated price effect. Actual 2010 season prices fluctuated between 0.5 to 0.6 US dollars per kilogram with only U.S.A experiencing higher prices of around $1 per kilogram, potentially due to their stricter phytosanitary requirements. Initial estimated opti mal quantities for the 2010 season indicate that China, Russia, and Saudi Arabia were oversupplied by approximately 34% and 28% (Russia and Saudi Arabia) respectively. As a consequence, stronger occurrences of price discrimination would result in order to obtain optimal market revenue, increasing Chinese import prices close to 80% and approximately 47% for Russia and Saudi Arabia. Comprising the majority of the imports, Netherlands was estimated at being undersupplied by 32%, where, if met, should see a price reduction of 14%. In comparison, the orange export industry has, financially, operated within 7% of this theoretical optimal export model for each of the 2008, 2009, and 2010 seasons.
81 Mandarins The majority of South African mandarin exports are impo rted by the Netherlands and the United Kingdom. Actual average weighted prices varied from 0.65 to 0.9 U.S Dollars per kilo during the 2010 season with only Belgium prices exceeding 1 U.S Dollar per kg. The Netherlands is once again estimated at being und ersupplied (36%) where, if optimal quantity was met, an optimal price 20% lower would follow. The United Kingdom was estimated at operating within 5% of its optimal quantity and price. Under optimal prices, the United States would experience stronger occ urrences of price discrimination. The mandarin export industry was estimated at operating within 7% of optimal revenue for 2009 and 2010, and within 6% for 2008. Grapefruit The Netherlands and Japan, and to a lesser extent, Russia and the United Kingdom, have dominated in terms of South African grapefruit imports for the past few years. Russia was excluded from the allocation procedure due to their insignificant price effect. Actual prices fluctuated between 0.45 (Germany) to 0.6 (Saudi Arabia) U.S Dollars during the 2010 export season. Under the estimated model, exports would increase by 8% and 17% for Japan and the United Kingdom respectively, causing a price reduction effect of 6% (Japan) and 11% (United Kingdom). Import quantities for the Netherlands are estimated at being oversupplied at approximately 12%. With the removal of Russia from the analysis, the grapefruit industry is estimated at operating within 4% of optimal revenue. Lemons and Limes Of the major importers of South African lemons, Russia, Saudi Arabia, and the United Kingdom purchased the commodity at an average weighted price of 0.64, 0.69
82 and 0.64 dollars per kg during 2010. As the second highest importer of lemons during 2010, the Netherlands purchased lemons and limes at a higher rate than the other leading importers approximately 0.82 dollars per kg. The Netherlands, Russia, and Saudi Arabia were all estimated at being oversupplied during the 2008, 2009, and 2010 season where, under the optimal revenue model, strong occurrences of p rice discrimination would result for the latter two countries. Not all resources were allocated across markets when estimating optimal lemon quantities, suggesting that the industry is currently oversupplying the world market. When comparing revenues, the lemon and lime industry of South Africa operated within 7%, 10%, and 11% of optimality during 2008, 2009 and 2010 respectively. Concluding Statements countries is built upon previ ous literature advocating for the inclusion of important predictors such as price, GDP and exchange rate. Extra consideration is used when utilizing the latter due to the different exchange rates inherited by the importing countries through the estimation period, and the different currencies exchanged within any individual transaction. While other literature might encourage more, less, or even different predicators, some important predictors were still not included, that being quality and size of the frui t. Operating under a deregulated system within South Africa, acquiring such information from the citrus industry on a global scale becomes somewhat of an impossible task. Theoretically it is well documented where higher quality fruit would be exported to a nd what sizes are traditionally preferred by certain countries; however, empirically, this information was not available to us.
83 The estimate price effects suggests that, even without quality information, the South African citrus industry has operated reaso nably well comparable to their projected optimal. Additionally, their current revenue is obtained through lesser extents of price discrimination, as observed under the optimal model. South Africa has proven itself as a dominant supplier of citrus to wor ld markets and continues to grow, as seen within these past 10 years. This has predominantly been achieved through increased supply from some of their larger markets. While this will most likely be the case in years to come, increased opportunities to ex pand production have been found through other smaller, but emerging, markets such as in Africa.
84 Figure 3 1. World fresh citrus e xports thousand t ons (Source: CGA, 2012)
85 Oranges Japan ( 3 1) China ( 3 1 ) S.Arab ( 3 1 ) USA ( 3 1) Russia ( 3 1) Japan ( 3 2) China ( 3 2) S.Arab ( 3 2) USA ( 3 2) Russia ( 3 2) Constant 27.95121 21.3007 43.7828 69.6771 *** 3.3398 30.4922 22.3894 40.1618 69.6921 *** 19.0898 Price 1.50908 *** 0.43 0.6036 *** 1.1701 ** 0.598 5 1.51698 *** 0.3541 0.5972 *** 0.858 0.5783 Exchange Rate 0.393332 6.0424 24.2649 0.2547 0.875172 1.9276 21.7328 1.0915 GDP 0.43885 1.9622 ** 0.3095 9.1232 *** 1.1235 *** 0.13247 4.5034 0.1238 18.484 *** 5.723 Deflator 1.91692 11. 2296 0.653 19.5734 6.3532 R 2 / DW 0.698 / 1.940 0.569 / 1.810 0.846 / 1.833 0.844 / 1.623 0.671 / 2.180 0.699 / 1.943 0.577 / 1.854 0.846 / 1.832 0.851 / 1.687 0.674 / 2.155 Mandarins Japan ( 3 1) China ( 3 1) S.A rab ( 3 1) USA ( 3 1) Russia ( 3 1) Japan ( 3 2) China ( 3 2) S.Arab ( 3 2) USA ( 3 2) Russia ( 3 2) Constant 226.589 34.1356 *** 1.75728 9.5335 48.9136 *** 53.66 ** Price 0.7218 0.2871 0.88333 ** 1.0589 0.3193 0.998 ** Exchange Rat e 178.3737 0.94214 5.2283 2.6112 *** GDP 0.4828 4.5858 *** 1.67469 *** 11.1655 25.5623 *** 9.0873 Deflator 15.9016 39.6404 *** 14.9346 ** R 2 / DW 0.544 / 1.457 0.653 / 1.543 0.692 / 1.824 0.572 / 1.618 0.718 / 1.715 0.734 / 1 .810 Grapefruit Japan ( 3 1) China ( 3 1) S.Arab ( 3 1) USA ( 3 1) Russia ( 3 1) Japan ( 3 2) China ( 3 2) S.Arab ( 3 2) USA ( 3 2) Russia ( 3 2) Constant 93.10547 245.9299 0.22094 89.39979 208.8896 7.4701 9 Price 1.3291 *** 0.4446 0.03544 1.44935 *** 0.441 0.04467 Exchange Rate 0.28852 183.693 0.935 3.51162 156.973 1.19969 GDP 5.19758 0.8114 1.13643 *** 0.51154 2.3772 0.45691 Deflator 19.2458 ** 4.8107 2.20359 R 2 / DW 0.713 / 1.856 0.349 / 1.850 0.627 / 2.178 0.749 / 2.093 0.354 / 1.865 0.628 / 2.181 Lemons and Limes Japan ( 3 1) China ( 3 1) S.Arab ( 3 1) USA ( 3 1) Russia ( 3 1) Japan ( 3 2) China ( 3 2) S.Arab ( 3 2) USA ( 3 2) Russia ( 3 2) Constant 41.0693 191.8932 5.13542 26.4924 260.1775 49.89523 Price 0.7117 0.686 ** 0.30011 0.8681 0.75 *** 0.40402 Exchange Rate 1.2411 138.109 0.025814 2.8287 187.144 1.67746 GDP 3.7764 0.536 1.432501 *** 4.9479 6.2433 ** 10.0274 Deflator 7.8388 8.6002 ** 15.8295 R 2 / DW 0.517 / 1.771 0.687 / 1.501 0.590 / 1.404 0.535 / 1.876 0.704 / 1.602 0.621 / 1.517 Figure 3 2 Price, income, GDP, deflator elasticities estimators (For countries operating under U.S dollar during t rade)
86 Oranges Belgium ( 3 3) Netherlands (3 3 ) Germany ( 3 3) U.Kingdom(3 3) Belgium ( 3 4) Netherlands ( 3 4) Germany ( 3 4) U.Kingdom(3 4 ) Constant 45.5045 *** 35.7052 *** 5.27271 3.30704 44.3769 *** 22.22237 43.3917 3.0832 Price 1.0252 ** 2.32883 *** 0.26682 1.56018 *** 1.1024 ** 2.24435 *** 0.8287 1.6216 *** GDP 2.7809 8.7778 8 *** 0.01387 2.08962 *** 8.099 27.58481 ** 6.1156 1.8311 Indicator 14.7519 *** 4.08553 *** 0.59903 34.8097 20.5764 ** 3.2615 Deflator 11.4457 41.7784 20.8807 7.5541 R 2 / DW 0.806 / 1.548 0.760 / 1.019 0.687 / 1.719 0.84 / 1.588 0.808 / 1.550 0.771 / 1.020 0.690 / 1.731 0.841 / 1.608 Mandarins Belgium (3 3 ) Netherlands (3 3 ) Germany (3 3 ) U. Kingdom( 3 3 ) Belgium ( 3 4) Netherlands 3 (4) Germany ( 3 4) U.Kingdom(3 4 ) Constant 8.15198 37.4488 *** 17.1781 *** 8.54928 63.6724 ** 16.2977 Price 0.78858 1.8174 *** 0.94658 *** 0.77609 1.8721 *** 0.93076 ** GDP 2.425358 8.3624 *** 4.4623 *** 4.73141 0.6555 4.94225 Indicator 5.266228 4.1448 *** 13.8625 3.7174 Deflator 4.72176 19.7108 0.95708 R 2 / DW 0.525/ 1.906 0.760 / 1.076 0.761 / 1.339 0.525 / 1.912 0.763 / 1.066 0.761 / 1.340 Grapefruit Belgium ( 3 3) Netherlands ( 3 3) Germany ( 3 3) U. Kingdom( 3 3 ) Belgium ( 3 4) Netherlands ( 3 4) Ger many ( 3 4) U. Kingdom ( 3 4) Constant 34.6227 ** 7.1031 69.9977 28.1256 *** 34.9475 ** 3.6397 129.735 ** 4.2864 Price 0.3748 0.8822 *** 1.9193 *** 1.5647 *** 0.4252 0.893 *** 2.0211 *** 1.2756 *** GDP 2.4851 3.1676 ** 10.5734 6.2442 *** 6.9397 7.0486 1.8256 20.4106 *** Indicator 10.9025 1.6301 ** 5.1443 27.6103 4.9853 3.7423 Deflator 9.2679 8.3153 35.7898 27.531 R 2 / DW 0.634 / 1.550 0.840 / 1.475 0.644 / 1.798 0.494 / 0.923 0.637 / 1.545 0.841 / 1.483 0.659 / 1.849 0.524 / 0.927 Lemons and Limes Belgium (3 3 ) Netherlands ( 3 3) Germany ( 3 3) U. Kingdom( 3 3 ) Belgium ( 3 4) Netherlands ( 3 4) Germany ( 3 4) U. Kingdom ( 3 4) Constant 15.8371 48.9413 *** 24.3469 13.0248 *** 18.027 4 40.6916 9.7905 22.714 *** Price 1.3275 *** 0.8189 1.2583 *** 1.6386 *** 1.25 *** 0.7972 1.2405 *** 1.752 *** GDP 0.3848 8.9824 *** 4.4668 4.2323 *** 11.0717 11.7402 11.604 1.1953 Indicator 3.291 5.2629 *** 3.0052 36.6647 7.6854 7.9932 Deflator 22.8251 6.0863 20.4728 10.6837 R 2 / DW 0.506 / 1.510 0.645 / 1.452 0.663 / 1.649 0.786 / 1.588 0.520 / 1.528 0.645 / 1.453 0.669 / 1.649 0.792 / 1.658 Figure 3 3 Price, income, GDP and deflator elasticities e stimato rs (For the United Kingdom and c ount ries currently operating under euro (domestic currency pre e uro)
87 A B C D Figure 3 4 Actual vs. estimated optimal orange q uantities (MT) 2010 A) Orang es, B) Mandarins, C) Grapefruit, D) Lemons
88 A B C D Figure 3 5 Actual vs. estimated optimal oranges p rices 2010 A) Orange s, B) Mandarins, C) Grapefruit, D) Lemons
89 Figure 3 6 Comparison between actual and estimated o pt imal r evenue A) Oranges, B) Mandarins, C) Grapefruit, D) Lemons
90 CHAPTER 4 VERTICAL INTERGRATION OF FOREIGN RETAILERS WITHIN THE SOUTH AFRICAN CITRUS INDUSTRY Introductory Statements The past 20 years have seen a rapid increase in the trade of fresh fruits and vegetables The estimated value of all fresh horticultural product trade in 1990 was approximately 51 billion USD. This figure increased to 160 billion USD in 2009 -a n increase of over 200 percent (Food and Agricultural Organization of the United Nations). Increase d trade in fresh fruits and vegetables can be attributed to a number of factors including improvement in logistics, decreased trade barriers, and increased consumer demand for these products. South Africa has become a major player in the world market for f resh fruit. Exports of fresh fruit from South Africa increased from 1.1 MMT in 1990 to 3.1 MMT in 2009, an increase of almost 200 percent (Food and Agricultural Organization of the United Nations). Fresh citrus has been an important role component of South increased participation in world fresh fruit trade. South Africa is now the second largest exporter of fresh citrus in the world, following only Spain, with exports of approximately 1.5 MMT in 2010 valued at over 5.8 billion rand 1 (Agricultural Forestry and Fisheries Department of South Africa, 2010) Problem Statement describe the system of trading fresh fruit. The fruit trade chain includes procurement, production, pac kaging, shipment, and delivery to the consumer. Within this chain, 1 Approximately U.S $640 million (2012)
91 numerous components are involved: picking, grading, packing, depots, exporters, importers, and more (Fresh Produce Export Forum, 2010). Within each component, failure or mismanagement of o ne element can affect the chain as a whole; moreover, different commodities may contain differing components within the chain. These operations typically involve separate organizations, each accruing some margin which is inevitably paid by the consumer. E ven though each operation adds additional cost to the consumer, logistics are coordinated by specialists from within that commodity, allowing for the best quality of fruit to reach consumer outlets. Recently the market ( specifically the fresh fruit trade c h ain) is reforming through the declared intentions of mass retail merchandisers predominately those located within the United Kingdom (UK), such as ASDA/Wal Mart and Tesco. Through backward integration, a process that will allow for the control of all, or most of the stages in the production and sales of their products, these mass retailers are actively increasing their within chain presence while also removing cost accruing but service providing players from the chain. With a larger share of the market, what consequences will this transformed chain have on the existing logistic infrastructure? Additionally, how have some of the traditional player s within the chain, such as specialized exporters, responded to this evolving market? Objectives Through a ca se study on South African citrus, we investigate a new export channel and how industry players, including producers and exporters, have responded to this change. P articipant s voluntarily agreed to an interview where they responded to a questionnaire. The se surveys were conducted face to face during a two week visit to
92 South Africa, specifically in the Western and Eastern Cape regions. Interviews were established through the aid of Mr. Peter Turner from Biogold South Africa. Participants were selected on their unique export programs, as well as on export volumes. The majority of South African exporters with respect to volume are reflected in the study; howe ver, this research also cater Additionally, researchers within academic and private industries were questioned, along with one of the southern which markets its own fruit through both direct and indirect channels. A summary of th ese interviews is displayed in Figure 4 1. These responses will as sist in uncovering how the industry perceives the introduction of a more direct export channel, as well as short and long term outcomes to such a change. The framework of the questionnaire was designed from literature proposed by new institutional economic s (NIE), a common source of literature in providing insight as to why institutions like markets, firms and their linkages exist, or how they arise, from various attributes that underlie economic activity. New Institutional Economics The vast array of frame works and propositions surrounding the act of vertical integration allows the learned scholar to draw from a number of resources. O ne needs to be conscious however, of the governance structure between the transacting parties. Williamson states that : Ve rtical integration is only one of many potential vertical governance arrangements that transacting parties may choose from and represents only one component of broader theories of the governance of contractual (1971, p .113).
93 Klein and Shelanski (1995) summarize the range of governance structures as fitting between the two extremes of spot markets: pure simplistic transactions and the fully integrated firm, where ownership and control is encapsulated within a unified bod y. Joskow (2010) explains that traditional views of vertical integration saw the economics. A most obvious foundational body of theory to dra w from would be Coase s nature of the firm article (1937). Coase broadened the research scope for economists by identifying that there are costs in operating within a market, and these costs factor in determining whether to produce internally to avoid these costs. He proposed transact ions would ultimately be undertaken within the firm if costs operating in a market were de emed to be higher. Arrow another pioneer in the field, acknowledged the consideration of positive transaction costs and its role within vertical integration: An in centive for vertical integration is replacement of the costs of buying and selling on the market by the costs of intra firm transfers; the existence of vertical integration may suggest that the costs of operating competitive markets are not zero, as is usu ally assumed in our theoretical analysis (1969, p.48). Much of the transaction cost economic (TCE) literature has built upon the approach of Coase, most notability by Williamson (1975, 1979), who looked at uncertainty, opportunism, frequency of transaction s, and asset specificity as the fundamental determinants in describing transactions. Williamson (1983) further explains asset specificity as inheriting different contexts within it -including site, physical, human asset specificity, as well as dedicated assets. Klein Crawford, and Alchian looked to further emphasize the role of transaction costs by stating that: His (Coase) primary distinction between transactions made within a firm and transactions made in the marketplace may often be too simplistic (1 978, p.326).
94 Their research examined the opportunistic costs, emphasized by Williamson (1975), in appropriating quasi rents of specific assets. Both Williamson and Klein et al. argue opportunistic behavior is likely to occur in situations where an invest ment is worth less outside a relationship than within. Their views are further supported by a world of incomplete contracts that will not fully identify all plausible considerations ex post. The conceptual breakthrough of this framework brought about much attention in terms of empirically testing this theory in the 19 80s. Some notable contributors governance choice include Monteverde and Teece (1982), Anderson and Schmittlein (1984), Levy (1985), MacDonald (1985), Joskow (1985), Anderson and Coughlan (1987), and John and Weitz (1988). In his review of empirical evidence supporting this theory, Joskow states: We have come a long way since 1937. The nature of the f irm and the nature of market relationships between firms ha ve attracted a lot of recent theoretical interest. Relationship specific investments, asymmetric information, and the costs of writing, monitoring, and enforcing contractual relationships have emer ged as the key factors explaining "nonstandard" vertical relationships.... I hope that theoretical and empirical work will continue to have a closer relationship to one another in this area than is typical in industrial organization ( 1988, p.115) Shelanski and Klein state, TCE tries to explain how trading partners choose, from the set of feasible institutional alternatives, the arrangement that offers protection for their relationship specific investments at the lower cost (1995, p. 337) Chiles and McMa s limited scope are behavioral assumption. Opportunism and bounded rationality, the other two behavioral assumptions, wer
95 interacting with these behavioral assumptions, is proposed to be incorporated into the theoretical model. Noorderhaven (1995) assum ption within TCE. He further emphasizes the trivial nature of economic organization without acknowledging bounded rationality and opportunism as important behavioral assumptions in TCE. While TCE was gaining much attention, a closely related body of theor y of vertical integration e merged. Grossman and Hart were fundamental in defining property right s theory (PRT) while identifying the role of vertical integration as an alternative measure to costly contracts. They emphasize ex post quasi rents as a critic al factor when it becomes too costly to specify the desired rights of that asset within the contract. Vertical integration is the purchase of the assets of a supplier (or of a purchaser) for the purpose of acquiring the residual rights of control (1986, p. 716). Hart and Moore (1990) view the Grossman Hart analysis as restrictive. They argue this pape research looks at an asset having ownership from a communal standpoint. They also view how the incentive structure changes according to the integration process. Grossman and Hart (198 6) and Hart and Moore (1990) both emphasize that vertical integration will itself create transaction costs. PRT is unsure of whether more or less vertical integration will result from more efficient contracting institutions (Acemoglu et al, 2009).
96 Whinsto n (2003) distinguishes TCE from PRT in three different ways. While they both identify incomplete contract and ex post quasi rents as explanatory variables for also mor cost of TCE. Thirdly, PRT assumes that, through any governance structure, opportunism is present, unlike TCE -which relaxes such severity when the transaction is brought within the firm. The growing literature of NIE understandably provides a relevant framework in constructing questions for our research objective. With it, the industry is able to provide insight as to why the newer, direct, sourcing of fruit was introduced. Addi tionally, producers provide an understanding of which export model they prefer, while traditional exporters attempt to explain their future role in the industry, as well as why some suppliers might find the newer, direct, supply chain preferable. UK Retai lers The United Kingdom (UK) and northern Europe ha ve long been among South immensely from the previous large bulk volumes bought at spot markets in Brussels in previous years, to a multidimensional plethora of traditional, hybrid, and direct channels. Today, the larger UK retail stores ar suppliers wanting to access consumers. This comes from restructuring changes in the 1980s and 1990s in which the retailers began to claim more value from manufacturers in the cost chain. Through mergers and acquisitions, they have gained additional control of supply chains through the investment of logistical procedures and development of distribution cent ers (Fresh Produce Export Forum, 2010)
97 Much debate has surrounded the introduction of UK retail stores in South Africa who have increased their presence within the fresh fruit supply chain. Stuart Symington, previous CEO of the South African Fresh Produce Exporters Forum and current CEO of the Perishable Products Export Control Board of South Africa, responded to this change during a session of the World Export Development Forum on supply chain implication of ethical business linkages. He had previously pr edicted that by 2010, 60% of fruit would be sold through based reduction charges in which shipping companies can give rebates of about $700 per container at the end of the s eason, if sufficient business was transacted through them. However, some of these rebates are not available anymore due to the acquisition of some logistical procedures. We can no longer produce [our packaging] in South Africa. And the supermarket that or ganized the deal gets the rebate on the packaging contract ... They get the rebates now. They are even buying farms. So they own the whole chain. Symington continues to ask the question: How ethical is this? All the money that we are supposed to be gleani ng in the chain on the supply side is being taken to the demand side. What does that mean? It simply means we pay our laborer s less. We pay our farmers less (International Trade Centre, 2009, p. 3). Symington points out the major ethical problems come from the expectation of over procurement to maintain full capacity of shelves and then, if the produce does not y recognition from buyers as the supplier of high quality fruit.
98 What are you supposed to do with fruit that you have packed in South 87% of fruit in the UK sells through supermarkets. You have to repack all that fruit and move it to the continent at massive expense. You never get told your price. You get told how much you must deliver, in what variety, in which quantity, in which weeks. Price is the wild card. Sometimes you are told wh en your product is on the water. Sometimes you are told when it is in a distribution centre. Sometimes you are even told else. They are debranding all of our products. They put the ir house colors on your fruit. By the time it reaches the UK and many places in Europe you (International Trade Centre, 2009, p. 4). Major Supply Channels In a comparison of traditional (indirect) and the more recent, direct, tra ding practices of UK retailers, the following chapter clarifies the operation of these two extreme export techniques. Additionally, popular alternate export channels to the direct and indirect are presented. These channels would, in some form, deviate un der certain components within the chain, either adding or removing cost accruing players during the logistic process. That said, the modern era export channel has evolved into a complex labyrinth of distribution networks leading to the consumer. Figure 4 2 presents various channels currently utilized between South African producers to UK and Northern European retailers. The first two channels operate under the traditional export method, that being through a pure export agent as well as an importer, and fo rm part of trade that is most closely associated with the European Mainland market. Under channel 1 the producer acts as his own exporter, an increasingly common occurrence within South Africa as smaller and medium producers continue to consolidate with on e another. Under these channels, trade is traditionally initiated by, first, the importer responding to consumer
99 demand. South African exporters are then informed of produce specifications from the importer. The packaging is ordered by the exporter and d farm. Each exporter in South Africa operates under some strategic model to serve their overseas clients. Some of these models include: Selecting only the best quality producers Seeking top end retailer clients Targeting volume based rebates from shipments Assuring favorable payment terms to producers Securing new varieties Providing market splits for the producer The exporter is also entrusted with securing contai ners for the shipping line, contracting road transport, and selecting cold store or port terminals. Lastly, the produce is required to clear customs and load the appropriate vessel before it is shipped out of South Africa. Once the importer receives the p roduct, it is then their responsibility to take control of the product to supply retail supermarkets, wholesale marke ts, hospitality industries, school feeding programs, or government institutions (Fresh Produce Export Forum, 2010). Today, UK retailers are pushing to ope rate under more direct channels, in comparison to channel s 1 and 2, by b ypassing previous cost accruing players taking into context the structure under which the retailer operates. In more extreme instances of vertical integration, retailers have acquired their own farms within South Africa, thus bypassing almost all external involvement. A more common occurrence of vertical integration is the attainme nt of fruit by retailers from the pack house. A medium sized grower that supplies seven retailers directly informed me he only classifies direct
100 2 ; however this definition will be relaxed for this discussion. In s upplying directly to retailers we look at ASDA/Walmart and Tesco, the two largest UK retailers operating under various direct souring programs. Since 2004, ASDA has operated through International Produce Limited (IPL), an importing company fully purchased and incorporated within ASDA in October 2009, as their supplier of fresh citrus. This joint venture was first created by shareholders Bakkavor and Thames in an attempt to create a unique supply chain model for fresh produce. By controlling more of the sup ply chain, some recognized benefits included: reduced costs to the consumer, improved shareholder returns, and greater sustainability for the growers. The success of this model led to Asda/Wal the model that is currently operate d under their franchise name. A direct benefit of their vertically integrated model is the quality control of their produce. Asda quality control presence at each distribution center has added to the success of this channel. This process allows 90% of IP L products to be inspected before shipment, where other suppliers would normally have only 2% checked. With higher quality produce being shipped and lower rejections returned from stores, IPL has offered Asda/Wal mart greater control of their produce whil (Scott, Lundgren, and Thompson, 2011). Supplying some US and Japanese Wal mart stores, IPL has been able to use its scale advantage with, ultimately, an international level impact (Randall and Seth 2011 ). T his form of export is most closely related to that of export channel three. Here, product is shipped directly from the producer to a supermarket client via its category 2 Interview with South African grower/exporter, January 24, 2013, West ern Cape, South Africa.
101 manager, that being IPL for ASDA. In comparison to the indirect channel, IPL removes the presence of the South African exporter. This includes taking control of road transport and shipment lines contracts, thus obtaining larger rebates. Control of fruit is exchanged to IPL traditionally at the pack house; however the risk associated with the product still remains with the producer until final clearance. This export channel also forms some part of Tesco importing distribution. Where IPL is wholly owned by ASDA, Tesco operates under two branches: MMUK and Group Food Sourcing ( GFS). MMUK is an importer of fresh citrus with an exporter division, AMC Fruit, in South Africa. However, they receive the majority of their fruit from other exporters, constituting roughly 75 other branch, GFS, handles the remaining 15 25 direct souring office in South Africa, obtaining most of its fruit from consolidated within the exporting country and, once received on land side, is sent directly to their distribution center. It is said this is where most cost cutting is attained through any distribution. 3 This form of export is most closely related to that of channel five. The reality of direct so ur c ing is that the same variables of operating in trade still remain in the distribution channel 4 A list of these variables include: a dherence to time and temperature protocols, capacity management, appointment of fruit inspection service providers, infor mation and logistic service providers, phytosanitary requirements as well as unfavorable weather conditions Although some of these are procedure s to be performed under any channel the overall management of these variables can (and 3 Interview with South African grower/exporter, January 25, 2013, Western Cape, South Africa. 4 Interview with South African citrus research association representative, January 21, 2013, Western Cape, South Africa.
102 does) vary by supply c hannel. During one interview, it was explained that the prioritization of these variables is influential in determining how successful one is at trading. In this researc h on direct sourcing, a variety of these distributions that all cater to some form of exclusion of traditional exporting players was investigated This includes a retailer owning their own importing component while also performing bagging, storage, and shipment out of their depots (Channel 3), fruit managed through third party importers p rocuring fruit directly from growers and exporters (Channel 4), or bypassing both the South African exporter and UK category manager (Channel 5). The Industry Responds In gaining an und erstanding of our research question, interviews were exporters. The questionnaires built upon concepts proposed in the NIE literature more specifically concentrating on uncertainty, frequency of transactions, opportunism, bounded rationality, and trust within the industry. Additional questions, more closely aligned with the functions currently performed w ithin the industry. A large scale far mer operating through both, non direct (eight years) and direct channels (six years) spoke cautiously of the evolving channels. Over the years, he has found merit in operating under both channels. There are exceptions to the direct channel where I have found some programs with supermarkets that pay very well. Twenty years ago, over 50% of the final price would go back to the farmer, now only 18 25% gets
103 returned. I will only deal with the direct channel if the retailer will keep my brand, creating a loyal following of my product 5 Financial security, payments, logistics, and management of product were seen as determining factors when deciding to operate through a direct channel. Credit checks are not needed with large r etailers as they will always follow through with the payment while working through the direct channel. The lack of credit checks was viewed Frequency of operating an d contract length was not seen as significant when deciding on which export channel to operate under, while bargaining leverage worked in favor for non direct channels. You get less bargaining leverage when working with the direct channel. With direct, t hey will start you off with higher prices and then their bargaining becomes more fierce. They attract you with the higher prices and then become more competitive Local (pure) exporters were still seen as an integral to the industry as they are still bein g used in some direct exports to European retailers, such as Tesco. The following farmer saw balancing produce between the two channels as key for long term successful returns: To avoid one channel completely would be detrimental to the grower. We will ha ve no negotiation power over supermarkets where an alternative crop directly ASDA, Tesco, Marks and Spence r, Morrisons and Sainsbury are some of the largest retailers in the UK, jointly accounting for over 80% of retail food sales in the country (Farfan, 2013). Sainsbury and Tesco have added further expecta tions to South African producers. 5 Inte rview with South African grow er/exporter, January 25, 2013, Western Cape, South Africa.
104 Traditionally, South African exporters have sent their grade 1 exports (highest quality fruit in terms of size, color, and sweetness among other conditions) to northern European countries while lower quality (grade 2 and 3) is sent to parts of the Middle East, Eastern Europe, and Russia. In their increasing pressure to enhance industry performance, Sainsbury and Tesco have now come to expect a grade 1+ from many South African producers without paying any additional premiu m price. 6 While some industry players are willing to abide to these extra conditions, some have looked elsewhere for better returns on their product. ASDA was a very transparent company when first operating in South Africa; however, became more restrictiv e on returns when International Produce Limited (IPL) became the exclusive importer of fruit into ASDA. This farmer has since looked to other retailers, such as Sainsbury and Marks & Spencer, who pay up to 40% more for higher quality. 7 IPL was d escribed a s very strict on its saw a strong future for the pure South African exporter due to the direct model being limited to part of the UK and Europe. Fruit moves where the money is. operate under the direct model He conti nued to point out that the (specialized ) exporter is the one who creates new markets and new opportunities. It is because of these newer markets around the world that other re tailers will always be dependent upon an exporter who creates these relationships. While also diversifying his crop between direct and indirect distribution channels, he noted that the most advantageous points in trading directly are financial 6 Interview with South African citrus research association representative, January 21, 2013, Western Cape, South Africa. 7 Interview with South African grower/exporter, January 2 5, 2013, Western Cape, South Africa.
105 security, lo gistics, managem ent of product, and information. In summary, he explained the direct model as a safe mode with a guaranteed payment; however, you do miss the some seasons Exporters have adopted different strategies to protect themselves within the market. Some of the l arger exporters have themselves vertically integrated and purchased their own farms for guaranteed b usiness. More importantly, as one exporter explained, he does not stand in the way of producers, who previously operated through his company but now want to trade directly with retailers. Instead of separating themselves from these farmers completely, they (exporting company) look towards an even more hybrid distribution where the farmers are allowed to trade directly; however, still able to use pa rts of their logistical model 8 When questioned, he summarized his strategy, Owning your own production unit and shi pping contract were two proposed methods to avoiding an uncertain future. This was contrary to a smaller pure exporter interviewed: our own farms. There is conflict in this. Wh o is shipping what? If we ship with our own fruit, other farmers might think we are taking preference 9 They conclude that their strength in the market belongs in their current relationships they have with their farmers. Along with providing all of the shipping rebates back to the farmer, an additional 25% of end year profits are distributed back to 8 Interview with South African grower/exporter, January 23, 2013, Western Cape, South Africa. 9 Interview with South African exporter, January 24, 2013, Western Cape, South Africa.
10 6 the grower. They too saw a long term future for their type of role within the trade channel. They saw their major advantage as having the ability to attain fruit from various sources, as well as assisting the grower in selling off their entire crop. While most retailers are concerned about the exact origin during the procurement of their fruit, they do compromise when faced with shortages. A grower wanting to trade directly would also face difficulties in selling off his entire crop due to the different sizes and color specifications. Through the traditional exporter, farmers are able to connect with various markets that have different specification requir ements. Other exporters were less willing to give their perspective on the matter by merely claiming, channel dominating over another 10 A large grower, packer, and exporter of citru s spoke of his perceived drivers as to why he felt retailers were in preference of trading di rectly. These drivers included t o increase financial margins within the chain ; c ontinue to convince shareholders of their safe and cost effective practices in eth ical trade ; h elp manage their risk of supply and opportunistic behavior 11 He further empathized his last point in describing the opportunistic behavior associated with growers when retailers are in short supply of produce. Here, through an event such as unfavorable weather to a particularly large grower, other suppliers of the 10 Interview with South African grower/ exporter, January 24, 2013, Western Cape, South Africa. 11 Interview with South African grower/exporter, January 24, 2013, Western Cape, South Africa.
107 This grower/exporter of citrus, operating under numerous different channels, believes the role of trad itional exporter will change in the near future. For long term longevity, along with the skills and passion to stay in the business, exporters would need to secure their own production source. He saw traditional exporters operating approximately 20/30% of fruit in the future and involving only niche varieties from small/medium enterprises (SMEs). Traditional exporters will also be further pushed from smaller farmers demanding fixed prices on their product. While he admitted to not getting necessarily bett er returns from trading direct, doing so does provide his business with a larger say in his product. Individuals from research institutions provided further feedback on the evolving export market. 12 13 Although there is no official number recorded, they con cluded exporters have most likely increased since the introduction of the direct channel. This goes against popular belief that this number has decreased; however, this is due to the consolidation of farmers forming their own export brand. This number is estimated to be larger than the number of pure exporters leaving the market. Financial security, when deciding on which channel to operate in, was explained as a double edged sword. Retailers such as Tesco and ASDA do not provide any financing options to growers. This limits smaller farmers, who are typically dependent upon production loans, to operate with traditional exporters who have access to financing options; moreover, this limitation can create a scenario where returns to this creating a locked in obligation due to these lagged payments. Some scenarios were 12 Interview with South African citrus research association representative, January 23, 2013, Western Cape, South Africa. 13 Interview with South African agricultural economist, January 23, 2013, Western Cape, South Africa.
108 also explained where, simply unable to function further on account of financial strain, farms have been bought out by their exporter. Regardless, this situation is less of an occurrence due to the continued consolidation of smaller growers. Concluding Statements Based on literature proposed by NIE, we interviewed farmers, exporters, and South African citrus research organizations. We discussed the evolving citrus export channel, including opinions and perspectives, with personnel from academic institutions, quality control, new variety branding companies, and other organizations. Most specifically, we look ed at uncertainty, frequency of transactions, opportunism, bounded rationality, and trust as the fundamental drivers in the development of our questions. Some of these drivers were found to be more pressing and influential during the questionnaire processe s. Through our interviews, it became increasingly clear these drivers have added to our understanding of why the South African citrus industry is experiencing vertical integration. Producers viewed uncertainty, our first driver in explaining vertical inte gration, as a defining element when deciding between direct or indirect channels. Uncertainty of payments was of less concern when dealing with reputable (direct) retailers, as these were seen as guaranteed payments after fruit was approved. Exporters saw uncertainty of supply as a strong driver for retailers to import directly. In supplying direct, retailers -without having to allocate from other, less desirable, markets. Res earchers viewed traditional exporters as a safer option for smaller farmers with uncertain financial options, as smaller farmers are able to finance their crops ahead of the season. Such a process is not possible in the direct model. Grower/exporters also widely recommended that
109 traditional exporters should safeguard their means of operation through the procurement of their own farms. Frequency of transactions, along with trust, were explained as the two largest drivers in explaining why farmers chose to o perate directly, and why traditional local exporters are still able to sustain themselves through the evolving export industry. Pure exporters explained their relationships with farmers, brought upon by frequent transactions, as elements that are consider ed in the highest regard. In building these relationships, farmers are not requesting fixed prices from their exporters, consequently trusting they will receive a fair market price. For large retailers, opportunism exists during short supply of produce. I n such a scenario, it is explained that exporters are able to manipulate market prices in their own interest. In avoiding such occurrences, retailers have imported directly, or even further, secured their own production source by purchasing citrus groves. Although not unanimous, most of the industry viewed IPL as the preferred option in supplying directly. This preference was not on account of the prices they offer, but Wal some UK retailers reject fruit not meeting certain shelf specifications, IPL were found more willing to negotiate lower prices for this less desirable product; furthermor e, IPL assumes all risk of the product after a certain number of days post inspection. Beyond price, management of the product is a critical determining factor when understanding the preferences of interview respondents. As one respondent noted,
110 Mar t has a certain number of days to inspect, the n the risk passes to them. In most Nevertheless, farmers were still hesitant to export their entire crop directly. It was recommended no more than 40 50% of their c rop should be allocated towards trading directly, therefore diversifying their risk between a stable market price (direct), and varying returns markets (indirectly). These varying returns market are essential for the farmer to receive some returns for low er quality fruit not accepted at higher end UK retailers, as well to receive premium prices for similar quality fruit received in UK stores (typically receiving market price). As increased consolidation occurs within the industry, trading directly becomes more of a financially stable solution. This means of exporting is still very new, and more aligned to methods used within the UK and parts of Europe. Traditional exporters have remained innovative in acquiring new markets -including those in Asia and th e Middle East. Such markets are becoming more desirable for many producers, as these markets also tradition ally favor trading indirectly with fewer product quality constraints. An uncertain future remains. While this holds true for any business or entrep reneur, opinions stand firm that there is room for both channels to operate. Producers have quickly ada pted to both channels, diversifying their crop and in turn, their risk, between the direct and indirect channels. Direct marketing has become accustomed to higher end retailers. In practice, the logistical procedure of diversifying an entire crop of different sizes and quality, that might not all be suitable for these high struggle under an all encompassing direct model industry; however, they will remain as
111 active through emerging African and Middle Eastern markets which continue to operate under traditional domestic importers. In conclusion, Tables 4 1 and 4 2 summarize these findings by presenting the estimated impact of each driver on the newly reformed institutional arrangement from producer and retailer perspectives respectively. A final observation some researchers believe retailers are pushing ethical trade boundar a proposed new addition involves retailers attaining all shipment rebates. Retailers have been able to acquire these shipping rebates through direct trade. IPL is seen as an a nomaly as most retailers still acquire the majority of their produce through traditional exporters. With traditional exporters, some researchers and exporters argue retailers will demand, even though exported through unaffiliated exporters with their own shipping contracts, these shipments should be awarded to them -in an attempt to generate greater rebates and larger margins within the chain. This remain s to be seen.
112 Type of Respondent Classification of Respondent Number of Interviews Exporter In direct 0 Direct 1 Both (Indirect and Direct) 4 Grower SME 2 Cooperative 1 Large 2 Research Academic 2 Industry 2 Total: 11 14 Figure 4 1. Summary of i nterviews 14 Eleven individual interviews. Some respondents fell under more than one category
113 Figure 4 2 Five major supply c hannels (Source: FPEF Advanced Manual, 2010 Adapted )
114 Table 4 1. Summary of producer drivers Producer Drivers Impact on Formation of Direct Channel High/Medium/Low Negative/Positive Uncertainty Security of Payment H P Competitive Price M N Finance O ptions M N Frequency of Transactions Management of Product M P Logistics M P Bargaining Leverage M N Rebates Received H N Bounded R ationality Guaranteed Market M P Price Fluctuations M N Trust H P Table 4 2 S ummary of retailer drivers Retailer Drivers Impact on Formation of Direct Channel High/Medium/Low Negative/Positive Uncertainty Supply H P Shareholders M P Frequency of Transactions H P Bounded R ationality Financial Security H P Opportunism M P
115 CHAPTER 5 CONCLUSIONS The governance structure of the South African citrus industry, and the implications of its transformation from a regulated to deregulated export body have been primarily researched following its ini tial undertaking. Most of this literature is attributed to Mather and includes his influential paper of Market Liberalisation in Post Apartheid H is approach in reviewing the citrus in dustry is not from traditional neo classical techniques as used in has received limited research in terms of its export performance and roles of its institutions, in pa rticular, how they have transformed the industry. In response, we identified key dimensions using both institutional and neoclassical approaches to explain economic growth and development, and to advance this body of literature. The objective of this rese arch is to analyze dimensions of industrial organization and export performance within the industry. For this, we adopt the framework s of structure conduct performance and new institutional economics to qualitatively explain the transformation of a pre reg ulated to, now, post deregulated export industry, and to explain the evolving governance structure within the citrus industry. Quantitative methods are used in determining influential export attributes and the efficiency of its allocating fruit to major im port destinations. Chapter 2 builds upon literature in analyzing and comparing the determinants of the South African citrus export market organization, behavior, and resulting success. Export channels, barriers to trade, labor and land issues, as well as g overnment intervention are some of the many dimensions explored within this study. Although the
116 South African government was the instrumental figure transforming the structure of the industry, it currently operates at a minimalist level -including fruit in spection and land ownership rights. It was through this limited assistance that the industry responded by implementing, and pursuing, a statutory levy for each exported box of citrus. This levy has continued to function as the primary financial source of f unding for research and development. Additionally, the consequences of apartheid still remain evident in large parts of the country. Most of the country continues to live in poverty, while unemployment has remained high, exceeding 30%. The citrus industry 's role has become increasingly vital in terms of employment for lower income black Africans. Neighboring the political unrest and financial constraint economies of Zimbabwe and Mozambique, natives of these countries have settled in South Africa seeking em ployment mostly through the agricultural sector. Employment opportunities are limited in an industry where smaller farmers have continued consolidating with one another. Such consolidation has resulted in an over 50% decline in the number of commercial far ms during the past 20 years. These commercial farms are historically reliant on seasonal workers. Phytosanitary conditions and overlapping production seasons have added to the limitations of citrus exports. In their struggle with the citrus black spot dise ase, only a minority of the citrus industry is able to export to the financially lucrative U.S market. In response, South African exporters continue exploring alternative trade routes to the Middle East and Asian markets. As described in Chapter 3 import demand models were estimated in determining significant predictors of citrus exports. In these models, price and importing countries'
117 GDP were identified as the most influential predictors in estimating import quantities of oranges, mandarins, lemons, and grapefruit during the deregulated export era. The effect of the countries' GDP became less influential when regressed with the country specific GDP deflator. All models accounted for both the domestic currency of the importing country, as well as the curre ncy exchanged during their specific trade agreement. Although this approach was proven to be mostly insignificant, it was achieved through the separation of a U.S dollar and European euro model. For the U.S dollar transactions, exchange rate becomes a pote ntially relevant factor for countries trading in a currency different from what is used domestically. The European euro transaction presented a different constraint. Here, the luxury of operating under one unified currency only becomes available from the y ear 2002. Previously, all European countries operated under their own, domestic currency. In this consideration, an indicator variable was included to determine if any deviation in trade of the different citrus varieties was evident from the pre euro to eu ro transaction period. This variable also proved to be mostly insignificant. In evaluating the efficiency of its export, a quantity constrained programming model was developed for the comparison of current export prices and quantities, to a theoretical op timal revenue maximizing allocation. For this procedure, each industry variety was estimated at operating within seven to eleven percent of optimality during the years of 2008, 2009, and 2010. These optimal revenues often included instances of severe price discrimination unlikely to suffice in a real world competitive market. The new institutional economics analysis of the South African citrus industry in Chapter 4 discusses the key governance changes witnessed within the recent export
118 chain reform. Here, s ome retailers have vertically integrated by procuring their own farms, or by acquiring the rights of additional logistic procedures. Interviews with local farmers and exporters were conducted for a market response as to why some retailers have adopted a mo re direct approach, and the potential consequences it has on the industry. Uncertainty, frequency of transactions, opportunism, bounded rationality, and trust were identified as the fundamental drivers in developing these questions. While some were less in fluential than others, all the above drivers were interpreted as influential components in explaining the act of vertical integration. Farmers were also confident of a secure trade future for exporters that operated in both direct and indirect trade busine ss models. Consolidated larger farmers, however, recommended that, for the preservation of a sustainable operational business future, traditional exporters excluded from the direct model should invest in the procurement of their own farms. Additionally, fa rmers have looked to diversify their risk by allocating their produce between both export models -the more financially secure, but potentially lower returning, direct model and the traditional variable returns indirect model. Such a result was explained a s a highly likely long term business model for most large scale farmers. Limitations of th e Research In the two fold research objective of estimating import demand models and the optimal trus export markets were identified. Literature encouraged the inclusion of price, GDP, and exchange rate as strong potential predictors of import quantities. While extra consideration was given to missing observations within the time series period, as wel l as the currency exchanges during a transaction, quality and size information of the fruit was inaccessible. This information is relevant in understanding the inherent nature of South African citrus
119 exports and its markets. Higher quality fruit is mostly sent to Western Europe and U.S markets, medium quality to the Middle East and Asia, while lower quality fruit is traditionally sent to Russia and Eastern European markets. In the construction of the programming model, the allocation model produced optimal prices and quantities in producing a maximized revenue for each citrus variety. This was obtained through the inverting and linearization of the import demand equations. It was necessary to linearize the import demand equations as the solver used to solve the allocation model has demonstrated in the past that it cannot deal with log linear equations (see Spreen, et al., 2003). In Chapter 4 the implications and consequences of vertically integration of retailers is explored. Th rough the assistance of empiri cal research, interviews were established with leading producers and exporters within the industry. In an attempt to validate our findings, exporters representing the majority of the industry, in terms of quantities, were interviewed. While it would be fav orable to comment on their exact participation, this information was private to members of the Fresh Produce Export Forum. Future Area of Research Quantity and price data attained from the World Trade Atlas assisted in the estimation of import demand mode Consequently, the efficiency of the industry is calculated by comparing actual export to theoretical optimal quantities for all significant citrus export markets. A future extension to this work would be to further estimate the performance over an extended period of time including the de regulated industry period. Such research will allow for an efficiency comparison between a deregulated and regulated citrus export period,
120 separated by citrus variety and ex port destination. Current available data restricted the study to a deregulated period study. Attempts to empirically estimate the increased controlled percentage of citrus transactions by retailers operating within the direct model proved to be unsuccess ful. Exporters were mostly unwilling to share transaction spreadsheets, as doing so could potentially compromise their relationship with the retailer, or their competitive position within the market. It was due to such difficulties associated with obtainin g this information that a qualitative approach was chosen to explore this research question. If attainable in the future, an empirical approach in estimating the market share captured by retailers adopting a direct trading channel would be most beneficial to the at present, limited literature.
121 APPENDIX A REGRESSION R CODE BelgiumGrapefruit < read.table("BelgiumGrapefruit.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary" ,"March","April","May","June","August","September","Oct"," November","Dec")) BelgiumGrapefruit$QuantityMMT = (BelgiumGrapefruit$Quantity/1000) BelgiumGrapefruit$PricePerMMT = (BelgiumGrapefruit$Price*1000) BelgiumGrapefruit$Log_Quantity = log(BelgiumGrapefruit$QuantityMMT) BelgiumGrapefruit$Log_Price = log(Belgi umGrapefruit$PricePerMMT) BelgiumGrapefruit$Log_ExRate = log(BelgiumGrapefruit$ExRate) BelgiumGrapefruit$Log_GDP = log(BelgiumGrapefruit$GDP) glm.linear < glm(BelgiumGrapefruit$Log_Quantity ~ BelgiumGrapefruit$Log_Price + BelgiumGrapefruit$Log_ExRate + B elgiumGrapefruit$Log_GDP + BelgiumGrapefruit$January + BelgiumGrapefruit$Febuary + BelgiumGrapefruit$March + BelgiumGrapefruit$April + BelgiumGrapefruit$June + BelgiumGrapefruit$August + BelgiumGrapefruit$September + BelgiumGrapefruit$Oct + BelgiumGrapefr uit$November + BelgiumGrapefruit$Dec) summary(glm.linear) BelgiumLemon < read.table("BelgiumLemon.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","Mar ch","April","May","July","August","September","Oct","November","Dec")) BelgiumLemon$QuantityMMT = (BelgiumLemon$Quantity/1000) BelgiumLemon$PricePerMMT = (BelgiumLemon$Price*1000) BelgiumLemon$Log_Quantity = log(BelgiumLemon$QuantityMMT) BelgiumLemon$Log_Price = log(BelgiumLemon$PricePerMMT) BelgiumLemon$Log_ExRate = log(Be lgiumLemon$ExRate) BelgiumLemon$Log_GDP = log(BelgiumLemon$GDP) glm.linear < glm(BelgiumLemon$Log_Quantity ~ BelgiumLemon$Log_Price + BelgiumLemon$Log_ExRate + BelgiumLemon$Log_GDP + BelgiumLemon$January + BelgiumLemon$Febuary +BelgiumLemon$March+ Belgiu mLemon$April + BelgiumLemon$May + BelgiumLemon$July +BelgiumLemon$August + BelgiumLemon$September + BelgiumLemon$Oct + BelgiumLemon$November+ BelgiumLemon$Dec)
122 summary(glm.linear) BelgiumMandarin < read.table("BelgiumMandarin.txt",col.names=c("Date", "Quantity","Price","ExRate","GDP","January","Febuary"," March","April","June","July","August","September","Oct","November","Dec")) BelgiumMandarin$QuantityMMT = (BelgiumMandarin$Quantity/1000) BelgiumMandarin$PricePerMMT = (BelgiumMandarin$Price*1000) Bel giumMandarin$Log_Quantity = log(BelgiumMandarin$QuantityMMT) BelgiumMandarin$Log_Price = log(BelgiumMandarin$PricePerMMT) BelgiumMandarin$Log_ExRate = log(BelgiumMandarin$ExRate) BelgiumMandarin$Log_GDP = log(BelgiumMandarin$GDP) glm.linear < glm(Belgium Mandarin$Log_Quantity ~ BelgiumMandarin$Log_Price + BelgiumMandarin$Log_ExRate + BelgiumMandarin$Log_GDP + BelgiumMandarin$January + BelgiumMandarin$Febuary + BelgiumMandarin$March + BelgiumMandarin$April + BelgiumMandarin$June + BelgiumMandarin$July + Be lgiumMandarin$August + BelgiumMandarin$September + BelgiumMandarin$Oct + BelgiumMandarin$November + BelgiumMandarin$Dec) summary(glm.linear) BelgiumOrange < read.table("BelgiumOrange.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January"," Febuary","Ma rch","April","May","June","August","September","Oct","November","Dec")) BelgiumOrange$QuantityMMT = (BelgiumOrange$Quantity/1000) BelgiumOrange$PricePerMMT = (BelgiumOrange$Price*1000) BelgiumOrange$Log_Quantity = log(BelgiumOrange$QuantityMM T) BelgiumOrange$Log_Price = log(BelgiumOrange$PricePerMMT) BelgiumOrange$Log_ExRate = log(BelgiumOrange$ExRate) BelgiumOrange$Log_GDP = log(BelgiumOrange$GDP) glm.linear < glm(BelgiumOrange$Log_Quantity ~ BelgiumOrange$Log_Price + BelgiumOrange$Log_ExRa te + BelgiumOrange$Log_GDP + BelgiumOrange$January + BelgiumOrange$Febuary + BelgiumOrange$March + BelgiumOrange$April + BelgiumOrange$May + BelgiumOrange$June + BelgiumOrange$August + BelgiumOrange$September + BelgiumOrange$Oct + BelgiumOrange$November + BelgiumOrange$Dec)
123 summary(glm.linear) ChinaOranges < read.table("ChinaOranges.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","Mar ch","April","May","June","July","September","Oct","November","Dec")) ChinaOranges$Quanti tyMMT = (ChinaOranges$Quantity/1000) ChinaOranges$PricePerMMT = (ChinaOranges$Price*1000) ChinaOranges$Log_Quantity = log(ChinaOranges$QuantityMMT) ChinaOranges$Log_Price = log(ChinaOranges$PricePerMMT) ChinaOranges$Log_ExRate = log(ChinaOranges$ExRate) C hinaOranges$Log_GDP = log(ChinaOranges$GDP) glm.linear < glm(ChinaOranges$Log_Quantity ~ ChinaOranges$Log_Price + ChinaOranges$Log_ExRate + ChinaOranges$Log_GDP + ChinaOranges$January + ChinaOranges$Febuary + ChinaOranges$March + ChinaOranges$April + Chi naOranges$May + ChinaOranges$June + ChinaOranges$July + ChinaOranges$September + ChinaOranges$Oct + ChinaOranges$November + ChinaOranges$Dec) summary(glm.linear) GermanyGrapefruit < read.table("GermanyGrapefruit.txt",col.names=c("Date","Quantity","Pri ce","ExRate","GDP","January","Febuary" ,"March","April","June","July","August","September","Oct","November","Dec")) GermanyGrapefruit$QuantityMMT = (GermanyGrapefruit$Quantity/1000) GermanyGrapefruit$PricePerMMT = (GermanyGrapefruit$Price*1000) GermanyGra pefruit$Log_Quantity = log(GermanyGrapefruit$QuantityMMT) GermanyGrapefruit$Log_Price = log(GermanyGrapefruit$PricePerMMT) GermanyGrapefruit$Log_ExRate = log(GermanyGrapefruit$ExRate) GermanyGrapefruit$Log_GDP = log(GermanyGrapefruit$GDP) glm.linear < gl m(GermanyGrapefruit$Log_Quantity ~ GermanyGrapefruit$Log_Price + GermanyGrapefruit$Log_ExRate + GermanyGrapefruit$Log_GDP + GermanyGrapefruit$January + GermanyGrapefruit$Febuary + GermanyGrapefruit$March + GermanyGrapefruit$April + GermanyGrapefruit$June + GermanyGrapefruit$July + GermanyGrapefruit$August + GermanyGrapefruit$September + GermanyGrapefruit$Oct + GermanyGrapefruit$November + GermanyGrapefruit$Dec)
124 summary(glm.linear) GermanyLemon < read.table("GermanyLemon.txt",col.names=c("Date","Quantit y","Price","ExRate","GDP","January","Febuary","Mar ch","April","May","June","August","September","Oct","November","Dec")) GermanyLemon$QuantityMMT = (GermanyLemon$Quantity/1000) GermanyLemon$PricePerMMT = (GermanyLemon$Price*1000) GermanyLemon$Log_Quantit y = log(GermanyLemon$QuantityMMT) GermanyLemon$Log_Price = log(GermanyLemon$PricePerMMT) GermanyLemon$Log_ExRate = log(GermanyLemon$ExRate) GermanyLemon$Log_GDP = log(GermanyLemon$GDP) glm.linear < glm(GermanyLemon$Log_Quantity ~ GermanyLemon$Log_Price + GermanyLemon$Log_ExRate + GermanyLemon$Log_GDP + GermanyLemon$January + GermanyLemon$Febuary + GermanyLemon$March + GermanyLemon$April + GermanyLemon$May + GermanyLemon$June + GermanyLemon$August + GermanyLemon$September + GermanyLemon$Oct + GermanyLemon $November + GermanyLemon$Dec) summary(glm.linear) GermanyOranges < read.table("GermanyOranges.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","M arch","April","May","June","August","September","Oct","November","Dec")) Germ anyOranges$QuantityMMT = (GermanyOranges$Quantity/1000) GermanyOranges$PricePerMMT = (GermanyOranges$Price*1000) GermanyOranges$Log_Quantity = log(GermanyOranges$QuantityMMT) GermanyOranges$Log_Price = log(GermanyOranges$PricePerMMT) GermanyOranges$Log_Ex Rate = log(GermanyOranges$ExRate) GermanyOranges$Log_GDP = log(GermanyOranges$GDP) glm.linear < glm(GermanyOranges$Log_Quantity ~ GermanyOranges$Log_Price + GermanyOranges$Log_ExRate + GermanyOranges$Log_GDP + GermanyOranges$January + GermanyOranges$Febu ary + GermanyOranges$March + GermanyOranges$April + GermanyOranges$May + GermanyOranges$June + GermanyOranges$August + GermanyOranges$September + GermanyOranges$Oct + GermanyOranges$November + GermanyOranges$Dec)
125 summary(glm.linear) JapanGrapefruit < r ead.table("JapanGrapefruit.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary"," March","April","May","July","August","September","Oct","November","Dec")) JapanGrapefruit$QuantityMMT = (JapanGrapefruit$Quantity/1000) JapanGrapefru it$PricePerMMT = (JapanGrapefruit$Price*1000) JapanGrapefruit$Log_Quantity = log(JapanGrapefruit$QuantityMMT) JapanGrapefruit$Log_Price = log(JapanGrapefruit$PricePerMMT) JapanGrapefruit$Log_ExRate = log(JapanGrapefruit$ExRate) JapanGrapefruit$Log_GDP = l og(JapanGrapefruit$GDP) glm.linear < glm(JapanGrapefruit$Log_Quantity ~ JapanGrapefruit$Log_Price + JapanGrapefruit$Log_ExRate + JapanGrapefruit$Log_GDP + JapanGrapefruit$January + JapanGrapefruit$Febuary + JapanGrapefruit$March + JapanGrapefruit$April + JapanGrapefruit$May + JapanGrapefruit$July + JapanGrapefruit$August + JapanGrapefruit$September + JapanGrapefruit$Oct + JapanGrapefruit$November + JapanGrapefruit$Dec) summary(glm.linear) JapanLemon < read.table("JapanLemon_fix.txt",col.names=c("Date" ,"Quantity","Price","ExRate","GDP","January","Febuary","M arch","April","May","July","August","September","Oct","November","Dec")) JapanLemon$QuantityMMT = (JapanLemon$Quantity/1000) JapanLemon$PricePerMMT = (JapanLemon$Price*1000) JapanLemon$Log_Quantity = log(JapanLemon$QuantityMMT) JapanLemon$Log_Price = log(JapanLemon$PricePerMMT) JapanLemon$Log_ExRate = log(JapanLemon$ExRate) JapanLemon$Log_GDP = log(JapanLemon$GDP) glm.linear < glm(JapanLemon$Log_Quantity ~ JapanLemon$Log_Price + JapanLemon$Log_ExR ate + JapanLemon$Log_GDP + JapanLemon$January + JapanLemon$Febuary + JapanLemon$March + JapanLemon$April + JapanLemon$May + JapanLemon$July + JapanLemon$August + JapanLemon$September + JapanLemon$Oct + JapanLemon$November + JapanLemon$Dec) summary(glm.lin ear)
126 JapanOranges < read.table("JapanOranges.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","Mar ch","April","May","June","August","September","Oct","November","Dec")) JapanOranges$QuantityMMT = (JapanOranges$Quantity/1000 ) JapanOranges$PricePerMMT = (JapanOranges$Price*1000) JapanOranges$Log_Quantity = log(JapanOranges$QuantityMMT) JapanOranges$Log_Price = log(JapanOranges$PricePerMMT) JapanOranges$Log_ExRate = log(JapanOranges$ExRate) JapanOranges$Log_GDP = log(JapanOran ges$GDP) glm.linear < glm(JapanOranges$Log_Quantity ~ JapanOranges$Log_Price + JapanOranges$Log_ExRate + JapanOranges$Log_GDP + JapanOranges$January + JapanOranges$Febuary + JapanOranges$March + JapanOranges$April + JapanOranges$May + JapanOranges$June + JapanOranges$August + JapanOranges$September + JapanOranges$Oct + JapanOranges$November + JapanOranges$Dec) summary(glm.linear) NetherlandsGrapefruit < read.table("NetherlandsGrapefruit.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","Januar y","Febu ary","March","April","June","July","August","September","Oct","November","Dec")) NetherlandsGrapefruit$QuantityMMT = (NetherlandsGrapefruit$Quantity/1000) NetherlandsGrapefruit$PricePerMMT = (NetherlandsGrapefruit$Price*1000) NetherlandsGrapefrui t$Log_Quantity = log(NetherlandsGrapefruit$QuantityMMT) NetherlandsGrapefruit$Log_Price = log(NetherlandsGrapefruit$PricePerMMT) NetherlandsGrapefruit$Log_ExRate = log(NetherlandsGrapefruit$ExRate) NetherlandsGrapefruit$Log_GDP = log(NetherlandsGrapefruit$ GDP) glm.linear < glm(NetherlandsGrapefruit$Log_Quantity ~ NetherlandsGrapefruit$Log_Price + NetherlandsGrapefruit$Log_ExRate + NetherlandsGrapefruit$Log_GDP + NetherlandsGrapefruit$January + NetherlandsGrapefruit$Febuary + NetherlandsGrapefruit$March + NetherlandsGrapefruit$April + NetherlandsGrapefruit$June + NetherlandsGrapefruit$July + NetherlandsGrapefruit$August + NetherlandsGrapefruit$September + NetherlandsGrapefruit$Oct + NetherlandsGrapefruit$November + NetherlandsGrapefruit$Dec)
127 summary(glm. linear) NetherlandsLemon < read.table("NetherlandsLemon.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary", "March","April","May","June","August","September","Oct","November","Dec")) NetherlandsLemon$QuantityMMT = (Netherlands Lemon$Quantity/1000) NetherlandsLemon$PricePerMMT = (NetherlandsLemon$Price*1000) NetherlandsLemon$Log_Quantity = log(NetherlandsLemon$QuantityMMT) NetherlandsLemon$Log_Price = log(NetherlandsLemon$PricePerMMT) NetherlandsLemon$Log_ExRate = log(Netherland sLemon$ExRate) NetherlandsLemon$Log_GDP = log(NetherlandsLemon$GDP) glm.linear < glm(NetherlandsLemon$Log_Quantity ~ NetherlandsLemon$Log_Price + NetherlandsLemon$Log_ExRate + NetherlandsLemon$Log_GDP + NetherlandsLemon$January + NetherlandsLemon$Febuary + NetherlandsLemon$March + NetherlandsLemon$April + NetherlandsLemon$May + NetherlandsLemon$June + NetherlandsLemon$August + NetherlandsLemon$September + NetherlandsLemon$Oct + NetherlandsLemon$November + NetherlandsLemon$Dec) summary(glm.linear) Neth erlandsMandarin < read.table("NetherlandsMandarin.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuar y","March","April","May","July","August","September","Oct","November","Dec")) NetherlandsMandarin$QuantityMMT = (NetherlandsManda rin$Quantity/1000) NetherlandsMandarin$PricePerMMT = (NetherlandsMandarin$Price*1000) NetherlandsMandarin$Log_Quantity = log(NetherlandsMandarin$QuantityMMT) NetherlandsMandarin$Log_Price = log(NetherlandsMandarin$PricePerMMT) NetherlandsMandarin$Log_ExRa te = log(NetherlandsMandarin$ExRate) NetherlandsMandarin$Log_GDP = log(NetherlandsMandarin$GDP) glm.linear < glm(NetherlandsMandarin$Log_Quantity ~ NetherlandsMandarin$Log_Price + NetherlandsMandarin$Log_ExRate + NetherlandsMandarin$Log_GDP + Netherlands Mandarin$January + NetherlandsMandarin$Febuary + NetherlandsMandarin$March + NetherlandsMandarin$April + NetherlandsMandarin$May + NetherlandsMandarin$July + NetherlandsMandarin$August + NetherlandsMandarin$September + NetherlandsMandarin$Oct + Netherlan dsMandarin$November + NetherlandsMandarin$Dec)
128 summary(glm.linear) NetherlandsOrange < read.table("NetherlandsOrange.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary" ,"March","April","May","June","August","September","Oct"," November","Dec")) NetherlandsOrange$QuantityMMT = (NetherlandsOrange$Quantity/1000) NetherlandsOrange$PricePerMMT = (NetherlandsOrange$Price*1000) NetherlandsOrange$Log_Quantity = log(NetherlandsOrange$QuantityMMT) NetherlandsOrange$Log_Price = log(Nethe rlandsOrange$PricePerMMT) NetherlandsOrange$Log_ExRate = log(NetherlandsOrange$ExRate) NetherlandsOrange$Log_GDP = log(NetherlandsOrange$GDP) glm.linear < glm(NetherlandsOrange$Log_Quantity ~ NetherlandsOrange$Log_Price + NetherlandsOrange$Log_ExRate + N etherlandsOrange$Log_GDP + NetherlandsOrange$January + NetherlandsOrange$Febuary + NetherlandsOrange$March + NetherlandsOrange$April + NetherlandsOrange$May + NetherlandsOrange$June + NetherlandsOrange$August + NetherlandsOrange$September + NetherlandsOr ange$Oct + NetherlandsOrange$November + NetherlandsOrange$Dec) summary(glm.linear) RussiaGrapefruit < read.table("RussiaGrapefruit.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary", "March","April","June","July","August","Sep tember","Oct","November","Dec")) RussiaGrapefruit$QuantityMMT = (RussiaGrapefruit$Quantity/1000) RussiaGrapefruit$PricePerMMT = (RussiaGrapefruit$Price*1000) RussiaGrapefruit$Log_Quantity = log(RussiaGrapefruit$QuantityMMT) RussiaGrapefruit$Log_Price = l og(RussiaGrapefruit$PricePerMMT) RussiaGrapefruit$Log_ExRate = log(RussiaGrapefruit$ExRate) RussiaGrapefruit$Log_GDP = log(RussiaGrapefruit$GDP) glm.linear < glm(RussiaGrapefruit$Log_Quantity ~ RussiaGrapefruit$Log_Price + RussiaGrapefruit$Log_ExRate + RussiaGrapefruit$Log_GDP + RussiaGrapefruit$January + RussiaGrapefruit$Febuary +RussiaGrapefruit$March+ RussiaGrapefruit$April + RussiaGrapefruit$June + RussiaGrapefruit$July +RussiaGrapefruit$August + RussiaGrapefruit$September + RussiaGrapefruit$Oct + Ru ssiaGrapefruit$November+ RussiaGrapefruit$Dec)
129 summary(glm.linear) RussiaLemon < read.table("RussiaLemon.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","Marc h","May","June","July","August","September","Oct","November"," Dec")) RussiaLemon$QuantityMMT = (RussiaLemon$Quantity/1000) RussiaLemon$PricePerMMT = (RussiaLemon$Price*1000) RussiaLemon$Log_Quantity = log(RussiaLemon$QuantityMMT) RussiaLemon$Log_Price = log(RussiaLemon$PricePerMMT) RussiaLemon$Log_ExRate = log(Russ iaLemon$ExRate) RussiaLemon$Log_GDP = log(RussiaLemon$GDP) glm.linear < glm(RussiaLemon$Log_Quantity ~ RussiaLemon$Log_Price + RussiaLemon$Log_ExRate + RussiaLemon$Log_GDP + RussiaLemon$January + RussiaLemon$Febuary + RussiaLemon$March + RussiaLemon$May + RussiaLemon$June + RussiaLemon$July + RussiaLemon$August + RussiaLemon$September + RussiaLemon$Oct + RussiaLemon$November + RussiaLemon$Dec) summary(glm.linear) RussiaMandarin < read.table("RussiaMandarins.txt",col.names=c("Date","Quantity","Price", "ExRate","GDP","January","Febuary"," March","April","June","July","August","September","Oct","November","Dec")) RussiaMandarin$QuantityMMT = (RussiaMandarin$Quantity/1000) RussiaMandarin$PricePerMMT = (RussiaMandarin$Price*1000) RussiaMandarin$Log_Quantit y = log(RussiaMandarin$QuantityMMT) RussiaMandarin$Log_Price = log(RussiaMandarin$PricePerMMT) RussiaMandarin$Log_ExRate = log(RussiaMandarin$ExRate) RussiaMandarin$Log_GDP = log(RussiaMandarin$GDP) glm.linear < glm(RussiaMandarin$Log_Quantity ~ RussiaMa ndarin$Log_Price + RussiaMandarin$Log_ExRate + RussiaMandarin$Log_GDP + RussiaMandarin$January + RussiaMandarin$Febuary + RussiaMandarin$March + RussiaMandarin$April + RussiaMandarin$June + RussiaMandarin$July + RussiaMandarin$August + RussiaMandarin$Sept ember + RussiaMandarin$Oct + RussiaMandarin$November + RussiaMandarin$Dec)
130 summary(glm.linear) RussiaOranges < read.table("RussiaOranges.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","Ma rch","April","May","June","July"," August","September","November","Dec")) RussiaOranges$QuantityMMT = (RussiaOranges$Quantity/1000) RussiaOranges$PricePerMMT = (RussiaOranges$Price*1000) RussiaOranges$Log_Quantity = log(RussiaOranges$QuantityMMT) RussiaOranges$Log_Price = log(RussiaOrange s$PricePerMMT) RussiaOranges$Log_ExRate = log(RussiaOranges$ExRate) RussiaOranges$Log_GDP = log(RussiaOranges$GDP) glm.linear < glm(RussiaOranges$Log_Quantity ~ RussiaOranges$Log_Price + RussiaOranges$Log_ExRate + RussiaOranges$Log_GDP + RussiaOranges$Ja nuary + RussiaOranges$Febuary + RussiaOranges$March + RussiaOranges$April + RussiaOranges$May + RussiaOranges$June + RussiaOranges$July + RussiaOranges$August + RussiaOranges$September + RussiaOranges$November + RussiaOranges$Dec) summary(glm.linear) SA Lemon < read.table("SALemon.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","March"," April","June","July","August","September","Oct","November","Dec")) SALemon$QuantityMMT = (SALemon$Quantity/1000) SALemon$PricePerMMT = (SAL emon$Price*1000) SALemon$Log_Quantity = log(SALemon$QuantityMMT) SALemon$Log_Price = log(SALemon$PricePerMMT) SALemon$Log_ExRate = log(SALemon$ExRate) SALemon$Log_GDP = log(SALemon$GDP) glm.linear < glm(SALemon$Log_Quantity ~ SALemon$Log_Price + SALemon $Log_ExRate + SALemon$Log_GDP + SALemon$January + SALemon$Febuary + SALemon$March + SALemon$April + SALemon$June + SALemon$July + SALemon$August + SALemon$September + SALemon$Oct + SALemon$November + SALemon$Dec) summary(glm.linear)
131 SAMandarin < read.t able("SAMandarins.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","Marc h","April","May","July","August","September","Oct","November","Dec")) SAMandarin$QuantityMMT = (SAMandarin$Quantity/1000) SAMandarin$PricePerMMT = (SAMand arin$Price*1000) SAMandarin$Log_Quantity = log(SAMandarin$QuantityMMT) SAMandarin$Log_Price = log(SAMandarin$PricePerMMT) SAMandarin$Log_ExRate = log(SAMandarin$ExRate) SAMandarin$Log_GDP = log(SAMandarin$GDP) glm.linear < glm(SAMandarin$Log_Quantity ~ SAMandarin$Log_Price + SAMandarin$Log_ExRate + SAMandarin$Log_GDP + SAMandarin$January + SAMandarin$Febuary + SAMandarin$March + SAMandarin$April + SAMandarin$May + SAMandarin$July + SAMandarin$August + SAMandarin$September + SAMandarin$Oct + SAMandarin$No vember + SAMandarin$Dec) summary(glm.linear) SAOrange < read.table("SAOranges.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","March" ,"April","May","June","August","September","Oct","November","Dec")) SAOrange$QuantityMMT = (SAOrange$Quantity/1000) SAOrange$PricePerMMT = (SAOrange$Price*1000) SAOrange$Log_Quantity = log(SAOrange$QuantityMMT) SAOrange$Log_Price = log(SAOrange$PricePerMMT) SAOrange$Log_ExRate = log(SAOrange$ExRate) SAOrange$Log_GDP = log(SAOrange$GDP) glm. linear < glm(SAOrange$Log_Quantity ~ SAOrange$Log_Price + SAOrange$Log_ExRate + SAOrange$Log_GDP + SAOrange$January + SAOrange$Febuary + SAOrange$March + SAOrange$April + SAOrange$May + SAOrange$June + SAOrange$August + SAOrange$September + SAOrange$Oct + SAOrange$November + SAOrange$Dec) summary(glm.linear)
132 UKGrapefruit < read.table("UKGrapefruit.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","Mar ch","April","June","July","August","September","Oct","November","Dec")) U KGrapefruit$QuantityMMT = (UKGrapefruit$Quantity/1000) UKGrapefruit$PricePerMMT = (UKGrapefruit$Price*1000) UKGrapefruit$Log_Quantity = log(UKGrapefruit$QuantityMMT) UKGrapefruit$Log_Price = log(UKGrapefruit$PricePerMMT) UKGrapefruit$Log_ExRate = log(UKGr apefruit$ExRate) UKGrapefruit$Log_GDP = log(UKGrapefruit$GDP) glm.linear < glm(UKGrapefruit$Log_Quantity ~ UKGrapefruit$Log_Price + UKGrapefruit$Log_ExRate + UKGrapefruit$Log_GDP + UKGrapefruit$January + UKGrapefruit$Febuary + UKGrapefruit$March + UKGrap efruit$April + UKGrapefruit$June + UKGrapefruit$July + UKGrapefruit$August + UKGrapefruit$September + UKGrapefruit$Oct + UKGrapefruit$November+ UKGrapefruit$Dec) summary(glm.linear) UKLemon < read.table("UKLemon.txt",col.names=c("Date","Quantity","Pric e","ExRate","GDP","January","Febuary","March"," April","May","June","August","September","Oct","November","Dec")) UKLemon$QuantityMMT = (UKLemon$Quantity/1000) UKLemon$PricePerMMT = (UKLemon$Price*1000) UKLemon$Log_Quantity = log(UKLemon$QuantityMMT) UKLe mon$Log_Price = log(UKLemon$PricePerMMT) UKLemon$Log_ExRate = log(UKLemon$ExRate) UKLemon$Log_GDP = log(UKLemon$GDP) glm.linear < glm(UKLemon$Log_Quantity ~ UKLemon$Log_Price + UKLemon$Log_ExRate + UKLemon$Log_GDP + UKLemon$January + UKLemon$Febuary + UK Lemon$March + UKLemon$April + UKLemon$May + UKLemon$June + UKLemon$August + UKLemon$September + UKLemon$Oct + UKLemon$November + UKLemon$Dec) summary(glm.linear)
133 UKMandarin < read.table("UKMandarins.txt",col.names=c("Date","Quantity","Price","ExRate"," GDP","January","Febuary","Marc h","April","May","July","August","September","Oct","November","Dec")) UKMandarin$QuantityMMT = (UKMandarin$Quantity/1000) UKMandarin$PricePerMMT = (UKMandarin$Price*1000) UKMandarin$Log_Quantity = log(UKMandarin$QuantityMMT) UKMandarin$Log_Price = log(UKMandarin$PricePerMMT) UKMandarin$Log_ExRate = log(UKMandarin$ExRate) UKMandarin$Log_GDP = log(UKMandarin$GDP) glm.linear < glm(UKMandarin$Log_Quantity ~ UKMandarin$Log_Price + UKMandarin$Log_ExRate + UKMandarin$Log_GDP + UKM andarin$January + UKMandarin$Febuary + UKMandarin$March + UKMandarin$April + UKMandarin$May + UKMandarin$July + UKMandarin$August + UKMandarin$September + UKMandarin$Oct + UKMandarin$November + UKMandarin$Dec) summary(glm.linear) UKOrange < read.table ("UKOrange.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","March", "April","May","June","July","August","Oct","November","Dec")) UKOrange$QuantityMMT = (UKOrange$Quantity/1000) UKOrange$PricePerMMT = (UKOrange$Price*1000) UK Orange$Log_Quantity = log(UKOrange$QuantityMMT) UKOrange$Log_Price = log(UKOrange$PricePerMMT) UKOrange$Log_ExRate = log(UKOrange$ExRate) UKOrange$Log_GDP = log(UKOrange$GDP) glm.linear < glm(UKOrange$Log_Quantity ~ UKOrange$Log_Price + UKOrange$Log_ExRa te + UKOrange$Log_GDP + UKOrange$January + UKOrange$Febuary + UKOrange$March + UKOrange$April + UKOrange$May + UKOrange$June + UKOrange$July + UKOrange$August + UKOrange$Oct + UKOrange$November + UKOrange$Dec) summary(glm.linear)
134 USMandarin < read.tabl e("USMandarins.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","Marc h","April","May","July","August","September","Oct","November","Dec")) USMandarin$QuantityMMT = (USMandarin$Quantity/1000) USMandarin$PricePerMMT = (USMandari n$Price*1000) USMandarin$Log_Quantity = log(USMandarin$QuantityMMT) USMandarin$Log_Price = log(USMandarin$PricePerMMT) USMandarin$Log_ExRate = log(USMandarin$ExRate) USMandarin$Log_GDP = log(USMandarin$GDP) glm.linear < glm(USMandarin$Log_Quantity ~ USM andarin$Log_Price + USMandarin$Log_ExRate + USMandarin$Log_GDP + USMandarin$January + USMandarin$Febuary + USMandarin$March + USMandarin$April + USMandarin$May + USMandarin$July + USMandarin$August + USMandarin$September + USMandarin$Oct + USMandarin$Novem ber + USMandarin$Dec) summary(glm.linear) USOrange < read.table("USOrange.txt",col.names=c("Date","Quantity","Price","ExRate","GDP","January","Febuary","March", "April","May","June","August","September","Oct","November","Dec")) USOrange$QuantityMMT = ( USOrange$Quantity/1000) USOrange$PricePerMMT = (USOrange$Price*1000) USOrange$Log_Quantity = log(USOrange$QuantityMMT) USOrange$Log_Price = log(USOrange$PricePerMMT) USOrange$Log_ExRate = log(USOrange$ExRate) USOrange$Log_GDP = log(USOrange$GDP) glm.line ar < glm(USOrange$Log_Quantity ~ USOrange$Log_Price + USOrange$Log_ExRate + USOrange$Log_GDP + USOrange$January + USOrange$Febuary + USOrange$March + USOrange$April + USOrange$May + USOrange$June + USOrange$August + USOrange$September + USOrange$Oct + USO range$November + USOrange$Dec) summary(glm.linear)
135 APPENDIX B OPTIMAL ALLOCATION GAMS CODE $Title Estimated optimal exports for South African citrus Sets i importing countries / Belgium, Germany, Japan, Netherlands, Russia, Saudi, UK / ; Parameter s grapefruit2010(i) grapefruit import totals for 2010 / Belgium 1265023.00 Germany 4129947.00 Japan 45553107.00 Netherlands 50097788.00 Russia 17424100.00 Saudi 683387.00 UK 11217808.00 / grapefruit2009(i) grapefruit import totals for 2009 / Belgium 10744810.00 Germany 6819982.00 Japan 91833620.00 Netherlands 72932189.00 Russia 19508884.00 Saudi 1041110.00 UK 14103542.00 / grapefruit2008(i) grapefruit import totals for 2008 / Belgium 7511318.00 Germany 2610031.00 Japan 59579724.00 Netherlands 47791271.00 Russia 14070373.00 Saudi 1212345.00 UK 18775880.00 / p_grapefruit2010(i) average weighted price for grapefruit import totals for 2010 in US dollars / Belgium 0.537809573 Germany 0.446343885 Japan 0.551161032 Netherlands 0.497525234 Russia 0.630419425 Saudi 0.608962725 UK 0.478491363 / p_grapefruit2009(i) averag e weighted price for grapefruit import totals for 2009 in US dollars / Belgium 0.438349317 Germany 0.312955561 Japan 0.25217941 Netherlands 0.290811223
136 Russia 0.375335353 Saudi 0.465858612 UK 0.289277933 / p_grapefruit2008(i) average weighted price for grapef ruit import totals for 2008 in US dollars / Belgium 0.572241395 Germany 0.516200662 Japan 0.46307999 Netherlands 0.47399891 R ussia 0.655669533 Saudi 0.528522615 UK 0.304095503 / elas_grapefruit(i) elasticity for grapefruit import calculated from import demand esti mations / Belgium 0.3748 Germany 1.9193 Japan 1.3291 Netherlands 0.8822 Russia 0.03544 Saudi 0. 4446 UK 1.5647 /; Parameter beta_grapefruit2010(i) OLS 2010 beta calculation from average price average quantity and elasticities ; beta_grapefruit2010(i) = p_grapefruit 2010(i)/(grapefruit2010(i)*elas_grapefruit(i)) ; Parameter beta_grapefruit2009(i) OLS 2009 beta calculation from average price average quantity and elasticities ; beta_grapefruit2009(i) = p_grapefruit2009(i)/(grapefruit2009(i)*elas_grapefrui t(i)) ; Parameter beta_grapefruit2008(i) OLS 2008 beta calculation from average price average quantity and elasticities ; beta_grapefruit2008(i) = p_grapefruit2008(i)/(grapefruit2008(i)*elas_grapefruit(i)) ; Parameter alpha_grapefruit2010(i ) OLS 2010 alpha calculation from average price average quantity and elasticities ; alpha_grapefruit2010(i) = p_grapefruit2010(i) (beta_grapefruit2010(i)*grapefruit2010(i)) ; Parameter alpha_grapefruit2009(i) OLS 2009 alpha calculation fr om average price average quantity and elasticities ; alpha_grapefruit2009(i) = p_grapefruit2009(i) (beta_grapefruit2009(i)*grapefruit2009(i)) ; Parameter alpha_grapefruit2008(i) OLS 2008 alpha calculation from average price average quantit y and elasticities ;
137 alpha_grapefruit2008(i) = p_grapefruit2008(i) (beta_grapefruit2008(i)*grapefruit2008(i)) ; display beta_grapefruit2010 display beta_grapefruit2009 display beta_grapefruit2008 display alpha_grapefruit2010 display alp ha_grapefruit2009 display alpha_grapefruit2008 Variables q2010(i) optimal quantities shipped to country i z2010 total revenue in dollars ; Positive Variable q2010 ; Equations rev_grapefruit2010 define objective function demand2010 demand constraint ; rev_grapefruit2010 .. z2010 =e= sum(i, q2010(i)*alpha_grapefruit2010(i) + ((q2010(i)*q2010(i))*beta_grapefruit2010(i))) ; demand2010 .. sum(i, q2010(i)) =l= sum(i, grapefruit2 010(i)) ; Model citrus2010 /all/ ; Solve citrus2010 using nlp maximizing z2010 ; Variables q2009(i) optimal quantities shipped to country i z2009 total revenue in dollars ; Positive Variable q2009 ; Equations rev_grapefruit2009 define objective function demand2009 demand constraint ; rev_grapefruit2009 .. z2009 =e= sum(i, q2009(i)*alpha_grapefruit2009(i) + ((q2009(i)*q2009(i))*beta_grapefruit2009(i))) ; demand2009 .. s um(i, q2009(i)) =l= sum(i, grapefruit2009(i)) ; Model citrus2009 /all/ ; Solve citrus2009 using nlp maximizing z2009 ; Variables q2008(i) optimal quantities shipped to country i z2008 total revenue in dollars ; Positive Vari able q2008 ;
138 Equations rev_grapefruit2008 define objective function demand2008 demand constraint ; rev_grapefruit2008 .. z2008 =e= sum(i, q2008(i)*alpha_grapefruit2008(i) + ((q2008(i)*q2008(i))*beta_grapefruit2 008(i))) ; demand2008 .. sum(i, q2008(i)) =l= sum(i, grapefruit2008(i)) ; Model citrus2008 /all/ ; Solve citrus2008 using nlp maximizing z2008 ; $Title Estimated optimal exports for South African citrus Sets i importing countries / Belgium, Germany, Japan, Netherlands, Russia, Saudi, UK / ; Parameters lemon2010(i) lemon import totals for 2010 / Belgium 88588.00 Germany 1961093.00 Japan 1264869.00 Netherlands 21186153.00 Russia 21896464.00 Saudi 19055626.00 UK 17102555.00 / lemon2009(i) lemon import totals for 2009 / Belgium 4060449.00 Germany 1754724.00 Japan 905246.00 Netherlands 13855271.00 Russia 28030722.00 Saudi 64323616.00 UK 15888758.00 / lemon2008(i) lemon import totals for 2008 / Belgium 6208750.00 Germany 2152251.00 Japan 7576286.00 Netherlands 27072934.00 Russia 8794456.00 Saudi 18559375.00 UK 23600886.00 /
139 p_lemon2010(i) average weighted price for lemon import totals for 2010 in US dollars / Belgium 0.4 Germany 0.490162379 Japan 0.899121419 Netherland s 0.818329372 Russia 0.649858007 Saudi 0.697307954 UK 0.683598655 / p_lemon2009(i) average weighted price for lemon impo rt totals for 2009 in US dollars / Belgium 0.811986866 Germany 0.475643611 Japan 0.610864538 Netherlands 0.555433239 Russia 0.235 626855 Saudi 0.162246131 UK 0.554298784 / p_lemon2008(i) average weighted price for lemon import totals for 2008 in US dollars / Belgium 0.53873201 Germany 0.505552856 Japan 0.208082799 Netherlands 0.556224012 Russia 0.456759017 Saudi 0.379351546 UK 0.618338734 / elas_lemon(i) elasticity for lemon import calculated from import demand estimations / Belgium 1.3275 Germany 1.2583 Japan 0.7117 Netherlands 0.8189 Russia 0.300113 Saudi 0.686 UK 1.6386 /; Parameter beta_lemon2010(i) OLS 2010 beta calculation from average price average quantity and elasticities ;
140 beta_lemon2010(i) = p_lemon2010(i)/(lemon2010(i)*elas_lemon(i)) ; Parameter beta_lemon2009(i) OLS 2009 beta calculation from average price average quantity and elasticities ; beta_lemon2009(i) = p_lemon2009(i)/(lemon2009(i)*elas_lemon(i)) ; Parameter beta_lemon2008(i) OLS 2008 beta calculation from average price average quantity and elasticities ; beta_lemon2008(i) = p_lemon2008(i)/(lemon2008(i)*elas_lemon(i)) ; Parameter alpha_lemon2010(i) OLS 2010 alpha calculation from average price average quantity and elasticities ; alpha_lemon2010(i) = p_lemon2010(i) (beta_lemon2010(i)*lemo n2010(i)) ; Parameter alpha_lemon2009(i) OLS 2009 alpha calculation from average price average quantity and elasticities ; alpha_lemon2009(i) = p_lemon2009(i) (beta_lemon2009(i)*lemon2009(i)) ; Parameter alpha_lemon2008(i) OLS 2008 alpha calculation from average price average quantity and elasticities ; alpha_lemon2008(i) = p_lemon2008(i) (beta_lemon2008(i)*lemon2008(i)) ; display beta_lemon2010 display beta_lemon2009 display beta_lemon2008 display alpha_lemon2010 disp lay alpha_lemon2009 display alpha_lemon2008 Variables q2010(i) optimal quantities shipped to country i z2010 total revenue in dollars ; Positive Variable q2010 ; Equations rev_lemon2010 define objective fun ction demand2010 demand constraint ; rev_lemon2010 .. z2010 =e= sum(i, q2010(i)*alpha_lemon2010(i) + ((q2010(i)*q2010(i))*beta_lemon2010(i))) ; demand2010 .. sum(i, q2010(i)) =l= sum(i, lemon2010(i)) ; Model citru s2010 /all/ ;
141 Solve citrus2010 using nlp maximizing z2010 ; Variables q2009(i) optimal quantities shipped to country i z2009 total revenue in dollars ; Positive Variable q2009 ; Equations rev_lemon2009 d efine objective function demand2009 demand constraint ; rev_lemon2009 .. z2009 =e= sum(i, q2009(i)*alpha_lemon2009(i) + ((q2009(i)*q2009(i))*beta_lemon2009(i))) ; demand2009 .. sum(i, q2009(i)) =l= sum(i, lemon2009(i )) ; Model citrus2009 /all/ ; Solve citrus2009 using nlp maximizing z2009 ; Variables q2008(i) optimal quantities shipped to country i z2008 total revenue in dollars ; Positive Variable q2008 ; Equations rev_ lemon2008 define objective function demand2008 demand constraint ; rev_lemon2008 .. z2008 =e= sum(i, q2008(i)*alpha_lemon2008(i) + ((q2008(i)*q2008(i))*beta_lemon2008(i))) ; demand2008 .. sum(i, q2008(i)) =l= sum(i, lemon2008(i)) ; Model citrus2008 /all/ ; Solve citrus2008 using nlp maximizing z2008 ; $Title Estimated optimal exports for South African citrus Sets i importing countries / Belgium, Netherlands, Russia, Saudi, UK, USA / ; Para meters mandarin2010(i) mandarin import totals for 2010 / Belgium 834072.00 Netherlands 18931923.00 Russia 12136007.00 Saudi 2569670.00 UK 45266409.00 USA 8448289.00
142 / mandarin2009(i) mandarin import totals for 2009 / Belgium 2735706.00 Netherla nds 14798256.00 Russia 8430244.00 Saudi 1407836.00 UK 55584542.00 USA 6426866.00 / mandarin2008(i) mandarin import totals for 2008 / Belgium 3282333.00 Netherlands 17718933.00 Russia 13028176.00 Saudi 2791476.00 UK 48206990.00 USA 4369877.00 / p_mandarin2010(i) average weighted price for mandarin import totals for 2010 in US dollars / Belg ium 1.097199378 Netherlands 0.732320516 Russia 0.657396606 Saudi 0.763403776 UK 0.702251451 USA 0.88 9386013 / p_mandarin2009(i) average weighted price for mandarin import totals for 2009 in US dollars / Belgium 0.767621393 Netherlands 0.715855849 Russia 0.774703716 Saudi 0.776744948 UK 0.496867571 USA 0.903022484 / p_mandarin2008(i) average weighted price for mandarin import totals for 2008 in US dollars / Belgium 0.682164579 Netherlands 0.662461873 Russia 0.484075049 Saudi 0.723422104 UK 0.596768977
143 USA 1.070793569 / elas_mandarin(i) elasticity for mandarin import calculated from import demand estimations / Be lgium 0.788582 Netherlands 1.8174 Russia 0.88333 Saudi 0.7218 UK 0.94658 USA 0.2871 /; Parameter beta_mandarin2010(i) OLS 2010 beta calculation from average price average quantity and elasticities ; beta_mandarin2010(i) = p_mandarin2010(i)/(mandarin2010(i)*elas_mandarin(i)) ; Parameter beta_mandar in2009(i) OLS 2009 beta calculation from average price average quantity and elasticities ; beta_mandarin2009(i) = p_mandarin2009(i)/(mandarin2009(i)*elas_mandarin(i)) ; Parameter beta_mandarin2008(i) OLS 2008 beta calculation from average p rice average quantity and elasticities ; beta_mandarin2008(i) = p_mandarin2008(i)/(mandarin2008(i)*elas_mandarin(i)) ; Parameter alpha_mandarin2010(i) OLS 2010 alpha calculation from average price average quantity and elasticities ; alpha_mandarin2010(i) = p_mandarin2010(i) (beta_mandarin2010(i)*mandarin2010(i)) ; Parameter alpha_mandarin2009(i) OLS 2009 alpha calculation from average price average quantity and elasticities ; alpha_mandarin2009(i) = p_mandarin200 9(i) (beta_mandarin2009(i)*mandarin2009(i)) ; Parameter alpha_mandarin2008(i) OLS 2008 alpha calculation from average price average quantity and elasticities ; alpha_mandarin2008(i) = p_mandarin2008(i) (beta_mandarin2008(i)*mandarin2008( i)) ; display beta_mandarin2010 display beta_mandarin2009 display beta_mandarin2008 display alpha_mandarin2010 display alpha_mandarin2009
144 display alpha_mandarin2008 Variables q2010(i) optimal quantities shipped to country i z2010 total revenue in dollars ; Positive Variable q2010 ; Equations rev_mandarin2010 define objective function demand2010 demand constraint ; rev_mandarin2010 .. z2010 =e= sum(i, q2010(i)*alpha_mandarin 2010(i) + ((q2010(i)*q2010(i))*beta_mandarin2010(i))) ; demand2010 .. sum(i, q2010(i)) =l= sum(i, mandarin2010(i)) ; Model citrus2010 /all/ ; Solve citrus2010 using nlp maximizing z2010 ; Variables q2009(i) optimal quantities sh ipped to country i z2009 total revenue in dollars ; Positive Variable q2009 ; Equations rev_mandarin2009 define objective function demand2009 demand constraint ; rev_mandarin2009 .. z2009 =e= sum(i, q2009(i)*alpha_mandarin2009(i) + ((q2009(i)*q2009(i))*beta_mandarin2009(i))) ; demand2009 .. sum(i, q2009(i)) =l= sum(i, mandarin2009(i)) ; Model citrus2009 /all/ ; Solve citrus2009 using nlp maximizing z2009 ; Variables q2 008(i) optimal quantities shipped to country i z2008 total revenue in dollars ; Positive Variable q2008 ; Equations rev_mandarin2008 define objective function demand2008 demand constraint ; rev_mandarin2 008 .. z2008 =e= sum(i, q2008(i)*alpha_mandarin2008(i) + ((q2008(i)*q2008(i))*beta_mandarin2008(i))) ; demand2008 .. sum(i, q2008(i)) =l= sum(i, mandarin2008(i)) ; Model citrus2008 /all/ ; Solve citrus2008 using nlp maximizing z2008 ;
145 $Title Estimated optimal exports for South African citrus Sets i importing countries / Belgium, China, Japan, Netherlands, Russia, Saudi, UK, USA / ; Parameters orange2010(i) orange import totals for 2010 Germany excluded due to positive price / Belgium 5480690.00 China 7188700.00 Japan 7503709.00 Netherlands 197812595.00 Russia 138938744.00 Saudi 88914223.00 UK 69711369.00 USA 34812949.00 / orange2009(i) orange import totals for 2009 / Belgium 33626052.00 China 4962018.00 Japan 7192064.00 Netherlands 151982174.00 Russia 98489709.00 Saudi 72 204469.00 UK 71441426.00 USA 28053305.00 / orange2008(i) orange import totals for 2008 / Belgium 32166606. 00 China 4043483.00 Japan 7972729.00 Netherlands 189822518.00 Russia 102747480.00 Saudi 138307621.00 UK 82855572.00 USA 32645490.00 / p_orange2010(i) average weighted price for orange import totals for 2010 in US dollars / Belgiu m 0.601777709 China 0.597387622 Japan 0.623611579 Netherlands 0.518146117 Russia 0.546347578 Saudi 0.537952074
146 UK 0.501940061 USA 0.968968756 / p_orange2009(i) average weighted price for orange import totals for 2009 in US dollars / Belgium 0.464792634 China 0.477551968 Japan 0.444916533 Netherlands 0.418135105 Russia 0.465806628 Saudi 0.421828 561 UK 0.415083592 USA 0.895213288 / p_orange2008(i) average weighted price for orange import totals for 2008 in US dollars / Belgiu m 0.518126353 China 0.417206987 Japan 1.286038088 Netherlands 0.436718655 Russia 0.401413874 Saudi 0.166357358 UK 0.419524985 USA 0.880371113 / elas_orange(i) elasticity for orange import calculated from import demand estimations / Belgium 1.02 52 China 0.43 Japan 1.509077 Netherlands 2.32883 Russia 0.5985 Saudi 0.6036 UK 1.56018 USA 1.1701 /; Parameter beta_orange2010(i) OLS 2010 beta calculation from average price average quantity and elasticities ; beta_orange2010(i) = p_orange2010(i)/(orange2010(i)*elas_ orange(i)) ; Parameter beta_orange2009(i) OLS 2009 beta calculation from average price average quantity and elasticities ; beta_orange2009(i) = p_orange2009(i)/(orange2009(i)*elas_orange(i)) ;
147 Parameter beta_orange2008(i) OLS 2008 beta cal culation from average price average quantity and elasticities ; beta_orange2008(i) = p_orange2008(i)/(orange2008(i)*elas_orange(i)) ; Parameter alpha_orange2010(i) OLS 2010 alpha calculation from average price average quantity and elasticiti es ; alpha_orange2010(i) = p_orange2010(i) (beta_orange2010(i)*orange2010(i)) ; Parameter alpha_orange2009(i) OLS 2009 alpha calculation from average price average quantity and elasticities ; alpha_orange2009(i) = p_orange2009 (i) (beta_orange2009(i)*orange2009(i)) ; Parameter alpha_orange2008(i) OLS 2008 alpha calculation from average price average quantity and elasticities ; alpha_orange2008(i) = p_orange2008(i) (beta_orange2008(i)*orange2008(i)) ; display beta_orange2010 display beta_orange2009 display beta_orange2008 display alpha_orange2010 display alpha_orange2009 display alpha_orange2008 Variables q2010(i) optimal quantities shipped to country i z2010 total revenue in do llars ; Positive Variable q2010 ; Equations rev_orange2010 define objective function demand2010 demand constraint ; rev_orange2010 .. z2010 =e= sum(i, q2010(i)*alpha_orange2010(i) + ((q2010(i)*q2010(i))* beta_orange2010(i))) ; demand2010 .. sum(i, q2010(i)) =l= sum(i, orange2010(i)) ; Model citrus2010 /all/ ; Solve citrus2010 using nlp maximizing z2010 ; Variables q2009(i) optimal quantities shipped to country i z2009 total revenue in dollars ; Positive Variable q2009 ;
148 Equations rev_orange2009 define objective function demand2009 demand constraint ; rev_orange2009 .. z2009 =e= sum(i, q2009(i)*alpha_orange2009(i) + ((q2009(i)*q2009(i))*beta_orange2009(i))) ; demand2009 .. sum(i, q2009(i)) =l= sum(i, orange2009(i)) ; Model citrus2009 /all/ ; Solve citrus2009 using nlp maximizing z2009 ; Variables q2008(i) optimal quantities shipped to co untry i z2008 total revenue in dollars ; Positive Variable q2008 ; Equations rev_orange2008 define objective function demand2008 demand constraint ; rev_orange2008 .. z2008 =e= sum(i, q2008 (i)*alpha_orange2008(i) + ((q2008(i)*q2008(i))*beta_orange2008(i))) ; demand2008 .. sum(i, q2008(i)) =l= sum(i, orange2008(i)) ; Model citrus2008 /all/ ; Solve citrus2008 using nlp maximizing z2008 ;
149 APPENDIX C QUESTIONNAIRES Exporter Interv iew Number: Company Name : Respondent Name: Contact E mail: Contact Number: How many years has your company been in the citrus export industry? Approximately what volumes by type of citrus are exported through your company each year ? T o what degree has your export volumes changed by each type of citrus during the past 10 years ? What are your major exporting destinations by each citrus type? Do you feel price pressure or threats to market share from alternative fruits? When custome r demand or purchasing habits change, what difficulties are associated with exporting alternate varieties? Do costs of exporting citrus di f fer across types? If yes, how and why? How do you di f same ty pes ( e.g. Product Innovation ? Packaging ? Branding?) Rank the following (1 lowest risk, 7 highest risk) in terms of establishing your position within the market Advertising : Customer Loyalty Government Regulations : Research and Development : Negotiations with Suppliers/Farmer s Negotiations with Buyers :
150 Interview Number: Spoilage/liability with Lost Products Investment in Storage and T ransportatio n How many growers export through you ? How d oes this compare to other exporters? What costs, regulations, and risks are associated with establishing, or incorporating, growing activities within your own industr y ? Outline a typical export transaction: Ho w and to what extent, is information exchanged between yourself and the buyer? Which currencies are exchanged through a typical transaction? (Are your transactions completed though a Custome r Foreign Currency (CFC) Account ? What is a typical CIF o f fer per quantity of citrus type ? What is a typical FOB price ? When is final payment completed?) In the typical transaction just described, when is ownership of the product exchanged? (Consignmen t Shipments vs. Fixed Price Shipments) Outline how this transaction would di f fer under the direct channel:
151 Interview Number: Briefly explain how these market factors differ when a producer decides on a n export channel route (direct vs. t raditional)? 1 ) Contract Length: 2 ) Financial Security: 3 ) Payments: 4 ) Logistics: 5 ) Management of Product: 6 ) Information Exchange: 7 ) Degree of Dependency on Current Channel: 8 ) Bargaining Leverage 9 ) Proximity to the Supplier 10) Th e Frequency of Operations 1 1) T raceability and Certification 12 ) Other How do you and your trade partners collaborate in managing risks associated with: 1 ) Supply shortages 2 ) Quality at Arrival: 3 ) Late Payments; 4 ) Political Risks: 5 ) W ea ther: 6 ) Sanitary and Phytosanitary Controls: 7 ) Unscrupulous Action: 8 ) Other: Briefly explain how the management of these risks might di f fer if operating under the direct channel:
152 Interview Number: What is an approximate farm to port price spread for your citrus varieties? Approximately what percentage of costs are associated with 1 ) T ransport (port to vessel ) % 2 ) Cold Storage % 3 ) Landside Charges (PPECB Lev y Inspection Fees, T erminal Handling ) % 4 ) Do cumentation Charges (Phytosanitary Certificates ) % 5 ) Forward Charges (Courier Fee % 6 ) Insurance (Marine, Credit Guarantee, Bank Charges ) % 7 ) Rebate s % How do rebates di f fer in terms of final destination? How do y ou expect this spread to change if operating under the direct channel? In comparison from traditional to direct, which entities within the trade chain do you expect to : obtain greater financial gain or greater participation? obtain financial loss or less participation? Which entity (produce r packing, exporter), and to what extent, is most at risk (job security) with the How has your export strategy altered with the introduction of an alt ernative export channel? How do you think your export channel has adapted to changes brought upon by the alternate channel: Short term responses?
153 Interview Number: Anticipated long term responses? In your opinion, is the indust ry sustainable under both channels? When comparing pre and post direct trade, how have the following changed: 1 ) Financial returns 2 ) Export yields 3 ) Frequency of transactions 4 ) Number of producers 5 ) Number of packers 6 ) Number of expor ters 7 ) Market Access 8 ) Market Risk Final Comments? Do you know of anyone else who would be willing to assist in my research?
154 Producer Interview Number: Company Name Respondent Name: Contact E mail: Contact Number: Are you asso ciated with, or form part of, a coop?: yes no If yes, what is the name of the coop? Which citrus varieties are you producing? How much of your total production does each variety represent (by percentage)? Do any of these varieties repre sent a specificity or niche market? If so, what factors led to the decision to produce these niche varieties? Overall, have your citrus crop varieties changed in recent years? If so, why: How do you di f ferentiate your product from other exporters same varieties (e.g. Product Innovation?,Packaging?,Branding?) What is an approximate farm to port price spread for your citrus varieties? Rank the following (1 lowest risk, 7 highest risk) in terms of establishing your position within the market Advertising : Customer Loyalty : Government Regulations : Research and Development : Investment in Orchard s Certification and T raceability Issue s Negotiation with Buyers : How large is your labor fo rce, how many are considered full/part time? How many hours constitute full/part time?
155 Interview Number: How are workers e f forts tracked within field labor? What financial incentives are o f fered in terms of commission for field laborers? H ow many nurseries supply you? What costs, regulations, and risks are associated with establishing, or incorporating, nursery activities within your own farm ? Which channe l would you currently be associated with: Direct Channel T raditional Channel Both Have you previously been associated with, or conducted trade through, the alternate channel? How many years have you conducted trade through this channel: Outline a typical transaction: Ho w and to what extent, is information exchanged between yourself and the buyer? Which currencies are exchanged through a typical transaction? (Are your transactions completed though a Customer Foreign Currency (CFC) Account ? What is a typic al CIF o f fer per quantity of citrus type?, What is a typical FOB price, When is final payment completed?) In the typical transaction just described, when is ownership of the product exchanged? (Consignment Shipments vs. Fixed Price Shipments)
156 Interv iew Number: In your opinion, outline how transactions would di f fer under the two channels. How has your decision to operate though this channel been influenced by: 1 ) Contract Length: 2 ) Financial Security: 3 ) Payments: 4 ) Lo gistics: 5 ) Management of Product: 6 ) Information Exchange: 7 ) Degree of Dependency on Current Channel: 8 ) Bargaining Leverage 9 ) Proximity to the Supplier 10) The Frequency of Operations 1 1 ) T raceability and Certification 12 ) Other Briefly explain how these market factors di f fer if operating under the alternate channel: How do you and your trade partners collaborate in managing risks associated with: 1 ) Supply shortages: 2 ) Quality at Arrival:
157 Interview Number: 3 ) Late Payments; 4 ) Political Risks: 5 ) W eather: 6 ) Sanitary and Phytosanitary Controls: 7 ) Unscrupulous Action: 8 ) Other: Briefly explain how the management of these risks might di f fer if operating under the channel you currently do not associate with? Are large producers capable of creating their own trade channel? Is a particular (direct or traditional) channel more favorable to large scale producers? Do SMEs have the means to participate in the direct channel? Can fa rmers participate in both channels? Approximatel y what percentage of costs are associated with 1 ) T ransport (port to vessel ) % 2 ) Cold Storage % 3 ) Landside Charges (PPECB Lev y Inspection Fees, T erminal Handling ) % 4 ) Docume ntation Charges (Phytosanitary Certificates ) % 5 ) Forward Charges (Courier Fee ) % 6 ) Insurance (Marine, Credit Guarantee, Bank Charges ) % 7 ) Rebate s % How do rebates di f fer in terms of final destination? In comparis on from traditional to direct, which entities within the trade chain do you expect to : obtain greater financial gain or greater participation? obtain financial loss or less participation?
158 Interview Number: Which entity (produce r pack e r exporter), and to what extent, is most at risk (job security) with the How has your production strategy altered with the introduction of an direct export channel? How do you think your export ch annel has adapted to changes brought upon by the direct channel: Short term responses? Long term responses? In your opinion, ss the industry sustainable under both channels? When comparing pre and post direct trade, how have the fo llowing changed: 1 ) Financial returns 2 ) Export yields 3 ) Frequency of transactions 4 ) Number of producers 5 ) Number of packers 6 ) Number of exporters 7 ) Market Access 8 ) Market Risk
159 Interview Number: Final Comments? Do you know of anyone else who would be willing to assist in my research?
160 LIST OF REFERENCES Acemoglu, D., Johnson, S. and Mitton, T. "Determinants of Vertical Integration: Financial Development and Contracting Costs." The Journal of Finance LXIV.3 (2009): 1251 1290. Agricultural, Forestry and Fisheries Department of South Africa. "A Profile of the South African Citrus Market Value Chain." Agricultural, Forestry and Fisheries 2011. "A Profile of the South African Citrus Market Value Chain." Agricultural, Forestry and Fisheries 2010. "Abstract of Agricultural Statistics." Agricultural, Forestry and Fisheries 2012. Anderson, E. and Coughlan, A.T. "International market entry and expansion via independent or integrated channels of distribution." Journal of Marketing 51 (1987): 71 82. Anderson, E. and Schmittlein, D. "Integration of the sales forces: An empirical estimation." Rand Journal of Economics 15 (1984): 385 395. Arize, A.C. "Traditional Export Demand Relation and Parameter Instability." Journal of 28 .6 (2001): 378 396. Arize, A.C., Osang, T., and Slottje, D.J. "Exchange Rate Volatility and Foreign Trade: Evidence from Thirteen LDC's." Journal of Business & Economic Statistics 18.1 (2000): 10 17. Arrow, K.J. "The Organization of Economic Activity: Issu es Pertinent to the Choice of Market Versus Nonmarket Allocation." The Analysis and Evaluation of Public Expenditure: The PPB System 1 (1969): 3 9 73. Aviphant, P., Lee, J., and Seale, J.L. "Trade Agreements and U.S. Citrus Exports to Japan." Paper presente d at the joint annual meetings of the American and WesternAgricultural Economics Associations, Vancouver, British Columbia, Canada, 1990. Bahmani Oskooee, M. "Determination of International Trade Flows." Journal of Develop ment Economics 20 (1986): 107 123 Bailey, D.G., Dresser, G. and Arnold, J.M. "Grapefruit Medication Interactions: Forbidden Fruit or Avoidable Consequences?" Canadian Medical Association 185.4 (2013): 309 316. Baldwin, R., DiNino, V., De Santis, R., De Santis, R. A. and Taglioni, D. "Stu dy on the Impact of the Euro on Trade and Foreign Direct Investment." European Economy Economic Papers 321, Directorate General Economic and Monetary Affairs (DG ECFIN), European Commission. 2008: 141.
161 Binkley, J. "The Relationship Between Elasticity and Least Squares Bias." The Review of Economics and Statistics 63 (1981): 307 309. Aggregate Import Demand Equation Comparison of Three European Countries." Journal of Internationa l Economics 10 (1980): 561 566. Central Intelligence Agency. The World Factbook 2009. 29 January 2012.
162 Food and Agricultural Organization of the United Nations. FAOSTAT Rome: Statistics Division of the FAO, n.d.
163 International Trade Centre. Consumer Conscience: How Enviromenment and Ethics are Influential Exports Switzerland: ITC, 2009.
164 Magee, S.P. "Price, Incomes, and Foreign Trade." Kenen, P.B. International Trade and Finance Cambridge University Press, 1976. 175 252. Mather, C. and Greenberg, S. "Market Liberalisation in Post Apartheid South Africa: The Restructuring of Citr us Exports after 'Deregulation'." Journal of Southern African Studies 29.2 (2003): 393 412. Mather, C. Liberalisation n.d.
165 Savin, N.E. and White, K.J. "Testing for Autocorrelation with Missing Observations." Ec onometrica 46.1 (1978): 59 67. Scott, C., Lundgren, H. and Thompson, P. Guide to Supply Chain Management Springer, 2011. Shelanski, H.A. and Klein, P.G. "Empirical Research in Transaction Cost Economics: A Review and Assessment." Journal of Law, Economics & Organization 11.2 (1995): 335 361. South African Government News Agency. Govt hails improvement on land reform 2012.
166 WageIndicator Foundation. Minimum Wage Report South Africa 2012.
167 BIOGRAPHICAL SKETCH Jean Paul Baldwin received his Bachelor of Science degree in m athematics and s tatistics at the Nelson Mandela Metropolitan Universi ty in Port Elizabeth, South Africa in 2005. He received his Honors degree in s tatistics at the same university the following year. He furthered his academic career by obtaining a m aster s degree in s tatistics at the University of Florida in 2009. He rece ived his Ph.D. degree i n Food and Resource Economics from the Universit y of Florida in the summer of 2013. Jean Paul was awarded the a lumni g raduate f ellowship during his Ph.D. tenure, the highest graduate student award at the University of Florida. Jean export efficiency of the South African citrus industry. During this study, he had the opportunity to partake in fie ldwork within South Africa -including interviews with citrus production, packaging, rese arch and export participants. His fields of specialization include international trade, econometrics, applied microeconomics and agribusiness.