Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Marketing cooperatives : A model of the output decisions of the Cloverdale lettuce and vegetable cooperative Latham, Susie 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1993_spring_latham_susie.pdf [ 4.8MB ]
Metadata
JSON: 831-1.0086336.json
JSON-LD: 831-1.0086336-ld.json
RDF/XML (Pretty): 831-1.0086336-rdf.xml
RDF/JSON: 831-1.0086336-rdf.json
Turtle: 831-1.0086336-turtle.txt
N-Triples: 831-1.0086336-rdf-ntriples.txt
Original Record: 831-1.0086336-source.json
Full Text
831-1.0086336-fulltext.txt
Citation
831-1.0086336.ris

Full Text

MARKETING COOPERATIVES: A MODEL OF THE OUTPUT DECISIONSOF THE CLOVERDALE LETTUCE AND VEGETABLE COOPERATIVEbySUSIE LATHAMB.Sc.(Agr.), The University of British Columbia, 1990A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDY(Department of Agricultural Economics)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIADecember, 1992© Susie Latham, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature) Department of Ayir.ti14wi.A^corttnwtiess.The University of British ColumbiaVancouver, CanadaDate 9c:. q 3 t.C1CaQDE-6 (2/88)iiABSTRACTMarketing cooperatives play an important role in agricultural activities. Institutionalsupport for cooperatives is based on the idea that, collectively, farmers can achievebenefits than, individually would be difficult to obtain. Head lettuce in the Lower Mainlandregion of British Columbia is marketed and distributed by a central selling agency whichis organized as a producer cooperative. Members of the cooperative are subject toregulations, in the form of market quota allocations which control the quantity of headlettuce they can sell through their cooperative. This study describes and analyses themarket structure of the head lettuce industry in British Columbia to ascertain and quantifythe source of benefits to producers from cooperative marketing within a regulatedmarketing environment.A model of the industry is constructed to characterize the market for head lettucein B.C. The parameters which affect consumer demand and farm supply are estimatedwith econometric equations. A feature of supply is that current production decisions areinfluenced by the producer's market quota allocation which, in turn, is determined by theproducer's past sales. The market quotas are believed to have constrained supplyresponse and this is borne out by the empirical results which indicates a highly inelasticsupply curve. The demand for head lettuce is also estimated to be inelastic. This resultis not surprising since head lettuce is regarded as a basic commodity by consumers.The estimated supply and demand elasticities are used to derive linear supply anddemand curves at the cooperative and wholesale levels. These are used with theobserved 1990 price and quantity levels to calibrate a model of the B.C. head lettuceindustry. A counterfactual model is then formulated to simulate a market with no controlson output. Given an inelastic wholesale demand, the simulation results indicate that forvery small increases in cooperative output, large decreases in price occur. Consequentlytotal revenues decline at every alternative assumption of supply increase. This resultsupports the hypothesis that output restrictions by the cooperative have the potential toincrease members' output prices.It is concluded that while the market quotas have in the past provided positivebenefits to cooperative members, the quotas may now be hindering the process ofadjustment to the loss of tariff protection and changing market conditions by makingproducers less price responsive.ivTABLE OF CONTENTSABSTRACT ^  iiTABLE OF CONTENTS ^  ivLIST OF TABLES  viLIST OF FIGURES ^  viiACKNOWLEDGEMENTS  viii1.0 INTRODUCTION ^  11.1^Problem Statement  ^21.2^Objectives of the Study  ^31.3^Organization of the Study  ^42.0 OVERVIEW OF THE B.C. FRESH VEGETABLE INDUSTRY ^ 52.1^Description of Market Structure  ^52.2^Description of the Cloverdale Lettuce and Vegetable Cooperative ^ 103.0 COOPERATIVE PRINCIPLES AND ECONOMIC THEORY ^ 243.1^Cooperative Principles ^  243.2^Implications of the Cooperative Principles ^  263.3 Economic Theory of Cooperative Formation  313.4^Price and Output Decisions and Membership Policies ^ 413.4.1 Closed Membership ^  433.4.2 Open Membership  463.4.3 Discriminatory Policy ^  463.5^Basis^for^Formation^of^the^Cloverdale^Lettuce & VegetableCooperative 484.0^A MODEL OF THE B.C. LETTUCE INDUSTRY ^ 524.1^Price Determination in the B.C. Lettuce Market ^ 534.2^Supply Model for B.C. Lettuce ^ 564.2.1^Estimation of Supply Curve 604.2.2^Supply Estimation Data ^ 624.2.3^Supply Estimation Results 654.2.2^Validation of Supply Estimation ^ 704.3^Demand Model for Lettuce in B.0 724.3.1^Estimation of Demand Curve ^ 724.3.2^Demand Estimation Data 734.3.3^Demand Estimation Results ^ 754.3.4 Validation of Demand Model 795.0^ANALYSIS OF A COOPERATIVE WITH SUPPLY CONTROL  825.1^Derived Supply and Demand Elasticities ^ 825.2^Derived Supply and Demand Curves ^ 865.3^Model Simulation ^ 885.4^Summary and Conclusions ^ 98REFERENCES ^ 106APPENDIX 110LIST OF TABLESTable 1^Fresh field-grown crops regulated by the B.C. VegetableCommission  ^6Table 2^Tariff schedule for regulated fresh field-grownvegetables (1986) ^9Table 3^Cooperative charges, 1991 (per carton basis)  ^15Table 4^Econometric results of lettuce supply estimation  ^66Table 5^Econometric results of lettuce demand estimation  ^76Table 6^Demand elasticities for lettuce ^78Table 7^Change in total revenue under alternative supplycurve assumptions, Wholesale demand elasticity: -0.26 ^93Table 8^Change in total revenue under alternative supplycurve assumptions, Wholesale demand elasticity: -0.14 ^94Table 9^Change in total revenue when imports increase ^97viLIST OF FIGURESFigure 1^B.C. head lettuce production and prices ^13Figure 2^Comparison of weekly wholesale-to-retail prices,1985-90, Jun/Oct ^20Figure 3^Optimum behaviour for a marketing cooperative when nosingle price will cover costs ^40Figure 4^Equilibrium output positions of a cooperative underalternative membership policies ^45Figure 5^Output equilibrium position of a cooperative with openmembership policy ^47Figure 6^Short-run model of price determination in theB.C. lettuce market ^54Figure 7^Observed and predicted values from econometricestimation of supply  ^71Figure 8^Observed and predicted values from econometricestimation of demand ^80Figure 9^Relationship between real wholesale prices,real retail prices and margins  ^85Figure 10^Model of B.C. lettuce market, 1990  ^87Figure 11^Model of supply and demand under alternativesupply assumptions  ^92VI Iv•diACKNOWLEDGEMENTSI would like to thank the members of my thesis committee, Mary Bohman, JimVercammen and Nancy Ryan. In particular, I would like to thank Mary, my main thesisadvisor for her contribution and encouragement.Thanks are also due to the support staff in the Department whose assistance hasbeen much appreciated. Special thanks are due to my fellow graduate students whowere always willing to lend a sympathetic ear.Finally, I would like to thank my family for their support over the years.11.0^INTRODUCTIONMarketing cooperatives play an integral role in Canadian agriculture. Cooperativemarketing of farm output is seen as a way for farmers to address market shortcomingsand as a means of price enhancement. Three possible ways to raise farm price areapparent:1. If the cooperative can market the farm product at a lower cost than existing non-cooperative firms, the margin separating farm and retail price can be reduced.2. In agricultural markets where there are few buyers of the raw farm output,marketing sector firms may have market power over farmers. In this non-competitive setting farmers may be paid less than what the product is worth.3.^If a producer cooperative can raise the retail price of the products produced fromthe farm commodity, a higher farm price will result. This may be accomplishedby improving the quality of the product or by restricting the flow of farm productto the market.Any action that raises the farm price will raise farm income and make the farmer betteroff. Two additional potential benefits of cooperatives are the reduction of risk, and theprovision of service in markets where no private marketing or processing firm is inoperation.21.1^Problem StatementCooperatives are a common feature in the agricultural sector in both developedand developing countries the world over. Governments actively encourage and supportcooperatives through a number of policies including the enactment of enabling legislation,financial assistance and development assistance. Support for cooperatives rests largelyon their perceived ability to bring about improvements in areas such as increasing marketaccess and competition, lowering processing and marketing costs and utilizing economiesof scale.The Cloverdale Lettuce and Vegetable Cooperative is the producer organizationresponsible for marketing of lettuce and other vegetables produced in the Lower Mainlandregion of British Columbia. It is empowered by provincial legislation as the exclusiveselling agency for head lettuce, celery and cauliflower. All producers of these crops aremembers of the Cooperative and are subject to market quotas which limit the amount ofoutput producers may sell through the Cooperative. An added feature of the B.C. marketis that imports enter freely subject to a seasonal tariff during the peak period of localproduction. Landed import prices are consistently higher than the price for locally grownproducts, suggesting the Cooperative has limited price setting ability. The source andextent of benefits to member-producers from this form of cooperative marketing is thefocus of this study. An issue of particular interest is the role of the market quotas.31.2^Objectives of the StudyThe purpose of this paper is three-fold: the first is to discuss the theoretical basisfor the welfare-improving qualities attributed to cooperatives. The theory will focus onagricultural marketing cooperatives and will be developed by examining the marketconditions which give rise to the formation of cooperatives, the way in which cooperativesinfluence these conditions to the producers' advantage, cooperative objectives and theirimplication on pricing and output behaviour. The theory will be applied to the case of theCloverdale Lettuce and Vegetable Cooperative which operates under regulations whichset it up as the sole seller of domestic lettuce but in a market structure which allowslimited price setting ability. Since the primary motivation for farmers' participation in acooperative is to improve their well-being, the question of interest is, does the Cooperativeprovide a benefit to its members not obtainable otherwise.The second purpose of the paper is to quantify the benefits of membership in theCloverdale Lettuce and Vegetable Cooperative by developing a model of the B.C. lettuceindustry. It is hypothesized that the inelastic demand which lettuce growers face givesrise to the imposition of output restrictions by the Cooperative as a way of preventing thelarge price decrease that will accompany an increase in supply. Furthermore, it isconjectured that farm supply has been effectively constrained by the market quotas andthat supply would be more price responsive in the absence of these market quotas.Therefore, the third objective of this paper is to use the lettuce market model to simulatenew equilibrium price and quantity outcomes under alternative assumptions of more4elastic supply.1.3^Organization of the StudyChapter 2 provides an overview of the structure of the B.C. fresh vegetableindustry with an emphasis on the operations of the Cloverdale Lettuce and VegetableCooperative. Chapter 3 of the study begins with a definition of cooperatives and ageneral overview of cooperative organizations. This is followed by a discussion of theeconomic theory of cooperatives as developed within the literature. Chapter 4 discussesthe estimation of supply and demand equations for the B.C. lettuce industry. Lineareconometric equations are used to derive the relevant elasticities. These are used inChapter 5 to obtain elasticities at the Cooperative and wholesale levels to develop amodel of the B.C. lettuce market. A counterfactual example is then used to showpotential changes in farm revenues by assuming alternative supply conditions to simulatea market in which the Cooperative does not use market quotas.52.0 OVERVIEW OF THE B.C. FRESH VEGETABLE INDUSTRY2.1^Description of Market StructureThe current marketing structure for fresh vegetables in British Columbia originatedin the 1930's. Producer returns below the cost of production and ineffective attempts atvoluntary price controls and cooperative selling convinced farmers that compulsory one-desk selling was essential if the industry were to survive. Petitioning by farmorganizations to the provincial government to provide the necessary legislation resultedin the proclamation of the Natural Products Marketing (B.C.) Act in 1934. Under this Act,primary producers may petition the government to hold a plebiscite to indicate the wishesof growers to become involved in the collective marketing of their particular commodity.If a favourable ballot is returned the Act enables the establishment of a producermarketing board which is then empowered to regulate the marketing, packing andtransporting of the specified product.1The B.C. Coast Vegetable Marketing Board and the Interior Vegetable MarketingBoard, both established a year after the Act, were amalgamated in 1980 to form the B.C.Vegetable Marketing Commission (BCVMC). The BCVMC, operating under the NaturalProducts Marketing (B.C.) Act, is the organization currently responsible for the control andregulation of the production and marketing of vegetables in B.C. To date, fifteen freshvegetables have been designated regulated commodities although the designation varies1Marketing boards in British Columbia, Province of British Columbia, Legislative Assembly, Select StandingCommittee on Agriculture.6somewhat between the Lower Mainland, Vancouver Island and Interior regions. Table 1provides a list of the regulated fresh field-grown crops and their status by region. Theseregulated crops represent approximately 59 percent of the value of fresh field-grownvegetable production in the province for the 1990 crop year.2 The major unregulatedfresh vegetable crops include leaf-type lettuce, greenhouse peppers, broccoli, brusselssprouts, corn, beans and various Chinese vegetables.Table 1:^Fresh field-grown crops regulated by the B.C. Vegetable Marketing Commission.Region 1:^Region 2:^ Region 3:Lower Mainland^Vancouver Island Okanagan & InteriorShare of BC Share of BC Share of BCProduction Production ProductionProduct Status Status StatusAsparagus -^12 - 19 + 69Beets + 80 + 13 + 7Cabbage - green +^66 + 20 + 14- red + 85 + 9 + 6Carrots +^50 + 40 + 10Cauliflower + 51 * 38 + 11Celery +^99 - 1 0Lettuce - head + 86 * 14 < .5Onions - yellow +^55 + 28 + 17Parsnips + 45 + 36 + 19Potatoes +^55 + 11 + 34Rutabagas + 77 + 10 + 12+ = regulated crop required to be shipped through an agency of the Commission- = unregulated* = not shipped through agencies of the CommissionSource: BCMAF, B.C. vegetable marketing guideAll regulated vegetables must be sold at the farm level through agenciesappointed by the BCVMC3 with a provision allowing these crops to be sold directly2Horticultural statistics 1990. Province of British Columbia. Ministry of Agriculture, Fisheries andFood.3Producersnorth of the 53rd parallel (Quesnel) are exempt from this requirement. However, this productionrepresents only a tiny proportion of total fresh vegetable production in the province.7through farmer owned roadside stands which are established on the producer's farmproperty or adjacent to it. The appointed agencies act as the single selling desk forspecified crops, and, by the terms of the Act, are organized as producer cooperatives.The cooperatives coordinate marketing activities for producers including processing,distribution, storage and transportation services. Membership in the cooperative is arequirement for prospective producers of regulated crops unless growers plan to sell fromon-farm outlets only.The percentage of regulated field crops marketed through on-farm outlets is smalland ranges from one percent (for head lettuce) to twenty percent (for beets). Anexception is asparagus where approximately 90 percent of farm output is sold from on-farm outlets.4 The Lower Mainland accounts for about 60 percent of total production ofregulated field crops with the Interior and Vancouver Island regions accounting for 20 and19 percent respectively. Table 1 lists each region's share of B.C. production for each ofthe regulated fresh field crops.To control the amount of produce supplied the BCVMC has instituted a marketquota system for regulated crops. The quota for an individual farmer for any particularyear is based on his/her average sales over a period of three to five years depending onthe commodity. A farmer's market quota is not an absolute fixed amount but representsthe proportion of the market to which the farmer is entitled, based on his/her historicalsales. The market quota limits the amount that the farmer may sell through theagency/cooperative, however this quota does not limit the amount of the commodity that4Nearly all asparagus production takes place in the interior of B.C. centred in the area around Armstrongand the northern Okanagan. Total provincial production in 1990 was 380,000 lb.8the farmer can produce. The cooperative calculates its selling price for the domesticproduct to be competitive with the landed price of imports. Given this selling price,product is sold on a share basis as long as market demand is there. If farmers haveexcess supply, e.g. more product than can be sold at the given selling price, thecooperative may attempt to sell all production at a lower price. Hence the BCVMC'smarketing scheme does not have quotas in the strictest sense of the word. The quotasallows the agency to restrict the quantity of product going to the market if it wants to keepdomestic product price as close as possible to the price of landed imports. Quotas haveno formally established market value with a farmer's quota being established over a fewyears. New producers would typically receive a market quota allotment in the first yearwith the build-up of quota in subsequent years dependent on the producers' actual sales.In general, quotas are also obtainable by leasing or buying farms with existing quota.Although the BCVMC has the power to set prices at the agency level for theproducts it regulates, in practice, the BCVMC is constrained by the landed import pricesof commodities from other production areas. The United States, in particular California,is the major source of imported fresh vegetables for Western Canada and local marketprices are based to a large extent on California costs and the size of its crops. There areno quantitative import barriers for fresh vegetables but seasonal tariffs are applied toimported crops during the peak period of local production. When sufficient volume of acommodity is available locally the BCVMC, on advice from the appropriate cooperative,makes a formal application to the Canadian Horticultural Council for application of thetariff. The tariff rate and the maximum duration of the tariff varies by commodity as does9the commencement date of tariff application. Table 2 summarizes the tariff rates, usualstarting time and maximum allowable duration. At all other times during the year freshvegetables enter duty free. Under the terms of the Canada-U.S. Trade Agreement, annualreductions in tariff rate, begun in 1989, will reduce the rate to zero over a ten year period.Tariffs on U.S. imports are charged according to the Most Favoured Nation (MFN)schedule, hence the current tariff rate for head lettuce, for example, is 9 percent advalorem.Table 2: Tariff schedule for regulated fresh field-grown vegetables (1986)Tariff Rate^ Maximum^Usual periodProduct^BP*^MFN** period allowable^of applicationAsparagus^free^$.055/lb^ 8 weeks^May 1 - June 26(but not less than 15% a.v.)Beets-topped^free^S.01/lb^ 34 weeks^Jul 14 - Mar 10(but not less than 10% ay.)Cabbage^free^S.0125/lb^ 34 weeks^Jul 1 - Feb 24(but not less than 15% a.v.)Carrots^free^S.05/lb^ 20 weeks^Jul 10 - Mar 31Cauliflower^free^$.01/lb 20 weeks Jul 15- Dec 2(but not less than 5% a.v.)Celery^free^$.02/lb^ 18 weeks^Jul 15 - Nov 18(but not less than 15% a.v.)Lettuce-head^free^S.0125/lb^ 16 weeks^Jun 24 - Oct 7(but not less than 15% a.v.)Onions^free^8.015/lb^ 46 weeks^Aug 30 - Mar 28(but not less than 15% a.v.)Potatoes^S.356/cwt All year*British Preferential^ a.v.^ad valorem**Most Favoured NationSource: British Columbia vegetable marketing guide. Province of British Columbia. 1986.The receipts for produce sold by the cooperative are pooled into product and,where applicable, grade pools and distributed to producers after operating expenses of10the cooperative have been deducted. Pool periods vary for different crops. Theexpenses cover all costs of services including overhead and administration incurred in themarketing, processing and distribution of the commodity. The deductions may be on aper unit or percentage basis and also vary between the different commodities. Thedeductions are made known to farmers usually at the beginning of the crop season. Thecooperative's board of directors decide whether any remaining funds at the end of thefiscal year are distributed to growers as patronage refunds or held in the cooperative.2.2^Description of the Cloverdale Lettuce and Vegetable CooperativeThe petition by growers to regulate head lettuce and sell through a cooperativecame about as a result of changing market conditions in the late 1960's. By that time,a post-harvest vacuum cooling process had become standard practice for head lettucein the California fresh vegetable industry. The process extends the shelf life andappearance of the product by removing the excess water that is a major cause ofblemishes and rot. The local produce importers and wholesale trade were familiar withand preferred the vacuum cooled imported head lettuce and wanted to see the locallygrown product processed in the same way. There was also an ongoing concern in thewholesale trade about the quality, mainly the variability in size, of the local head lettuce.Undersized deliveries could be rejected but due to transportation lags in obtainingimported lettuce the wholesaler would be short on supply unless an alternative localsource could be quickly located. Prior to the establishment of the Cooperative, local1 1growers negotiated with the wholesale trade on a one-to-one basis for orders and prices.The local procurement system undoubtably involved higher costs for the wholesaler thandealing with a single produce broker in California who could supply all the producedesired through one transaction.On the growers' part, there was an increasing concern about their lack ofnegotiating power in that wholesalers were often able to pay different prices to growerseven when no quality differences were apparent. Discussions with producers in businessat this time suggest that most farmers were aware of the range of prices within whichlocal produce was selling, although exact prices were not shared except between closeneighbours and friends. It was also difficult to know the quality of product shipped sincefarmers made their own deliveries to the wholesalers. Growers had no way of knowingwhether the prices they received were comparable to prices of similar products thewholesalers imported from the U.S. For local farmers the cost of acquiring informationon U.S. product prices, transportation and exchange rates on a regular basis is very highand unlikely to be justified by their small scale of operation. The lack of information wasa distinct disadvantage for a grower in negotiating prices and a potential source ofmonopsony power for wholesalers. Coupled with this concern was the awareness thatthe quality of local lettuce had to keep pace with California imports if the local productwere to maintain its market share. To achieve this, the issues of consistent quality andpost-harvest processing had to be addressed.Given these circumstances, both farmers and wholesale trade were receptive toa producer agency for marketing head lettuce. There was a general consensus amongst12all parties involved that an agency with regulatory powers was the only way to ensureparticipation by all growers. Regulated marketing through a single agency would: 1)ensure maximum through-put for the vacuum cooling equipment; 2) facilitate theinspection process to ensure that all product would be subject to the same qualitystandards; 3) give growers a collective voice in negotiating price; 4) eliminate price-cutting by individual producers, and; 5) ensure that market shares be allocated togrowers who have historically served the local market and to control the growth ofproduct supply to the market. Although lettuce growers had never organized before, anumber of other field vegetables (see Table 1) were marketed in a similar way and theindustry was familiar with the structure.The Cloverdale Lettuce and Vegetable Cooperative was formed in 1969 anddesignated by the BCVMC as the exclusive marketing channel for head lettuce. At thetime of its formation head lettuce was the only product marketed by the CloverdaleCooperative. In subsequent years celery and cauliflower were added to theCommission's list of regulated crops and the Cooperative was designated the soleagency for their marketing. At present, in addition to the three regulated crops theCooperative undertakes to market, on a contract basis, other fresh vegetables grown byits members. These include leaf type lettuces, hothouse peppers and a variety ofChinese and specialty vegetables.Lettuce production expanded quickly after the formation of the Cooperative from11.4 million lb. in 1969 to 16.2 million lb. just two years later, an increase of over 40percent (see Figure 1). Although there are considerable year to year fluctuations.13production levels remained fairly constant through the 1970's and up to the mid-late1980's. Since that time, output levels have declined somewhat. Changing consumertastes and a substitution to a wider variety of fresh vegetables has contributed to thisdecline. Figure 1 also shows an upward trend in real producer prices since 1969. Partof the reason for this trend may be attributed to the decline, in real terms, of themarketing charges levied on members by the Cooperative.Figure 1: B.C. head lettuce production and pricesSource: B.C.M.A.F.F. Production of vegetable crops together with an estimate of farm value.Although local growers supported the petition to market head lettucecooperatively, they were unwilling (or unable) to finance the purchase of the vacuumcooling equipment. To overcome this problem, the newly formed Cooperative contracted14with a private firm, West Coast Cooling, who would invest in the equipment and providethe vacuum cooling service. The Cooperative would provide marketing and otheradministrative services.The Cooperative continues to contract out the vacuum service at annuallynegotiated rates.5 Services provided by the Cooperative itself for the marketing andprocessing of farm product are charged to members on a fixed per unit basis. The lattercharges are established annually and vary among the different vegetables depending onthe amount of processing performed by the Cooperative. For example, farm packedhead lettuce carries a relatively low charge since it is farm packed and processing islimited to vacuum cooling. Cellophane wrapped lettuce, celery and cauliflower aredelivered in bulk and are washed, trimmed, graded and packed in cartons and/or poly-wraps before distribution to the wholesale trade and therefore have a higher charge.Table 3 provides a breakdown of the Cooperative's charges for selected products for1991. In addition to those listed the charges may include additional levies for specialprogrammes and promotional or lobbying activities undertaken by the Cooperative onbehalf of its members.6 Packing materials are provided at cost to members and theCooperative attempts to minimize the cost of these materials for members through bulkpurchases, particularly shipping cartons which represent a substantial portion of totalmarketing costs.Charges for inputs and services provided by the Cooperative are calculated to5The rates for vacuum cooling are set equal to rates charged by U.S. shippers on produce imported fromCalifornia.6Proposals for special programmes are presented and voted on during general meetings of Cooperativemembers. Once a resolution has passed, any necessary levies are assessed on all members, output.1 5assure that total costs are covered. These charges are deducted from the Cooperative'sselling price to obtain the members' net price per carton of product delivered to and soldby the Cooperative. To achieve an operating at cost position, excess service chargescollected are redistributed back to members on the basis of patronage. Membership inthe Cooperative is available for a nominal fee of one dollar, therefore members'investment in the Cooperative is contributed only through these service charges whichinclude amounts for servicing and repayment of financed capital investments.7Table 3: Cooperative charges, 1991 crop (per carton basis)HeadLettuce(farm-pack)HeadLettuce(cello)Celery(farm-pack)Celery(shed-pack)Cauli-flower(farm-pack)Cauli-flower(cello)Administration .60 .60 .60 .60 .60 .60BCVMC Levy .075 .075 .075 .075 .075 .075Claims .03 .03 .03 .03 .08 .03Inspection .05 .05 .05 --- .05 ---Aphid Monitoring .03 .03 --- --- ---Research .03 .03 .04Processing .30 .35 .60Carton .95 1.20 1.13Cello Products .70 --- .45Trash Removal .01 .20 .10Utilities .02 .20 .20$ .815 $2.795 $ .745 $2.695 $ .805 $3.185*vacuum cooling charges add $0.75 per carton on farm packed productsSource: B.C. Vegetable Marketing CommissionMembership in the Cooperative is necessary for growers who want to marketlettuce, celery or cauliflower and is available to all producers. However, the amount of7Although members have never been required to invest lump sums in the Cooperative, the original membershipwere required to sign promissory notes in order to secure the start-up bank loan. These notes were held by thebank until the loan was repaid. The original membership have made a greater contribution to the Cooperativebut formalizing share or equity arrangement is difficult as some of these members are no longer farming. Onesuggestion has been to base members, equity in the Cooperative solely on the basis of total historic patronage,however a few original members have raised objections citing their greater contribution.16production a new producer can sell to the Cooperative is controlled via a marketallocation or "quota" system. A new grower of lettuce is allowed to deliver 2000 cartonsto the Cooperative during his/her first season of operation.8 In subsequent years agrower's market allocation is determined by a three year moving average of his/herannual sales. The grower's allocation is expressed as a percentage of total sales by theCooperative over the same period. To illustrate, for the 3-year period 1988 to 1990, theCooperative sold an average of 287,000 cartons of head lettuce per year9. Assume arepresentative farmer's average annual sales are 15,000 cartons over the same period.This farmer's allocation or market quota for 1991 is the ratio of his/her average annualsales to the Cooperative's average annual sales. Therefore the farmer in this examplewould be entitled to supply 5.23 percent of all sales made by the Cooperative in 1991.However, if the farmer is unable to make his/her allocated delivery on any day, themissed delivery is not carried over to another time.At the beginning of the production season, the Cooperative estimates sales for thecoming year based on historical sales, and using the percentage allocations approximatesthe amount of product each grower could expect to sell. Actual sales depends on currentmarket conditions and there is no guarantee that a grower will, in fact, be able to sellhis/her "quota" amount. If current sales are lower than historic averages, then eachmember's current sales are also lower. At the same time a grower who plans production80iscussions with people in the lettuce industry indicates that in the Cooperative's history, it is veryrare that a new grower can supply this maximum allotment in the first year of production. Therefore the quotasshould not pose a significant barrier to entry.98ased on production figures published by the B.C. Ministry of Agriculture, Fisheries and Food inProduction of vegetable crops together with an estimate of farm values.1 7closely to historical quotas may miss out on sales if market conditions are favourablesince "excess" demand is allocated to members on a proportional basis. In theory, themarket allocation quota comes into effect only when supply exceeds demand. Inpractice, however, with product readily available from California and other U.S. sources,the quota is used to control local supply such that prices remain close to the level of U.S.imports.A farmer who wished to increase the amount he/she markets must lease or buyfarms with existing "quotas" or market allocations. Quotas are tied to a land base,therefore when a farm is sold the quota goes with the land. Similarly when land withquota is rented, the renter may have use of the quota. In a lease arrangement, the ownerof the farm temporarily gives over his/her quota rights to the person leasing the farm.When the lease arrangement has ended, all the quota rights revert back to the owner,including any changes in the quota. The owner may then use the quota if he/she plansto farm that land himself/herself, or set up a new lease arrangement. While the land isleased the owner has no access to the quota. During the lease period, any increase(decrease) in quota is used by the renter (although the increase reverts back to theowner at the end of the lease period).The farmer could also expand production beyond the predicted allocation andspeculate that market demand will increase, thereby increasing the sales base from whichfuture allocations are calculated. There are no quotas on production  per se, and it shouldbe emphasized the Cooperative is not required to accept all its members' farm output.The quota works by allocating current market sales among producers based on their1 8historical share of market sales, and by controlling the amount of new production enteringthe market.To minimize discounts for lesser quality the Cooperative's manager is authorizedto withhold a member's market allocation if the quality of that member's crop is deemedto be below current market standards. This allocation is distributed to the rest of themembership based on the market quota held by each individual. When allocations aremissed due to substandard quality, to gaps in supply, or any other reason they are notheld over for subsequent time periods.An unintended consequence of the market allocation system is that it maycontribute to supply volatility. Obviously, it provides an incentive to members to maintaina constant supply of quality produce since missed sales will result in lower annual salesand hence, smaller market allocations in future years. However, it may also be anincentive for members who miss sales early in the season to speculate by planning extraproduction for later in the season in an attempt to make up to their normal level of annualsales. Over the longer term however, this system is intended to shift market allocationinitially held by unreliable, low-quality growers to those with consistent supplies of qualityproduct.1°All the fresh vegetables marketed by the Cooperative are available in largequantities from other areas of supply, particularly from California. The Cooperativecompetes with these alternative sources and the price at which the Cooperative sells islargely determined by the landed import price of produce from these sources. Landed1^-°Discussionswith Cooperative members/representatives. There is a general consensus that, over the years,the quality of head lettuce shipped by the Cooperative has improved as a result of this "weeding-out" process.19import prices are estimated weekly by the Cooperative. These are calculated byconverted FOB prices (prices for graded and packed product at origin, e.g. Salinas,California) into Canadian dollar equivalent and adding seasonal tariffs, transportationcosts, vacuum cooling charges, brokerage and handling charges. The Cooperativedetermines it selling prices so as to be competitive with the calculated landed importprice.C:)CDCDCDcn00Dre)- 1-'4. 0(10111111111111111111111CD^CO^CN1^COCN^C■1U04-J03 J e d $00z CVa- )CY)CD—Figure 2: Comparison of weekly wholesale-to-retail prices, 1985-90, Jun/OctSource: Agriculture Canada. Fruit, vegetable and honey crop and market report.  Ottawa.2021Data limitations do not allow a direct comparison of the Cooperative's sellingprices and landed import prices for head lettuce. However, by assuming that wholesalersmark-up domestic and imported lettuce in the same way, the wholesale price quotationsfor domestic and imported lettuce should reflect the cost at the wholesale level. A graph(Figure 2) of weekly wholesale-to-retail prices of domestic and imported lettuce for theperiod 1985 to 1990 shows that (except for a few observations) domestic prices are belowthose of imports.In general, the landed import price is regarded by the Cooperative as themaximum it could sell its produce at." However, depending on current marketconditions in both California and British Columbia, it is possible the Cooperative's pricecould be higher, lower or the same although the first instance is rarer. Quality differencesbetween the local and imported California product are the usual cause for discounts onthe local product. If the supply of imported product is very weak, the Cooperative isbetter placed to price nearer to landed import levels. Conversely, during normal or strongsupply conditions, Cooperative prices lag behind the landed import level after allowing forany quality differences. Throughout the production season the Cooperative's managercarries out weekly inspections of member's crops to determine the quality and quantityof produce available. A weekly bulletin is sent to the wholesale trade advising ofquantities available and the selling price which is set after consideration of both local andCalifornia supply conditions. Quoted prices are still subject to change after the bulletinis issued. For example, if FOB California prices decreased, the Cooperative's selling price11^-Discussionswith Cooperative members/representative.22would follow accordingly. However, increases in the imported product price are rarelymatched by the Cooperative (within the same week). This suggests the wholesalers havesome market power over producers, even when acting collectively.During periods when strong supply conditions prevail, maintaining price near thelanded import price is more difficult if the Cooperative is to sell all its members' output.The Cooperative could enforce quota allocations so that only the quantity clearing themarket at the current price is sold, or attempt to sell all the available product at a lowerprice. Conceptually, we may think of the Cooperative as manipulating its supply curve.Selling all members output would be equivalent to a rightward shift of the Cooperative'ssupply curve from its initial position where quota allocations are enforced and supplyrestricted. Given a downward sloping demand curve, more product will sell only at alower price. The quota allocations allow the Cooperative to maintain supply at levelsconsistent with the calculated "target" price.The discount necessary to move excess supply depends on the interactions ofa number of factors. If the quality of the local product is consistently high amongst allgrowers, and if prices of California imports are firm (implying weaker supplies) then thediscount to move all local produce will be small relative to the discount necessary whenCalifornia prices are weak. Where possible, the Cooperative attempts to negotiate withwholesalers for special orders (e.g. supermarket "specials") such that the selling price isdropped only on the "special" orders with "normal" prices applying on other Cooperativesales. Since members receive the same price for their deliveries, discounting affects allmembers whether or not the individual member has excess supply. For example, if a few23growers have excess supply (e.g. more product than can be sold at a calculated marketprice competitive with import prices) and the Cooperative makes a decision to lowerprices to move this "excess" output, then all product is affected unless the Cooperativehas negotiated special sales prices. A member who does not have output in excess ofhis/her market quota would also receive the lower price. It is clear that given the CIFprice, a major concern of the Cooperative will be to control supply by members to a levelthat would permit achieving this target CIF price.243.0 COOPERATIVE PRINCIPLES AND ECONOMIC THEORY3.1^Cooperative PrinciplesThe preceding chapter described the operations of a cooperative organization, theCloverdale Lettuce and Vegetable Cooperative, but with no attempt to relate cooperativeprinciples to the organization's observed behaviour. This chapter will review the literatureon cooperatives, and discuss how it may apply to the Cloverdale Lettuce Cooperative.To better understand the effectiveness of cooperatives in a market-orientedeconomy, it will be useful to distinguish cooperatives from other types of businessenterprises. The basic premise underlying any cooperative action is that through jointeffort and mutual self-interest individuals may collectively achieve objectives unattainableby acting alone (Dunn, 1989). Although this premise could apply to other organizationsa number of principles uniquely define the cooperative organization. The very earliestprinciples were established by a group of weavers in Rochdale, England who formed apurchasing cooperative in 1844. Many of their principles are accepted to this day (citedby Coifed!, 1987, p.174):open membership;one vote per member;provision of capital by members with limited return on investment;net margins returned to members on basis of patronage;market prices to be charged, cooperative savings or profits returned as patronage25refunds to members;cooperative management to be under the control of membership-elected officers,andregular, comprehensive financial accounts made available to members.Exceptions exist, for example, agricultural cooperatives restrict their membershipto farmers and Staatz (1987a) notes that marketing cooperatives may restrict membershipin the short run due to limitations of plant capacity, product quality control or otherreasons. Many cooperatives obtain financing from commercial banks for investment andoperating capital. Nevertheless, these principles remain the basis of modern agriculturalcooperative associations.A number of writers have catalogued three essential characteristics that setcooperatives apart from other types of business organization (Staatz, 1987a; Dunn, 1989;Barton, 1989):1. The user-owner principle:  Those who have capital invested in the cooperative arethe majority users of its services. 122. The user-owner-control principle: Control of the cooperative is by those who useand own it. Furthermore, control is not proportional to equity investment. Votingpower is limited on a one member/one vote basis, by patronage or some otherform of restriction.3.^The user-owner-benefits principle: The benefits owners receive from capital12Thereare exceptions to this, for example, cooperatives may accept business from non-members as a way ofusing up excess plant capacity but these users have no say in cooperative decisions nor share in patronagerefunds.26invested in the cooperative is allotted on the basis of their usage or patronage.This principle is often stated as business or service-at-cost.The above principles define the farmer cooperative: a user-owned and user-controlledbusiness from which benefits are derived and distributed amongst owners on the basisof their usage. 13In addition, an important characteristic of agricultural cooperatives is the elementof vertical integration from the farm level. For several reasons (discussed in detail in thefollowing section) it may be beneficial for producers to move into secondary productionstages closer to the consumer. However, at the secondary processing stage theminimum size of plant needed to operate efficiently usually preclude individual farmproducers expanding operations beyond primary production activities through verticalintegration. Thus all cooperatives represent the joint action of a group of farm producersto integrate upstream (agricultural inputs supply cooperatives) and/or downstream(marketing and processing cooperatives).3.2^Implications of the Cooperative PrinciplesThe cooperative principles described above result in a form of businessorganization with incentives and decisions different from those faced by investor-oriented13SeeLeVay (1983, p.3-6) for a review of some other definitions and characteristics of what constitutesa cooperative.27firms (10Fs).14 To the extent that stockholders influence a firm's decisions, it could beexpected that the decisions of a firm will be different if its stockholders were major usersof its services than if they were not (Staatz, 1987a). Some implications arising fromcooperative principles and characteristics are discussed below.The vertically integrated nature of the farm and cooperative imply that their ownershave an incentive to jointly maximize the profits of the farm/cooperative system. As aresult, cooperatives may give greater consideration to their member's fixed costs andhave different pricing practices than that of 10Fs. Members' fixed costs are likely to bemore important in cooperative decisions than in an 10F because the market transformsthe fixed costs of an 10F's customers or suppliers into variable costs for the 10F.Processing cooperatives, for example, often have a policy of accepting all farm productdelivered by their members because of the need to account for the member's fixed on-farm costs. An IOF usually does not have to deal with its suppliers' fixed costs; they aretransformed via the market into the raw product price paid by the 10F, which the IOFprocessor considers as a purely variable cost (Staatz, 1987a).One of the most important consequences of owners being users of the firm'sservices is that pricing decisions are of vital interest to the owners since the income ofmember-farmers are more dependent on the individual prices charged or paid by thecooperative than on its overall profitability (Staatz, 1987a). In contrast, shareholders in1^-4Firmsother than cooperatives may be referred to as investor-owned or proprietary firms, and profit,private, ordinary or corporations. However, cooperatives are also private and investor owned organizations andthey seek to increase profits or the economic well-being of their members. The distinction between the twoorganizational types is that cooperatives return net income on the basis of usage or patronage whereas otherbusinesses return net income on the basis of investment. Thus the term investor-oriented makes the distinctionclearer (Barton, 1989). Throughout this paper it is assumed that 10Fs do not integrate backward into primaryfarm production activities.28an 10F are only interested in the "bottom line". Pricing decisions in a cooperative affectnot only the firm's performance but have a direct effect on the member's willingness topatronize and invest in the organization. Furthermore, certain price pooling policies willaffect the distribution of income amongst members. In multi-product cooperatives, thesensitivity of owner-users to price policies will limit the use of price-cutting or cross-subsidizing strategies to increase the cooperative's share of the output market.The user-owner principle may also impose limitations on the potential for acquiringequity capital. Farmer cooperatives can only increase their equity base by applying tocurrent members or through increasing membership (Staatz, 1987a). The latter optionmay not be viable due to the absolute limit of producers in the cooperative's area. Theformer may be difficult as ownership of the cooperative provides few direct financialbenefits to the members since limited returns are paid on their equity and the interestdoes not reflect the profitability of the cooperative (Corman and Fulton, 1990).Furthermore, shares in cooperatives are not traded on secondary markets making itdifficult to gain financially by selling the shares should they appreciate in value. Sincereturns from investment in the cooperative are low and benefits are achieved throughusage, members have an incentive to use the cooperative but not to invest in it. Inaddition, the one-member/one-vote rule implies an individual member cannot increasehis/her control over cooperative affairs by investing in the cooperative.To alleviate this equity problem many cooperatives have adopted the practice ofretaining a portion of earnings in an attempt to force members to invest in theorganization. Although retained earnings are critical for the future performance of the29cooperative, such an investment also means members are deprived of an immediatebenefit in the way of a cash patronage payment (Corman and Fulton, 1990).The one-member/one-vote rule followed by most cooperatives is intended toprovide all members with an equal say in cooperative affairs. A result of this rule is thepossibility that a few producers who contribute a large part of the patronage may bedominated by a majority of small-scale patrons. The fact that the large producers maytake their patronage elsewhere should ensure that their voices be heard. However, if thecooperative is the only firm in the market, no such outcome is assured and the decisionmaking process could be contentious.In summary we can review these implications with specific reference to theCloverdale Lettuce and Vegetable Cooperative to define goals that may be implied fromthe observed behaviour of the Cooperative. Most pertinent to our topic of interest is thatprice and output decisions of the Cooperative need to reflect the interests of all itsmembers. As a member controlled organization the Cooperative has an obligation bothto members willing to accept lower prices in order to move their over-supply and tomembers who have, in a sense, controlled their production to minimize the need todiscount on price. In periods of above normal supply, given the usual demandassumptions, all output will be sold only by lowering price below the landed FOB price.Since members receive the same price, some conflict is likely to arise between memberswith over-supply and those without. The existence of market quotas implies theCooperative has discretion over the amount of farm output it accepts. If theCooperative's objective is price maximization, (where price maximization is taken to be30achieving the price level of California imports) then the Cooperative must fully enforce themarket quotas to ensure that over-supply does not push price downward. However,there is evidence the Cooperative does take into consideration members' supplyconditions in determining its weekly selling price, and attempts to sell product on specialterms to the wholesale trade. Therefore, a plausible assumption is that the Cooperativetries to strike a balance between maximizing price (as defined) and maximizing sales.As the sole agency for marketing head lettuce, membership in the Cooperative isno longer voluntary but mandatory. Regulating the marketing of these crops serves to:1) Coordinate pricing of farm output, 2) Facilitate application of grading standards, 3)Ensure the fixed costs of marketing and processing will be spread over the maximumamount of output available, and; 4) Ensure that investment capital is contributed byproducers on the basis of their output and use of Cooperative services and facilities. Onthe other hand, individual producers may feel Cooperative goals do not coincide with theirown but have no alternative except to continue marketing with the Cooperative. In thelonger run, however, dissatisfied producers could switch to non-regulated crops, whichwould shrink the Cooperative's operations if no new producers enter the lettuce market.While head lettuce production has declined slightly since the early 1980's (referto Figure 1), it is uncertain whether this is due to a structural change in lettuce demandor to dissatisfied growers switching to other crops. However, it is certainly the case thatduring the 1980's growers did diversify into other (non-regulated) crops. Leaf lettuceproduction, for example, increased from about one million pounds annually in 1981 to31approximately ten million pounds by 1989.15 Since head lettuce production did notdecline by a similar amount, the increase in leaf lettuce was probably a result of bringingnew land into production.3.3^Economic Theory of Cooperative FormationAn important characteristic of agricultural cooperatives is the element of integrationfrom the farm level. The economic incentives for vertical integration via a marketingcooperative will be discussed by looking at the relationship between the amount that amarketing firm is able to pay for raw farm output and its selling price of the produce:^IR) = PM ; m^(1)where:r = maximum price paid by marketing firm for the raw farm output, R, where r is afunction of the amount of R purchasedp= output price received by marketing firm for it final product, Y, where p is a functionof the amount of Y producedY = final product, where Y = R/kM = marketing margin or cost, including normal profit, of labour, energy, capital andmaterials to market and process the raw productk = conversion factor for the number of units of raw farm product (R) needed toproduce one unit of final product (Y)15B.C.Ministry of Agriculture, Fisheries and Food. Production of vegetable crops together with an estimateof farm value. Victoria, B.C. various issues.32If the raw input and final output markets are competitive and the conversion factor is one,the above relationship becomes:= p - M^ (2)This relationship also shows how farm gate price is related to the Cooperative's sellingprice and is an important component of the model developed in the subsequent sections.In a competitive market firms function at the lowest cost possible to survive, thusthe margin, M will be as low as possible. Competition amongst firms to secure suppliesof raw farm product will bid up prices paid to farmers to the maximum indicated inequation (1). However, in markets where competitive forces are weak or lacking, theabove relation shows three possibilities where integrating cooperatively could increasefarmer's welfare (Sexton and Iskow, 1988):1. The cooperative may be able to lower the marketing margin, M, thus increasingthe raw product price, rf, paid to farmers.2. The presence of a producer cooperative may prevent the exercise of monopsonypower by downstream firm(s) with market power who may force farmers to accepta smaller price than rf defined in equation (1).3.^The cooperative may be able to increase rf by influencing p. In the extreme case,the cooperative may act as a producer cartel and select output so as to maximizeproducers' joint profit.Each of these possibilities will be examined below. Additional conditions existwherein cooperative action will benefit producers. One is that there are no firms in themarket to perform the marketing functions, a second is that the cooperative may be able33to mitigate uncertainty, and a third is that cooperatives, when viewed as a public good,may have positive indirect effects on farmers. These will also be discussed in thefollowing sections.Marketing MarginsThe scope for reducing margins and raising rf through cooperative verticalintegration depends on two factors: 1) whether the cooperative can obtain lower priceson inputs used in the processing stage, and/or 2) the ability of the cooperative to performmore efficiently than other firms. These factors are equally relevant for 10Fs, thereforethe question is whether a marketing cooperative has any advantage in these two areas.Since cooperatives operate in the same environment as 10Fs, they are likely toface the same set of variable input prices. However, Sexton (1986) and Sexton and lskow(1988) show the cost of capital is less for low tax firms, such as cooperatives, who canpass net income through to their members without paying tax on it.16 They note thatwhile this tax advantage is available to any vertically integrated firm and is not, in itself,sufficient reason to form a cooperative, it does confer an advantage for a cooperativecompeting with non-vertically integrated 10Fs.Conducting exchanges within a vertically integrated enterprise can be moreefficient than exchanges through markets if vertical integration reduces transactions costs.These include the costs of acquiring and analyzing information, decision-making, andnegotiating and carrying a contract to fulfilment (Staatz, 1987b). In agricultural markets,16Thereferences cited discuss U.S. cooperatives. Taxation of cooperatives in Canada is different. Theyare taxed at the corporate rate but allowed to deduct patronage dividends paid or credited, provided that aprior legal obligation existed to do so (Taylor, 1971).34transactions costs may be particulary high due to the existence of sunk costs or assetfixity at the farm level.Asset fixity or sunk costs occur when the difference between asset acquisitionprice and its salvage value is high. The more specific an asset to a use the larger thesunk cost. When sunk costs are high, one firm may take advantage of another throughopportunistic behaviour, for example, by reneging on contractual obligations.Staatz (1987b) notes that the risk of opportunistic behaviour arising from assetfixity is a major factor for the prevalence of cooperatives in marketing fresh fruits andvegetables and milk. Producers of these commodities are particularly vulnerable toopportunism since the harvested product is a sunk asset with few resale opportunitiesbecause of its perishability. Vertical integration via a producer cooperative minimizes thepotential for opportunism since producers have control of both primary and secondaryactivities. Hence a producer cooperative may indeed have lower transactions costs. Acooperative may be able to operate with lower margins and thus return a higher price toits producer members.Avoidance of Market PowerIf a commodity market is characterized by only one (few) buyer(s), farmers facemonopsony (oligopsony) power and may gain by forming a cooperative to break ormitigate the monopsony. The cooperative in this circumstance seeks to achieve the sizeeconomies (reducing M, the per unit processing cost) usually implied by the presence ofmarket power (Sexton, 1986). The avoidance of market power is the most familiarjustification for cooperatives in a market-driven economy. With few buyers of the raw farm35product, competitive pressure that causes raw product price, rf to be bid up are weak orlacking. In addition, monopsony or oligopsony power permits the exercise of pricediscrimination, that is, paying different prices to different producers for no cost justifiedreason.The ability of cooperatives to counter market power depends on several issues.The first concerns costs and relates to the shape of the firm's cost curve. A cooperativeentering an industry to counter buying power must be able to attract sufficient volume tobe able to operate in the flat portion of the average cost curve. If this is not the case,then the cooperative's marketing margin, M, will be higher than that of other firms and theprice it is able to pay to its members may be less (Sexton and lskow, 1988).A second point to consider is that a bargaining association may be sufficient toprevent price discrimination and offset buying power. By playing processing firms offagainst one another and/or threatening to withhold the product by forming a cooperativeto market and process the raw output themselves, a farmer's bargaining group maycause prices to be bid up. The bargaining association can achieve price enhancementgoals without the financial commitments and concerns about achieving efficient scalerelevant to formation of marketing cooperatives (Sexton and lskow, 1988).The third concern relates to entry into the marketing sector by a cooperative.Sexton and Sexton (1987) writes that price-cost deviations in an industry signal profitableentry opportunities for all firms, not just cooperatives, so that the presence of marketpower beyond the farm gate does not adequately explain a cooperative's presence. Hesuggests that differences in the determinants of the entry decision are the major36consideration. An 10F's concern lies solely with behaviour in the post-entry market whilea cooperative is most concerned with market characteristics if it does not get involved.Thus potential entry by a cooperative can be deterred by a limit price. However, invokinga limit price to deter entry requires an incumbent firm to pay farmers a sufficiently highprice so that there is no incentive for cooperation formation. Price "wars" between theincumbent firm and the new entrant involve paying farmers more for their output. Limitprices that deter cooperative entry are conceptually similar to sustainable prices andtherefore confer potential cooperatives with a powerful welfare enhancing role (Sexton andSexton, 1987).Potential for Producer CartelsFormation of producer cartels to exercise market power yields maximum pricebenefits for farmers. As the sole seller of a commodity the cartel may exercise monopolypower by selecting the output level which will maximize profit. However, such cartels aresubject to cheating by producers who collectively benefit from cartel power, but, becauseprice usually exceeds marginal cost, individually have an incentive to supply more output.The conditions for successful producer cartel include: representation of a majorproportion of producers' output; the ability to restrict potential imports into the cartel'smarket; the establishment and effective enforcement of a quota system to control outputand/or a system to control the timely release of product to the market. Any of theseconditions may be difficult to achieve. For example, production may be spread over sucha wide geographic area that organization is very costly or perishability of the commoditymay render market supply control impractical, therefore the scope for cartelization via37producer cooperatives is limited.However, a producer cooperative may be able to achieve a degree of marketpower arising from cartelization if it controls the only processing and marketing facilitywithin a particular region (Sexton, 1986). The great weakness of cartels is the lack of aneffective means of enforcing output restrictions but if producers must sell through thecooperative, output control will likely be more feasible (assuming imports to the regioncan also be controlled). The existence of a single plant market suggest there may bebarriers to entry at the processing/marketing stage due to substantial fixed costs and/orlarge volumes necessary to achieve the minimum efficient scale. Therefore, acquiringmarket power will probably be most feasible for cooperatives which handle commoditiesrequiring capital intensive processing (implying presence of scale economies) and arestorable. It would be least effective for perishable fresh fruits and vegetables whichrequire minimal processing and must be sold quickly.Reduction of Risk and UncertaintyAssured access to upstream or downstream markets is an incentive for verticalintegration common to all business. With the assumption farmers are risk adverse,Sexton and lskow (1988) discuss two basic market risks: 1) Markets where prices areinflexible and markets do not clear and producers may be unable to sell all they want atthe market price, and; 2) Markets where prices are flexible and large fluctuations in priceincrease the farmer's exposure to income uncertainty. In the first instance a cooperativecan alleviate the risk by strategically choosing membership so that the probability ofmarketing all its members' output is high. Risk is thus shifted from members to producers38outside the cooperative. Alternatively, an open membership cooperative could providea "home" for all farm output. Risk is pooled and the costs of sales rationing is spreadover all producer-members to minimize their effect on individuals, for example a non-cooperative buyer may purchase all output from some farmers and none from others.In the second instance fluctuations in market price expose farmers to price andincome uncertainty. Contracting can help reduce this price risk but may lock tradingpartners into unfavourable prices due to uncertainty of market conditions at time ofharvest. More flexibility is offered through contingency pricing as practised by producercooperatives. Producers receive an interim payment for deliveries with final price receivedcontingent on the cooperative earnings and market price at time of final sale. Staatz(1987b) suggests cooperatives may be more successful than 10Fs in using contingentpricing because producer-members have access to the cooperative's financial accountsand are less likely to believe the cooperative is withholding information or is usingcontingency pricing to act opportunistically towards them.Risk due to price fluctuations may be diversified in multi-commodity cooperativesthat allocate returns from a single price pool so that low prices in one commodity may beoffset by high prices for another. In single commodity pools where prices may fluctuateweekly or even daily, pooling spreads price risk among member to minimize their effectfor any one individual.39Cooperatives as Public GoodsIn addition to the direct effects, Staatz (1987b) suggests that cooperatives maybe viewed as a type of public good with positive indirect impact on the agricultural sector.First, cooperatives may play a role in bringing about a more socially desirable regionaldistribution of income. This claim rests on the argument the 10Fs extract profits from ruralcommunities and re-invests them in metropolitan areas. In contrast, cooperatives rebatenet margins to its members who re-invest locally thus stimulating greater regionaldevelopment. Second, cooperatives formed to compete directly with existing 10Fs mayimprove economic efficiency by compelling the 10Fs to expand output and increase theirX-efficiency. Third, cooperatives may be more likely to encourage and invest in thedevelopment of new crops and production techniques. An 10F may be reluctant toengage in farm improvement activities because producers can potentially use their newskills to patronize competing firms. Unified ownership of farming and marketing activitiesreduces potential opportunism on the farmer's part since their investment in thecooperative gives greater assurance of their continued patronage.Provision of ServiceFarmers have often faced situations where no private firm is willing to serve themarket. However, a cooperative, by recognizing the joint profitability of farm productionand marketing, may be able to extract value from marketing where the benefits of doingso exceeds the cost (Sexton and Iskow, 1988). To illustrate this point, refer to Figure 3below.1717Thefollowing analysis is taken from Sexton and Iskow (1988).40The farm's supply curve is labelled S and indicates the minimum price they needto cover variable costs. The maximum a downstream firm could pay per unit for anygiven quantity of the raw farm product while covering all costs and a normal return oninvestment is given by the net average revenue product (NARP) curve. The net marginalrevenue product (NMRP) curve shows how much each incremental unit of the farmproduct is worth.Figure 3: Optimum behaviour for a marketing cooperative when no single price will cover costs.Source: Sexton and Iskow (1988)As drawn, for all output levels the maximum a downstream firm could pay liesbelow the minimum farmers needs to cover variable costs, hence no firm could profitablyserve the market. However, the NMRP curve does lie above the supply curve for somelevels of production. Assuming farmers do form a cooperative, the optimal farm outputwill be Q*, where the marginal cost of production is equal to the value of the marginal unitof raw product to the processor. Total net revenues from production and marketing theproduct is represented in the graph by the large dotted area minus the small cross-41hatched area. At the optimal output level Q*, the cooperative must pay the producer P1for each unit. However, at P1 the cooperative is not covering total costs. To recover this,it could then charge annual membership fees equal to the amount (P1 - P2)Q*. It is thisflexibility in pricing which enables a cooperative to profitably operate where an IOF couldnot. Farmers will benefit as long as the profit from growing the crop is greater than themembership fee, or producer surplus (area acPi) is greater than membership fees (areaP2bcP.1).3.4^Price and Output Decisions and Membership PoliciesThe previous section discussed how a producer cooperative could increasefarmers' welfare by integrating into post-farm production activities. The following sectionwill present a model (Pauly and Redisch, 1973) of a cooperative to illustrate howmembership policies can alter equilibrium output and price positions. A closedmembership policy is shown to have an effect on cooperative output analogous to thequota restrictions used by the Cloverdale Lettuce and Vegetable Cooperative.To develop the discussion, a cooperative is assumed to produce one output (Y).Capital (K), labour (L), and the raw farm product (R) are used to produce Y according tothe cooperative's production function:= RI<^ (3)The raw produce, R, is supplied by member-farmers while the cooperative supplies labourand capital. Members have control of cooperative operations and ensure that services42provided to them are produced in such a way as to maximize their net return from the rawproduct they supply to the cooperative. Farmers supply the raw input according to theirmarginal cost schedules at or above their minimum average variable cost. For simplicitywe may assume that farmers have constant marginal and average costs without seriouslyaffecting the model's general results. This implies that the supply curve of R is horizontalat the price received by farmers.The demand for the final product, Y, can be written:Y = nPy) ; a IlaPy < 0^ (4)where P,, the price of the cooperative's output. By assuming a competitive outputmarket, the cooperative faces the inverse demand curve:^P = 7:7^ (5)Y^YThe price, Pc, that the cooperative charges its members for services provided isgiven by: PcY = wL + cK, where w is the wage rate of labour and c, the user cost ofcapital. The cooperative operates to maximize the net incomes of its members and itsobjective is to maximize (PyY - PcY) subject to the production function (3) and demandfunction (5) above. The difference between (PyY - PcY) is the surplus amount aftercovering the cooperative's processing costs. This problem is identical to that of a profitmaximizing firm although profit is the cooperative surplus which is distributed to members.The marginal conditions for optimal employment of labour and capital are the same, thatis, the first order conditions for maximization yields the condition that marginal factor costmust equal the marginal value product for inputs. By assuming perfect competition in43input markets, marginal factor cost is merely the competitive market price of the input.The amount of raw material supplied by member-farmers may be variable overtime, or the cooperative could control deliveries by enforcing restrictive quotas on itsmembers. The problem of the cooperative is to determine the optimal amount of productto process given its objective is to maximize the net income of its members. The abilityto control the quantity of product processed depends on the cooperative's membershippolicies. Equilibrium output results of three alternative policies are discussed: (1) closedmembership; (2) open membership; and (3) a policy in which the cooperative acceptsnew members on a discriminatory basis.3.4.1 Closed MembershipAlthough open membership policies were espoused by early cooperatives, manymodern agricultural cooperatives have closed policies. The decision to admit newmembers is made by the existing members who, by assumption, have control over thecooperative's operations. All members share in the residual income of the cooperativeon the basis of their share of total output. If the objective of the cooperative is maximizethe net return per unit of raw input (R) supplied by members, then members will be willingto add new members thereby increasing the amount of R, as long as the additional inputof R causes the net return per unit of R to rise (or not fall).Therefore, the objective function to be maximized by the cooperative is:The necessary first order conditions (FOC) for a maximum are:44(6)PvY - wL - cKPr = 's.t. Y = (1_, K, R)Py = Py15- afil., K, Fl)w =  YaLi''^= aRL, K, FOc ^YaKT am., K, FOP r =. Y aRThe third condition above (equation 9) states that the quantity of raw product processedby the cooperative is determined by equating the marginal revenue product of R with thenet average revenue product of R. Thus a closed cooperative will accept additional rawproduct from new members as long as its contribution to total revenues is greater thanthe average revenue per unit of R which members currently receive. Satisfying the thirdcondition also implies that there are farmers willing to supply the raw output at the priceoffered by the cooperative, or in other words, the equilibrium value of Pr must be at leastequal to the farmer's marginal cost of producing R.Figure 4 illustrates the long-run equilibrium position of a cooperative whosemembers are assumed to have constant marginal costs. Within the cooperative'sprocessing activities capital, labour and other inputs take on short-run optimal values asthe amount of raw product processed, R, varies along the X-axis. The curve labelledR45NARP (net average revenue product) indicates for each level of R the maximum netrevenue per unit or price the cooperative can pay members for their output. Returns toscale in cooperative operations initially result in the upward sloping segment of the NARPcurve but decreasing returns and diminishing marginal revenue eventually cause the curveto turn down. The maximum per unit net average revenue for the raw product, Pr*, isreached at the intersection of the net marginal revenue product (NMRP) and the NARPcurves, where a closed membership cooperative would process R* of raw product.Figure 4: Equilibrium output positions of a cooperative under alternative membership policies.Source: Pauly and Redisch (1973)Pauly and Redisch (1973) make note of some seemingly paradoxical conclusionsof this cooperative m'odel in that supply response by the cooperative to changes inproduct and factor prices are opposite to that predicted by models of 'orthodox' firms.For example, an increase in the given price of output could result in lower output levelsand smaller quantities of R supplied by members. This results because an increase inP produces a proportionate increase in the marginal revenue product of R, assuming the46K and L are held constant but the return per unit of R will rise more than proportionately.In terms of Figure 4, an increase in output price shifts both the NMRP and NARP curvesupward but the extent of the shift in the NARP curve is greater so that the intersection ofthe two curves will now occur to the left of its current position and the maximum NARPwould be attained at a smaller quantity of R.3.4.2 Open MembershipThe open cooperative does not restrict membership and accepts any amount ofraw product by members subject to covering its processing costs. The equilibriumquantity of raw product processed is reached where the average net revenue per unit ofR is equal to the member's marginal supply price. In Figure 4 this is the point labelledA0:• Alternatively a cooperative facing an upward sloping supply curve for R whichintersects NARP curve to the left of its maximum will be in equilibrium at R00 as shownin Figure 5. Such a cooperative will pursue an open membership policy as it would liketo increase the quantity of R processed in order to move up the rising part of its NARPcurve and increase per unit revenues for its members. A closed policy could then beadopted when the quantity of R processed reaches R*.3.4.3 Discriminatory PolicyAs the name implies, a discriminatory policy also requires that a cooperative beable to restrict its membership or control the amount of raw input supplied to it. Referringto Figure 4, the cooperative faces a horizontal supply curve for R and sets the quantity47it processes in the same way as would a profit maximizing firm. The quantity of productprocessed will increase until the NMRP of R is equal to the member's marginal supplyprice. Equilibrium quantity of product processed is Rd and members receive a per unitreturn for Rd equal to its marginal cost. However, members share in the profits of thecooperative according to their share of total patronage, or the amount Yr+ per unit of Rsupplied. Patronage refunds are essentially an adjustment to price to bring thecooperative to operation at cost.Figure 5: Output equilibrium position of a cooperative with open membership policy.Source: Pauly and Redisch (1973)To achieve an equilibrium at (S = MC = NMRP) requires an initial assumption thatfarmers consider only the price received at delivery when making production decisionsand regard patronage refunds as windfall gains (Cotterill, 1987). Failing that, the48discriminatory policy (or closed membership) requires multipart pricing schemes or quotasto preserve the optimum delivery levels of the raw product. If members are unable todistinguish between the market price paid and the patronage dividend, the cooperativemay be forced to operate at the position where price paid is equal to the NARP. Notethat the models of the discriminatory and open membership policy cooperatives do notexhibit the potentially perverse supply response present in the closed membershipmodels. Increases in output price will lead to increases in output and the amount of Rsupplied by members.The models discussed above provides a framework to evaluate the benefits ofcooperative marketing. It turns out that benefits depend to a large extent on themembership policy of the cooperative. Membership policy may be viewed as analogousto a form of output control, indeed, even a closed membership cooperative would, in thelonger run, have to institute some form of output restriction. Otherwise, ceteris paribus.in the long run members' supply curve could shift out as more farmers join thecooperative in response to the higher returns offered, and erode the benefits ofcooperative marketing.3.5 Basis for Formation of the Cloverdale Lettuce & Vegetable CooperativeIn this section, aspects of the economic theory of cooperative formation will beapplied to the Cloverdale Lettuce and Vegetable Cooperative to establish the basis ofbenefits of the Cooperative to its members. Several elements would appear to have49relevance with respect to the Cloverdale Cooperative:1. Provision of Service. One primary concern of lettuce growers was the need fora central standardized inspection service. This could have been provided by agovernment agency paid by a user fee or it is possible that a private firm couldhave undertaken this service as well as provide the post-harvest vacuum coolingprocess. However, such a function creates another market tier. Verticalintegration on the part of farmers reduces the margin between retail and farmprices by returning profits from this marketing service to farmers as patronagerefunds or as higher prices. An added advantage to a cooperative performing thisfunction is that member-farmers are more likely to accept inspection processsince they have a say in the conduct of cooperative business.Given the high initial investment cost of the vacuum cooling facility, someguarantees of maximum throughput were necessary to ensure that all operatingcosts could be recovered. Thus, a cooperative with a binding membership wasthought to be a logical way of achieving this aim.2. Avoidance of Market Power. Ex post, it is difficult to establish whether significantmonopsony power was present before the Cooperative was formed. Althoughtwo or three wholesaler buyers dominated the market, there were numerous smallwholesalers as well. However, the high cost of acquiring market information byindividual farmers is an important factor in conferring market power towholesalers. This information asymmetry permitted wholesalers to exercise pricediscrimination when dealing with domestic producers. A cooperative can achieve50a scale of marketing which greatly decreases information costs and thus putproducers on a more even negotiating basis with wholesalers.3.^Producer Cartels. Although scale economies and enforcement of gradingstandards are cited as reasons necessitating compulsory membership in theCooperative, the potential for output restriction as a means of raising prices mustbe admitted as another reason. This potential is given credibility through theexistence of the market quota scheme supervised by the Cooperative. While theCooperative has control over domestic production, imported lettuce is free toenter B.C. subject only to a seasonal tariff. Therefore the Cooperative's ability torestrict product to the market is severely hampered and, assuming the sameproduct quality, any attempt to raise local lettuce price above landed import levelswould eventually result in the loss of the entire market. However, the Cooperativecan use the market quota system to restrict the supply of local lettuce reachingthe market. By restricting supply the Cooperative's supply curve would be shiftedto the left of where it would be if no market quota allocation were in place. Theadvantage for members to comply with the market quotas appears to stem fromcollectively holding local supply to a level which allows domestic prices to be setnear those of the imported product.The first two sources of cooperative benefits discussed above are difficult toquantify  ex post due to the unavailability of data and non-market factors such as greaterloyalty and trust by farmers to a member owned organization. It is possible however to51examine the basis and extent of benefits to members arising from the Cooperative'smarket quota scheme. The balance of this study will be concerned with the analysis ofthese benefits. The next chapter discusses the empirical estimation of supply anddemand parameters which will be used in the last chapter to simulate a lettuce marketmodel where it is assumed no market quotas are in place.524.0 A MODEL OF THE B.C. LETTUCE INDUSTRYThe benefits of cooperative marketing were discussed in the previous chapter and,as noted, it is difficult to quantify certain benefits due to the unavailability of data. Of thepossible benefits, benefits of the quota used by the Cloverdale Lettuce and VegetableCooperative to constrain supply may be examined by modelling the B.C. lettuce industryand simulating alternative supply parameters.This section will discuss estimation of demand and supply parameters for headlettuce in B.C. These parameters will be used to calibrate a model of the B.C. lettuceindustry under the current system of market quotas. This price and output equilibriumwill be used as a reference point to compare changes in farmer's revenues in theabsence of the Cooperative's marketing system. It is hypothesized that the Cooperativehas altered the market structure to the benefit of its members by influencing producers'supply elasticity and supply curve through the use of market quotas. The premise for theanalysis is that supply has been restricted and rendered more inelastic. The lower outputlevel which results enables the Cooperative to set higher prices than they wouldotherwise. For a higher price to lead to higher revenues, the demand curve must beinelastic. The demand for basic food commodities is considered quite inelastic, howeverdemand for lettuce in B.C. will be estimated as part of the analysis. Price and outputsolutions under alternative supply assumptions will be used to calculate changes in farmerrevenues if no market output controls existed.Primary market demand and farm supply will be modelled by linear equations.53The derived wholesale demand and the derived supply by the Cooperative will beestimated by assuming a marketing margin relationship between the primary andwholesale/cooperative levels.4.1 Price Determination in the B.C. Lettuce MarketBefore proceeding to the discussion of demand and supply components, Figure6 provides an overview of a proposed short-run model of the B.C. lettuce industry. Animportant assumption of the model is that the California lettuce industry stronglyinfluences the price of the locally grown product. California is a major exporter of headlettuce to the rest of the U.S. and Canada, accounting for approximately 75 percent of allU.S. shipments in 198818 and over 95 percent of B.C. imports of head lettuce.18 Itis assumed that B.C. is a price-taker in the lettuce market since provincial consumptionand production is very small relative to California's production (total B.C. consumption isless than one percent of California production).In the short run, and we may think of this as a one week period, the quantity ofimported lettuce is fixed at the amount that wholesalers have purchased and is in transitto the B.C. market at this given time. The price that wholesalers pay for the importedlettuce, Pa, is exogenous and is also fixed in this short run period (in the sense that the18—-u.S.Departmentof Agriculture, Agricultural Marketing Service. Fresh fruits and vegetable shipments bycommodities, states, and months, 1989, FVAS-4 Calendar Year 1988.19AgricultureCanada. Agri-Food Development Branch. Horticulture and Special Crops Division. Annual unloadreport: fresh fruits and vegetables. Various issues.Q(Co)D(w)D(w, r)S(Coop)S(Coop)0.(c) Ci(c)^0(t)^Qfy(b) Output restriction by the CooperativeS(Coop)^PriceP(cif)P.(c)P(c)D(w)D(w, r)Q(t)^Qty(a) B.C. demand for al lettucePriceP(cif)P(c)54price of the lettuce in transit has already been negotiated).In Figure 6, panel (a), the wholesale demand curve for all lettuce is representedby the curve, D. Although not explicitly shown, this wholesale demand is derived fromthe retail demand for lettuce in B.C. Given the fixed quantity of imports, the Cooperativefaces the residual demand curve of wholesalers, labelled Dwr. The horizontal distancebetween the Dw and Dw j. curves represents the pre-committed quantity of importedlettuce purchased by the wholesalers, s'aca. In the usual trade model where B.C. is theprice taker, the quantity of lettuce the Cooperative would supply is given by theintersection of the Cooperative supply curve, Scoop, and the Po price line.Figure 6: Short-run model of price determination in the S.C. lettuce market55However, because the demand for domestic lettuce is a residual demand thequantity the Cooperative actually supplies, 0coop, is given by the intersection of the-Cooperative supply and the residual wholesale demand curves. We may think of therebeing an unobserved Cooperative supply curve that lies above the observed supplyschedule, Scoop. The way to look at this relationship is that the Cooperative must givethe wholesalers more lettuce at any given price to account for the lower quality, orconversely, take a lower price for the same quantity of lettuce. At the quantity, QCoop,the price the Cooperative receives is Pcoop, which is less than P. The differencebetween the cif price and the Cooperative's price may be decomposed into two parts.The distance between the Cooperative supply curve, Scoop, and the unobservedCooperative supply curve represents the quality discount, which is assumed here to beconstant at all quantities sold. The distance between the unobserved Cooperative supplycurve and the Ppif price line represents the discount to sell the Cooperative's availablesupply.In the short run with the price and quantity of imports fixed, it is possible that theCooperative could restrict its supply and move up the residual wholesale demand curve.In Figure 6, panel (b), the observed supply curve of the Cooperative has been pivotedleftward to S'oop. There is a decrease in the quantity of lettuce supplied by theCCooperative to acoop, for which they receive a higher price, P'coop. The total discountneeded to sell the available supply decreases, and the Cooperative selling price is closerto the cif price, Ppif• The quantity of imports remains the same at Oca, but the totalquantity of lettuce supplied to the B.C. market has decreased by the Cooperative's56amount.A key question in our analysis is to quantify the price changes due to outputrestriction and how revenues will be affected. The next two sections will describe supplyand demand models for head lettuce in B.C. Estimates of elasticities will be computedfrom the empirically estimated supply and demand equations. The elasticities will be usedto calibrate the supply and demand model to estimate the equilibrium price and quantityposition of the Cooperative. This equilibrium will then be used to compare alternativeequilibria under an assumption of no Cooperative market quotas or output restriction.4.2 Supply Model for B.C. LettuceIdeally a measure of the degree to which the Cooperative has altered supplyresponse would require a supply elasticity for the period before and after cooperativeformation. Due to data limitations only the post-Cooperative supply function can beestimated. Pre-Cooperative supply is hypothesized to be unrestricted and more elasticand plausible values will be assumed to estimate the supply curve in the absence of theCooperative.Estimation of the supply of lettuce at the farm level may be usefully approachedby describing a behavioral model of farmers' production decisions. This behavioral modelis useful for interpreting the coefficients of the empirical model and any restriction impliedby the functional form chosen. The following model includes the major factors orvariables related to output decisions:57Desired Output = F (Expected Price, Input Prices, Quota Effect, Expected Prices of Other Crops)The desired farm output of lettuce for a given year is a function of the expected price oflettuce, the price of production inputs, the expected price of alternative crops and themarket allocation quota.In the empirical model, the dependent variable is measured as the current year'sacreage planted to lettuce. Actual output sold may differ considerably from planned ordesired output because of environmental factors beyond the farmer's control. Thedivergence may also occur as a result of unfavourable market conditions, for example,low cost imports from California driving prices down. The area actually planted to lettuceis, to a much greater degree, under the farmer's control and, thus, a better proxy of forplanned output. As a proxy for planned output, acreage planted has two shortcomings.First, land is only one of many inputs in production and Behrman (1968) suggests thatan index of all inputs to be devoted to the crop is preferable. Most non-land inputs arenot unalterably committed to one crop at the beginning of the production season but havemany substitution possibilities in response to market or environmental factors. Hence themeasure of all inputs actually used on a crop may not correspond to the originallyplanned input use. Second, land is not a homogeneous factor. If land is sufficientlyheterogeneous and if there are other inputs that constrain production, it is possible thata farmer may increase planned output of a specific crop by devoting less but better landto that crop. In the case of lettuce growing farms in the Lower Mainland, the quality ofland is fairly homogeneous and acreage will be an appropriate proxy for planned output,given an expected yield based on a given fixed level of inputs.To show the relationship between planned production and planted acre, we canlook at their respective elasticities:ao* P aA P aY. PäP • Q* aP• A aP Y*E0p = E,,kp + Ey..pwhere Q* is the planned output, P is the price of Q, A the acreage planted to 0, and Y*the expected yield per unit of A, for a given level of input use. The elasticity of interestis own price or EQ.,p whereas the elasticity obtained through empirical estimation is EAR.From the above it can be seen that the smaller is (3Y*/(3P • P/Y*), the closer is EAR anapproximation for Ecr,p. In developed country agriculture there are few physical barriersto acquiring or using more inputs, therefore changes in expected yields in response tooutput price changes may not always be presumed to be zero. However, a case may beargued for assuming Ey*R to be close to zero in our model by noting that lettuce pricesmay fluctuate considerably from week to week. Thus an increase in output price at timet, when inputs can be effectively used, does not imply the same price will prevail at time(t+ 1) when the crop is harvested. Under these circumstances, a farmer's yieldexpectations may be constant making the elasticity of acreage equivalent to the elasticityof planned output with respect to price.Expected prices are formulated as a moving average of the price in the currentperiod, the price in the previous period and the price two periods past. The choice of athree period moving average is somewhat arbitrary. However, it is sensible to assumea finite period of carry-over effects. Prices of fresh vegetables tend to show considerableyear to year fluctuations and it would be unrealistic to assume farmers only consider last58(10)59year's prices when making output decisions for the current year. For example, highprices in the last period as a result of a poor crop due to inclement weather conditionswould not be viewed by farmers as anything but a one-period anomaly. Therefore,historical average prices are more appropriate. The current year's price is included in theexpected price because lettuce is planted on a continuous basis from early April to midAugust. Although there is no assurance that prices received in the early part of theseason will prevail, a farmer's planting decisions for the rest of the season are apt to beinfluenced by the observed trend of current prices.To incorporate the effect of competing crops in the farmers' decision-makingprocess, the price variable is expressed as the 3-year moving average price of headlettuce relative to the 3-year moving average price of leaf lettuce. The farms producinglettuce also produce other vegetables. Substitution possibilities among the various cropssuggests that a price index of all farm output in the Lower Mainland would also beappropriate. Data limitations preclude the construction of such an index therefore theprice of leaf lettuce alone will be used. Leaf lettuce is a good alternative crop sinceproduction techniques and type of inputs used are almost identical to those used in headlettuce production.The quota effect on output decisions is measured by a 3-year moving averageacreage. Since the Cooperative allocates market shares to members on the basis ofhistoric share of sales, current planned output levels are likely to be strongly influencedby past production. The 3-year moving average used by the Cooperative to calculatemembers' market share is used in this model to capture the effect of the market quota60allocation.The effect of input prices on desired output is measured by a share-weightedcomposite inputs price index. The index is composed of variable inputs whose level ofusage may conceivably be altered in the course of the growing season in response toprice signals.4.2.1 Estimation of Supply CurveThe behavioral model described in the previous section translates directly into thesingle equation empirical supply equation that will be estimated. Choice of data to proxythe variables from the behavioral model was based on the theoretical considerationsdiscussed and availability. The supply equation is estimated by:ACRES = a + 13 RPt + (3 2INPUTSt + 133MAVAs +^ (11)61where:ACREt^= current year's acreage planted to lettuceRPt^= 3-yr. moving average price of head lettuce relative to the 3-yr movingaverage price of leaf lettuce, where the years are t, t-1, and t-2INPUTS t = share weighted price index of the variable inputs labour, fertilizer, pesticideand non-mortgage interest of the current yearMAVAt = 3-yr. moving average acreage planted to head lettuce, where the years aret-1, t-2 and t-3The model hypothesizes that the acreage planted to head lettuce in period t is alinear function of the price of head lettuce relative to the price of an alternative crop, leaflettuce, the price of inputs and past lettuce acreage. The output price variable isexpected to be positively related to the dependent variable. Furthermore, the use of arelative index implies that as the price of head lettuce increases relative to the price of thesubstitute crop, then ceteris paribus, more acreage would be devoted to head lettuce.The inputs price index is expected to be inversely related to acreage. The measure of thequota effect (MAVA) should be positively correlated to current planted area. Assumingno major farm level structural changes it is expected farmers, at minimum, plan tomaintain production to historical levels and, possibly, shift out their supply curves.Estimation of a supply equation presupposes that firms engaged in production areassumed to obtain the largest possible profit given operational constraints. Marketdemand forces constrain the amount of product that may be sold (at a profitable price)and physical production requirements constrain the level of output that may be obtained.Under the profit maximization objective, firms supply output to the point where marginal62cost is equal to market price. The aggregate supply curve for the industry is thehorizontal summation of the portions of the individual firm marginal cost curves that lieabove minimum average variable cost. Hence the supply curve presumes that output isproduced by the least cost combination of inputs at any given output price.A major limitation of a linear supply curve is the implication of fixed marginal costs,in other words, the cost of production remains constant regardless of the output level.In many production systems this result is unrealistic, however by hypothesizing that therange of output levels under consideration is limited we may, with greater confidence,assume that input costs do remain constant within this range.The empirical model contains an explanatory variable (MAVA) constructed fromlagged values of the dependent variable. In our model, the MAVA variable is constructedas:MAVAt =E ACREt_tm(12)Although ACREt_i depends on pt_i and all previous disturbances, it should not becorrelated to the current error term pt. Assuming that the current error term is not seriallycorrelated, then the lagged regressor ACREt_i will also be uncorrelated with the error termpt thereby satisfying an important OLS assumption that the explanatory variable is notcorrelated to the error term.4.2.2 Supply Estimation Data63Primary data used to construct supply variables was taken from an annualpublication of horticultural statistics issued by the BCMAFF,23 Farm Inputs PriceIndex,21 and Agriculture Economic Statistics.22. The following is a more detaileddiscussion of how the variables are constructed:ACRE. In order to confine the supply estimation to producers belonging to theCooperative, the values of ACRE are the acres planted to lettuce for the Lower MainlandRegion only.RP. Crop prices collected by the BCMAFF represent selling prices (per lb.) towholesalers from the farm or by agencies. In the case of lettuce, the reported price is theCooperative's selling price. The farm gate price is calculated by subtracting theCooperative's charges which include administration fee, carton cost and vacuum coolingcost. A complete data set was not available for these charges. The administration feewas calculated from revenues reported in the Cooperative's financial statements. Cartonprices were not available for the period 1973 to 1987. To approximate this cost, a priceindex of corrugated boxes23 was used to compute a nominal price series. Data onvacuum cooling cost was also not available for the period 1973 to 1987. Approximatingthis cost is more problematic as there is no clear alternative 'product' for which time20B.C.Ministry of Agriculture, Fisheries and Food. Production of vegetable crops together with an estimateof farm value. Victoria, B.C. 1969 through 1988.21StatisticsCanada. Catalogue No. 62-004.22StatisticsCanada. Catalogue No. 21-603E.23StatisticsCanada. Industry selling prices. Catalogue No. 62-011.64series price data is available. However, anecdotal evidence suggests this cost tends tobe very stable, for example, from 1988 to the present the cooling cost has remained at$0.75 per carton of product. Therefore, the missing data was interpolated from theobserved endpoints. Since the vacuum cooling cost is only one component of totalCooperative charges, errors introduced by systematic patterns in the data should bediluted. The farm gate price of leaf lettuce was calculated in a similar manner. Theadministration fee was adjusted (down) to account for levies not applicable to leaflettuce24 and the carton cost was reduced to reflect its lighter construction. The relativeprice index is calculated as:RPt = 314 PUN (13)where PL is the price of head lettuce and PLL is the price of leaf lettuce. Using a relativeprice variable excludes other vegetable prices from the farmer's decision in planninglettuce production. Using other vegetable prices as explanatory variables assumesfarmers can switch freely to these other crops in response to price increases. For LowerMainland farmers this may not be the case since many other crops are also regulated(see Table 1) and require quota in order to deliver product to the designated cooperative.Leaf lettuce, however, is not regulated and farmers may grow as much as they think willbe profitable.INPUTS. The INPUTS index is constructed from the farm price indexes of four variableinputs: labour, fertilizer, pesticide and non-mortgage interest. The total index is simply the24Leaflettuce is not a regulated crop hence B.C. Vegetable Marketing Commission levies are not applicable.65sum of individual price indexes each weighted by its share of the total of these variableinputs. Shares are calculated from farm operating expense data for British Columbia.MAVA. A 3-year moving average of past acreage planted.4.2.3 Supply Estimation ResultsThe empirical model specified in Section 4.2.1 was estimated using the regressionpackage SHAZAM. The supply equation was estimated using a linear functional form.The results are presented in Table 4. The model was initially estimated using ordinaryleast squares. All coefficients showed the expected relation to the dependent variable andwere significant except for the relative price variable, RP. The adjusted R2 value indicatesthe model is a reasonable fit of the data explaining 80 percent of the variation in thedependent variable.Collinearity between the explanatory variables does not appear to be a seriousproblem as the correlation between regressors is less than 0.6 although with more thantwo explanatory variables, this is a less precise guide to detecting collinearity. However,auxiliary regressions indicate that the moving average acreage variable, MAVA is collinearwith the other explanatory variables. Since this variable was included to account for whatis believed to be an important factor in supply decisions, it would be incorrect to omit thisvariable. With a fairly high R 2 and significant t statistics multicollinearity does not pose aserious problem.Although the model is specified on the assumption of non-correlation between the66explanatory variable(s) and the stochastic disturbance term, the Durbin-Watson statistic(2.694) lies in the interval of indecision as to whether there is negative autocorrelation.In cases where the estimated d value lies in the indecisive zone, the modified d testprocedure can be used. Given that the critical value of du = 1.685, the modified d testindicates that the null hypothesis of RHO =0 should be rejected, that statistically there issignificant evidence of negative autocorrelation.Table 4: Econometric results of lettuce supply estimationn = 19^df = 15Dependent Variable: ACREOLS^ AUTOEstimated^Elasticity^Estimated^ElasticityVariable^Coefficient^at Means Coefficient^at MeansRP^ 0.165^0.018^0.179^0.020(1.401) (2.174)**INPUTS^ -0.006^-0.096^-0.006^-0.099(5.284)** (7.491)**MAVA^ 0.476^0.478^0.480^0.481(3.040)** (4.300)**CONSTANT^ 3.206^0.601^3.200^0.599(3.500)** (4.856)**RHO^ -0.383(1.808)Durbin-Watson^2.694^ 2.367R2 (adj.) 0.801 0.831Note:^ 1. t-statistics in parenthesis2. ** significant at 95%3. Mean value of dependent variable: 5.3384It is suspected that the source of autocorrelation probably lies with the averageacreage variable since values of ACRE are likely to be correlated to each other and henceto the disturbance term. To check whether excluded variables was causing67autocorrelation, a time trend was added to the regression. If the excluded variable affectsthe acreage planted, then the error term will reflect a systematic pattern. The new OLSregression had a Durbin-Watson statistic of 2.74 and, using the modified d test, the nullhypothesis of no negative autocorrelation is rejected. In addition, the time trend variableitself was not statistically significant from zero and the adjusted R2 suggests this modelgave a poorer fit of the data. It should be noted that adding a time trend to test forexcluded variables is limited to excluded variables that exhibit a linear pattern over time.With time series data heteroscedasticity is not believed to be a problem and nodiagnostics were carried out in this regard. On the basis on the modified OW test, themodel was corrected for serial correlation of the residuals. With the correction, regressionresults improve with all variables statistically significant at 95 percent. The recalculatedDW statistic still lies in the interval of indecision but the modified d test indicates the nullhypothesis of no negative autocorrelation is just barely rejected suggesting that theproblem is less serious.The elasticity with respect to price shows that for a 10 percent increase in theprice of head lettuce relative to the price of leaf lettuce (where all prices are movingaverages), acreage is increased by a mere 0.2 percent. Such a low elasticity measureimplies that output, as measured by acreage, is not at all responsive to price. Inagricultural sectors where physical factors are limiting, price insensitivity is more probablesince there may be no viable alternative crops. However, this is not likely to be the casein the Lower Mainland where a wide range of crops can be grown and substitutionpossibilities high. Output response may also be limited by the level of management and68technical expertise of the farmer and/or rigidities in market structure. In an 'orderly'marketing scheme as for head lettuce, market shares are allocated on the basis ofhistorical sales. Thus, unless farmers believe with great certainty that demand willincrease substantially, they will use past sales as the main guide to planning output forthe next period. The estimated coefficient of the quota variable (MAVA) is significant andindicates past sales do indeed have a strong effect on output. Furthermore, since futuremarket shares are calculated on the basis of current sales, farmers may be reluctant tocut back on production in response to what may be perceived as a one-period decreasein price. It should be noted that many ,of the other crops grown by Cooperative membersare subject to the same market sharing arrangement and increasing output of lettuce mayjeopardize future market allocations of other crops. In the long run, of course, farmerscould acquire more land if they believe price increases and/or demand shifts to bepermanent. As a result of the market quotas, response in acreage with respect to pricemay be highly inelastic.Other proxies for the price variable were tested, including the 3-year movingaverage price of lettuce deflated by the Farm Products Price Index (FPPI) for B.C. Thisdid not yield comparable results possibly because the FPPI covers a wide range ofagricultural products including livestock, dairy, fruit and grains. A regression was alsoestimated with the 3-year moving average prices of leaf lettuce and head lettuceseparately, where both set of prices were deflated by the FPPI. The leaf lettuce variablewas of the expected sign but with an insignificant t statistic.Individual input price indexes were used in an early version of the model. The69estimated coefficients were of the correct sign but with insignificant t statistics. Since theoverall test of significance (F test) indicated that the four inputs did have a influence onACRE, a composite index was used to model this effect.The INPUTS variable is highly significant and shows that for a 10 percent increasein the price of inputs there is a one percent decrease in output (acreage planted) implyingan inelastic response. There are no guidelines, a priori, as to an expected elasticitymeasure. If the inputs included in the variable represent some minimum threshold levelof inputs needed to produce the output then very little response could be expected withrespect to price increases. On the other hand, lettuce is continuously planted throughthe season so there is scope for acreage response to lower input prices. In general,input prices are fairly stable (i.e. relative to output prices) and are often known ornegotiated at the beginning of the production season.It was hypothesized that past acreage would have a positive influence on currentacreage and this is borne out by the estimated coefficient which is both of the expectedsign and significant. For a 10 percent increase in the 3 year average acreage there willbe approximately 5 percent increase in current acreage. Reinterpreting this variable asthe market quota allocation effect, if farmers strictly adhered to past market allocationsin planning output (i.e. quotas are very important in determining current output) thenacreage in the current period would exactly equal the 3 year average acreage, and wewould expect the coefficient of the MAVA variable to be unity as well as the elasticityvalue. On the other hand, if past quotas have no effect on current output decisions thenthis coefficient would be close to zero. The estimated MAVA coefficient lies midway70between the two extremes and suggests past quota allocations may have an impact inrestraining current output. In fact, our results show that past quota allocations have agreater impact on current output than the price variable.The conclusions regarding the quota effect are made on the assumption that thedata for the ACRE variable has been correctly measured. It is possible that the actualacres planted may be under-represented due to reporting errors. Reported data equatesacres harvested with acres planted and ideally we would like to have acres planted as thedependent variable and average acres harvested as an explanatory variable.4.2.2 Validation of Supply EstimationOverall the results of the regression equation indicates a fair fit of the data to theregression line. The model (corrected for autocorrelation) predicts 83 percent of thevariation in the dependent variable, with all explanatory variables of the expected sign.To examine the validity of the estimated model, we first plot the observed andpredicted values (Figure 7). The observed values show considerable fluctuation from1979 to 1981.25 Although the model captures some of the turning points it does notpick up these large fluctuations, however the predicted values do follow the trend of theobserved values during other periods when fluctuations are small. It is possible that overa larger sample the fit would improve.A second test of the predictive ability of the model is to look at the deviations of25Thelow acreage in 1981 was due, in the most part to heavy rains during early summer which flooded cropsthat would normally be harvested in August.71the predicted values from observed values, where the smaller the deviations the better isthe fit of the model. The root mean square error (RMSE) value calculated is 0.193, orexpressed as a percentage of the mean observed dependent value, is 3.62 percent. Inother words, over the sample population the model predictions deviate by less than 4percent from the observed mean value.Figure 7: Observed and predicted values from econometric estimation of supplyAs a final test of the model, the supply equation was regressed omitting the lastobservation to provide a forecast. The predicted value is 5.1226 (512.26 acres) with acalculated residual of -0.0266. This deviation of the predicted from the final observedvalue represents less than one-half percent deviation from the mean value in this sample.These results suggest that the predictive ability of the model is acceptably accurate andwe conclude it captures the major elements of the hypothesized behavioral model offarmer's output planning process.724.3^Demand Model for Lettuce in B.C.4.3.1 Estimation of Demand CurveThe demand for lettuce will be estimated as a single linear equation.26 Thecommodity lettuce would appear to satisfy the conditions that allow our assumption ofzero bias due to omitted variables as it represents a small proportion of expenditures onfresh vegetables and has few good substitutes. The quantity of lettuce is taken to be afunction of its own price and income. The inclusion of substitutes or complements wasconsidered but no single or group of fresh vegetables appeared to be appropriate. Theselection problem is more acute in vegetables than in other food commodities as it isoften difficult to determine if individual vegetables are substitutes or complements for eachother. For example, head lettuce is often consumed in salads but may be served with orwithout other fresh vegetables. Other leafy type lettuces are believed to be weaksubstitutes for head lettuce which is consumed for its unique crunchy texture. A timetrend is included to account for secular changes in demand. The demand equation isspecified as:PCOLt = per capita consumption of head lettuce in period t (lb.)26Estimatinga demand equation presupposes the maximization of a representative consumer's utility functionsubject to the consumer's budget constraint. The first order conditions for a maximum yields a set of partialderivatives of the utility function with respect to each consumer commodity. These are solved to obtain a setof demand equations where the quantity demanded of a commodity demanded of a commodity is expressed in termsof its own price, other commodity prices and income. To satisfy the budget constraint, the sum of expenditureson all commodities must be equal to (or less than) total income (see Deaton and Muellbauer for discussion ofrestrictions on the set of demand equations). Invoking separability permits the estimation of subsystems ofdemand equations. In estimating single linear equation, omitted variables are assumed to be zero. The moreclosely related the omitted commodities are to the one of interest the larger the bias on the includedcommodities.73PCOLt = a + 1 MPL86t + P2PCDIt + 1334 + 11(^ (14)MPL86t = price index of lettucePCDIt = per capita disposal incomeTt^= time trend, from 1 to 23Since our objective is to estimate market equilibrium in the period when localproduction is available, the time frame for the demand equation is limited to coincide thiswith period. The period from June to September represents the local season, althoughsupplies are thin in the first week or so of June and limited supplies are still available inthe early part of October. The delineation of this time period was necessitated by datalimitations and by the need to approximate as closely as possible the period during whichtariffs apply on imported lettuce.The estimated coefficient of lettuce price is expected to be negative. Since lettuceis assumed to be a normal good, the estimated income coefficient is expected to bepositive. A priori statements about trend variables are difficult to make since they often,as in the present case, may be accounting for other non-specified explanatory variables.The usual application of time trend is to mark shifts in demand over time due to changein tastes, and may be of either sign. In addition the trend variable could pick up theeffects of other variables. If an omitted variable is a substitute, then the trend would beexpected to have a negative coefficient.4.3.2 Demand Estimation Data74Data for the demand model was taken from provincial reports of productionstatistics,27 Consumer Prices and Price Indexes,28 unloads reports,29 and ProvincialEconomic Accounts. Annual Estimates.30 The following section describes the variablesused. All per capita values were obtained by dividing by the estimated B.C. populationfor the respective year.PCQL. The quantity of lettuce consumed in B.C. for the period June to September istaken to be imports during this period and the total annual local production. There aresome exports of B.C. grown lettuce to other provinces but the quantities are small relativeto total production. The imports data are taken from weekly unloads reports compiledby Agriculture Canada based on surveys at wholesale points in Vancouver. Since thereports only account for the imports made by the wholesalers contacted, the unloadsunder-state imports.31 A monthly provincial import data series was not available for theyears considered. Despite these problems, it is believed the B.C. per capita consumptionas calculated for the period is more appropriate than national consumption figures.MPL86. The price variable is approximated by the national price index for lettuce.Monthly indexes are available and this variable represents the median value for the four27B.C.Ministry of Agriculture, Food and Fisheries. Production of vegetable crops together with an estimateof farm value. Victoria, B.C. 1969 through 1990.28StatisticsCanada. Catalogue 62-010.2 9AgricultureCanada. Agri-Food Development Branch. Fruit, vegetable and honey crop and market report.Ottawa.38StatisticsCanada. Catalogue No. 13-213.31Adiscussion with Agriculture Canada representative indicates they are aware of the under-reporting andestimate the published figures to be approximately 20 percent less than actual imports. They did not offer anexplanation for the chronic discrepancy.75month period under consideration. Average retail price data collected for variousCanadian cities is available for July, however prices tend to be at their lowest during thismonth, therefore a median national 'summer' price is likely to be more representative.PCDI. Provincial monthly per capital disposable income is not available and this variableis constructed by simply dividing annual income by three, implicitly assuming the incomedoes not vary seasonally. The PCDI value was deflated by the median value of theVancouver consumer price index for the four month period June to September.TIME (T). Time trend variable which increases by increments of 1 for each successiveobservation.4.3.3 Demand Estimation ResultsDemand parameters were estimated using ordinary least squares technique. Apriori there is no theoretical basis to specify the functional form of the regressionequation. Double log forms are often used in demand studies but the selection is justifiedby a better fit of the data rather that forehand knowledge. A double log form was triedbut rejected in favour of the simple linear form because of the implication of constantelasticity associated with the log form. As with most perishable produce, the price ofhead lettuce may fluctuate considerably over short time periods, or from a given timeperiod in one year to the same time period in the next year. Using a log form presumesthe own price elasticity would remain constant despite large changes in price. In a linearrelationship, the price level effect states that as prices increase, so does the absolute76value of own price elasticity. If the range of prices under consideration was small, aconstant elasticity assumption is acceptable, but this is not the case with head lettuce.The regression results are summarized in Table 5. All estimated coefficients showthe expected relationship with the dependent variable and are statistically significant. Theadjusted R2 value of 0.67 suggests a satisfactory fit of the data considering the problemsin specifying appropriate variables.Table 5: Econometric results of lettuce demand estimationn = 23^ df = 19Dependent variable:^PCOLEstimated^ ElasticityVariable^Coefficient at MeansMPL86^ -0.314^ -0.260(1.750)*PCDI^ 0.002^1.129(3.037)**T^ -0.179^ -0.331(2.060)**Constant^ 3.259^0.463(1.657)RHODurbin-Watson2R (adj.)0.1221.5930.674Note:^1. t statistics in brackets2. ** significant at 95%3. * significant at 90%4. Mean of the dependent variable: 7.033A common problem in demand estimation is that price and income data seriestend to be correlated to each other. It is natural that prices in an economy will move77together and there is often no satisfactory solution to the collinearity problem. In seriouscases, the simplest means to correct the problem is to omit the variable causing thecollinearity. However this introduces a new problem of specification bias. The correlationmatrix of variables indicates high levels of correlation between regressors. The R2 valuesobtained from auxiliary regressions of each explanatory variable on the other confirmshigh levels of collinearity. The time trend variable is often a source of collinearity and aregression was run omitting the time trend. The pairwise correlation between the tworemaining regressors is in excess of 0.8 indicating the multicollinearity problem remains.The practical consequences of multicollinearity are large variances and standard errorsof the estimators, however estimates remain unbiased and consistent. Since the standarderrors in the model are small and all t statistics are significant no remedial measures forcollinearity are carried out.Graphical analysis of calculated residuals did not indicate the presence ofcorrelation between the disturbance terms and other explanatory variables. Thecalculated Durbin-Watson (DVV) statistic is 1.59 and given the critical DW values with k' =3and n = 22, lies in the indecisive zone as to presence of positive autocorrelation. Themodified DW test rejects the null hypothesis of no positive autocorrelation by a very smallmargin. On the basis of the residuals plot and the modified DW test, the problem ofautocorrelation is assumed to be negligible and no remedial measures were carried out.Demand elasticities as calculated at the mean value of the dependent variable arepresented in Table 6 together with elasticities from other studies. Huang's (1985) resultsare estimated using a large scale demand system utilizing a constrained maximum78likelihood method with parameter restrictions derived from classical demand theory.Hammig and Mittelhammer (1980) utilized OLS and 2SLS to estimate a model of supplyand demand in the U.S. lettuce industry. The model estimated by Curtin et al (1987) mostresembles the author's model but included price of tomatoes, potatoes as well as anindex of other fresh vegetables. The latter two studies also imposed some parameterrestrictions.Table 6:^Demand elasticities^for^lettuceThis Study HuangCurtinet alHammig &MittelhammerPeriodOwn PriceIncome1969-1991(annual)-0.261.131953-1983(annual)-0.140.231965-1984(annual)-0.170.251954-1977(annual)-0.120.19Sources:Curtin, L., Theoret, D. and Zafiriou, M. Demand for foods in Canada: Recent estimates.Agriculture Canada. Policy Branch. Working Paper 13/87. December, 1987.Hammig, M.D. and Mittelhammer, R.C. "An imperfectly competitive market model of the U.S.lettuce industry". Western Journal of Agricultural Economics (5) 1980: 1-12.Huang, K.S. U.S. demand for food: A complete system of price and income effects.  U.S.D.A.Economic Research Service. Technical Bulletin No. 1714. Washington, DC: December 1985.The calculated own price elasticity is larger (in absolute terms) than elasticitiesfrom other studies, but like the others, lies in the inelastic range. Few true substitutesexist for head lettuce and, furthermore, lettuce may be regarded as a near staplecommodity by consumers. The larger elasticity from our model may be a result ofrestricting the sample to account for summer consumption only. Increased availability ofother fresh vegetables could account for slightly more elastic demand in the summer.79The calculated elasticity with respect to income from our model is much largerthan the other income elasticities cited. Income elasticities for basic food items tend tobe low in value, rarely being greater than unity except for highly prepared and luxuryfoods. Therefore, this figure must be interpreted with some caution.There is no theoretical basis to predict either the sign or magnitude of the timetrend variable. Growing consumer awareness of dietary needs would suggest that thedemand curve would be shifting rightward over the period of estimation. However,increasing choices available in produce markets and more 'sophisticated' taste may bedecreasing absolute consumption levels. The estimated coefficient is negative implyingan overall reduction in head lettuce consumption. The time variable will also be pickingup the effects of omitted explanatory variables that are related to the dependent variable.In this case, the time variable may be acting as a proxy for the price index of theunknown bundle of other fresh vegetables/foods which influence lettuce consumption.4.3.4 Validation of Demand ModelAlthough the own price elasticity obtained from our model compares favourablywith other results, the overall fit of the data is modest with only 67 percent of the variationin consumption explained by the regressors. However, the joint F test statistic rejects thenull hypothesis that all explanatory variables are equal to zero and all coefficients wereof the expected sign and all were significant.The first measure of model validation is a graph of the observed and predictedObserved Quantity -4*-- Predicted Quantity80values. Figure 8 shows that while the predicted values follow the general trend ofobserved consumption, it smooths out the observed fluctuations. A priori, demandquantities of staple food commodities are expected to be rather stable over time. Dataerrors may account for the large observed fluctuations making prediction less precise.Figure 8: Observed and predicted values from econometric estimation of demandThe second measure is to examine the root mean square error (RMSE) termcalculated from the deviations of the observed and predicted values. The calculatedRMSE value is 1.241. Dividing the RMSE by the mean of the dependent variable givesa measure that expresses the deviations of predicted from the mean as a percentagevalue. The calculated value is high, indicating the predicted values deviate by asubstantial 17 percent from the observed mean.To test the model's predictive ability, the demand equation was regressed omittingthe last observation to obtain a forecast value of 4.574 for that observation with a residual81of -0.451. The deviation of the forecasted value from the observed mean is 2.194 percentand indicates a better prediction ability. The problem with the model's predictive abilityappears to lie in the unusually high values observed in 1973 and the unusually low valuesin 1981. The latter value result from the exceptionally low output level due to poorweather in that year.Although the predictive ability of the demand model is moderate, it is possible thatthis is a sample phenomenon and that over a larger sample the predicted values willapproach the true ones. The results from the model are therefore considered acceptablefor calculating derived elasticities.825.0 ANALYSIS OF A COOPERATIVE WITH SUPPLY CONTROLIn this section the elasticities obtained from the empirical estimation of supply anddemand will be used to calibrate a model of the B.C. lettuce market. The model is thenused to show changes in total revenues by assuming an unrestricted supply curve thatis hypothesized to exist in the absence of the market sharing quotas operated by theCloverdale Lettuce and Vegetable Cooperative. Since the empirical estimation is basedon retail and farm level data, the first part of this chapter will discuss the estimation ofderived elasticities from primary level elasticities. This is followed by a discussion of thecalibration of the supply and demand model using the derived elasticities. Both supplyand demand will be modelled as linear curves. The third part reviews the implications ofa cooperative with output control and simulates the effects of a marketing environmentwithout such controls. This will be done by assuming an unrestricted farm supply curvewithin the calibrated model to look at changes in producer/member's welfare. Shiftingor pivoting the supply curve will yield new supply elasticities. The final part of the chapterwill provide a summary and conclusion of the analysis together with recommendations forfurther research.5.1^Derived Supply and Demand ElasticitiesThe own price elasticities calculated in the previous chapter represent supply anddemand response at the farm gate and consumer levels respectively. For our purposes83the elasticities of interest are derived supply and derived demand elasticises or those atthe Cooperative level with respect to supply and the wholesale level with respect todemand. The exact relationships between the two levels depends on how the primaryand derived curves are related. Since the two are separated by a schedule of marketingmargins, the relationship depends on how the marketing margin behaves. Tomek andRobinson (1990) discuss three possibilities.The first is where the marketing margin is a combination of absolute andpercentage mark-ups which in linear may be written:M = c + aPP M = margin between primary and wholesaleP = price at primary levelThe mark-up structure implies that the constant portion (c) represents fixed costs whilethe percentage portion (a) represents marginal costs of marketing each unit of output.The elasticity relation between the primary and retail level is given by:Ed = Ed [1^(1-a)Pd i^(15)where:Ed^= derived level elasticityp^= primary level elasticityc^= constant,a^= proportion; Ca < 1P^= primary level priceA second, and the simplest possibility in specifying a margin relationship is thecase of a fixed absolute amount, where the margin is a constant regardless of the amountmarketed. The primary and derived curves are parallel separated by a constant amount.84This specification is often not realistic but appears to fit the behaviour of the CloverdaleLettuce and Vegetable Cooperative with respect to its mark-up above farm level prices.As discussed in Chapter 2, the Cooperative sets agency charges for marketing servicesannually.32 Furthermore, these charges are fixed to all members regardless of the levelof output. Therefore, in our analysis a fixed margin will be assumed to characterize therelationship between primary (farm) level and Cooperative (derived) level supply. Therelationship between elasticities at the two levels is given by:(16)where Pd is the Cooperative (derived) level price and other notation as noted above.Given an empirically estimated primary elasticity value, it is possible to obtain the derivedlevel elasticity. Since prices at the Cooperative level will always be higher than those atthe farm level, the price ratio (Pd/Pp) is greater than one. Thus supply at the Cooperativelevel will be more elastic than supply at the farm level.The third alternative is a fixed percentage marketing margin where the derivedlevel prices are related to primary level prices by:Pd = aPpwhere Ck a <1. With the constant value (c) equal to zero, the right-hand side of theidentity (8) reduces to one. Assuming a constant percentage margin regardless of the32Asdiscussed in Section 2.2, capital investments are financed through the service charges collected frommembers. Any surplus remaining at the end of the year after all outstanding receivables have been met isdistributed back to members on the basis on their patronage. The surplus may be due to lower input pricesnegotiated after the service charge has been set, larger volumes which will reduce the fixed cost component ofthe service charge or other reasons. Patronage refunds represent a reduction in the service charge paid bymembers on every unit of product they market through the Cooperative. Information on the frequency and amountof patronage refunds is not available. Omitting patronage refunds from our calculations means that we mayoverstate the Cooperative margin in some years.85quantity marketed, the price elasticities at the two market levels will be the same for agiven quantity marketed, or (Ed = E p). Primarily due to a lack of data about marketingcosts between the retail and wholesale levels, a constant percentage margin will beassumed on the demand side. One consequence of constant percentage margins is thatas prices increase, the gap or margin between wholesale and retail prices increase. Anexamination of a graph of real retail and wholesale prices together with the marginprovides some evidence for constant percentage margins. Figure 9 shows that along withthe decline in the real retail price of lettuce, margins have also fallen while at the sametime, the real wholesale price has remained stable. Thus the gap between retail andwholesale prices has indeed narrowed as retail prices have decreased.Figure 9: Relationship between real wholesale prices, real retail prices and marginsSource: B.C.M.A.F.F. Production of vegetable crops together with an estimate of farm values.Statistics Canada. Consumer prices and price indexes.865.2^Derived Supply and Demand CurvesIn this section a model of the B.C. lettuce industry will be calibrated for the year1990 using linear supply and demand curves calculated using the estimated elasticities.On the demand side, the calibration of the wholesale demand for local lettucebegins with the consumer demand for all lettuce. In Figure 10, panel (b), the curve Dr isthe consumer demand for lettuce, where Dr is calculated using the empirically estimateddemand elasticity, the observed retail consumer price and observed consumption duringthe 1990 production season. Consumption is assumed to consist of domestic output andimports. Dr has an elasticity of -0.26 through the point (17.66 M.Ib.; $0.59).Also in panel (b) is the estimated wholesale demand for all lettuce, Dw, calculatedby applying a proportional margin to consumer demand. The proportion is the ratio ofthe wholesale demand price ($0.22) to the retail price ($0.59) as discussed above. Usinga proportional margin, the wholesale demand also has an elasticity of -0.26 through thepoint (17.66 M.Ib.; $0.22). This is shown as point B in the figure.Having established the demand for all lettuce in panel (b), the residual wholesaledemand for local lettuce, Dw,r, is derived by subtracting imports from total wholesaledemand. In Figure 10, panel (a), the residual wholesale demand curve for local lettuce,Dwr lies below and is parallel to the total wholesale demand curve, D. The distancebetween the two curves represents the fixed quantity of imported lettuce. The residualwholesale demand curve for local lettuce is known to pass through the quantity 11.8 M.Ib.(the domestic production) at a price of $0.22 (the Cooperative selling price). Using this(^11.8(a) B.C. demand for load lettuce----- 0(w)^ D(w, r)(Mb.)Output17.663(M.b.)Outputoll lettuceS(Coop)S(form)(11.8;0.22)PriceP(cif)P(c)Q(Ca)=5.863(b) B.C. demand for87point and the slope value from the equation of total wholesale demand, DIA,,, is calculatedto have an intercept at the price $0.78 and an elasticity of -0.39 at the point (11.8 M.Ib.;$0.22).Figure 10: Model of B.C. lettuce market, 1990On the output side, the farm level supply curve has an assumed elasticity of 0.020at the 1990 production level of 11.8 M.Ib. and the 1990 average farm price of $0.15 perlb. The absolute constant margin between Cooperative and farm level supply price iscalculated as:Administration FeeCarton CostVacuum Cooling Cost$0.81 / carton1.45 / carton0.75 / carton$3.01 / carton88Based on an average of 43 lb. per carton, the margin of $0.07 per lb. yields aCooperative supply price of $0.22 per lb. Using the elasticity relationship set out in theprevious section the Cooperative supply curve has an estimated elasticity of 0.029 at anoutput quantity of 11.8 M.Ib. at a price of $0.22 per lb., i.e. the point where theCooperative supply curve intersects the residual wholesale demand curve of local lettuce.Prices received by member-farmers are determined by the farm level supply curve, which,as noted is $0.15 per lb. at a quantity of 11.8 M.Ib. The Cooperative and farm supplycurves are labelled Sc and Sf respectively in Figure 6, panel (a). The two curves areparallel, separated at every output level by the Cooperative's fixed marketing margin.5.3^Model SimulationA working hypothesis of this paper is that cooperative marketing institutionsprovide positive benefits to its members. With respect to the Cloverdale Lettuce andVegetable Cooperative, the issue of particular interest is whether this single desk sellingcooperative, using a market allocation scheme, increases its member's welfare (or totalrevenues) relative to an unregulated marketing environment.The underpinnings of the Cooperative's welfare improving role for members89depends on its ability to manipulate or restrict output by members in order to move upthe market demand curve, which in this instance is the residual wholesale demand. Theextreme case of output restriction is, of course, the monopolist seller who sets outputequal to marginal revenue to maximize profits. This result assumes the monopolistoperates on the elastic portion of the demand curve. In an analogous manner thebenefits of the more limited output control of the Cooperative also depend on thewholesale demand elasticity. With an elastic demand, increasing output will cause priceto decline by a smaller amount thereby increasing revenues. With an inelastic demandthe consequences of increased output are not quite so salutary since price will fall by agreater amount causing total revenues to decrease. The more inelastic is demand thelarger are the price decreases relative to output increases.Given that the estimated demand curve for lettuce is inelastic, the purpose ofoutput controls by the Cooperative must be seen as an effort to maximize revenues bypreventing the large price decreases that would accompany a relatively small increase inoutput supply. It is hypothesized and supported by the estimated supply equation thatthe market quota allocation scheme used by the Cooperative has been a constraint onfarm output response and that the supply curve is more inelastic than it would be in theabsence of the constraining market quotas. Consistent with this hypothesis the resultsof the empirical estimation indicates that supply is indeed quite inelastic with respect toprice.To assess changes in revenue due to the Cooperative's output control scheme,ideally we should have an estimate of supply elasticity both before and after formation of90the Cooperative. A pre-Cooperative supply elasticity is not available therefore it will beassumed that pre-Cooperative supply curve is more price responsive. For the simulationmodel we want to model a market supply to reflect the higher output levels assumed tobe the case when market quotas did not exist. To simulate increased output theestimated supply curve will be pivoted rightward by adjusting its slope by discreteincrements while keeping the intercept term constant. The flatter supply curve yields anew intersection with the residual wholesale demand curve at a higher output level anda lower price. In our simulation, the precise amount of discount from the cif price will notbe explicitly addressed since data on the price wholesalers pay for the imported lettuceis not available. Only the change from the observed Cooperative price is modelled.Higher output levels may also be simulated by shifting the supply curve rightward,however this assumes the new supply curve has the same slope and that the error termis the same at all levels of production. It should be noted that a more elastic supply maybe simulated by rotating the observed supply curve through the observed equilibriumprice and quantity point. This would give a more elastic supply at the observed point butwill not show any impact on output in the current period. A multi-period simulation wouldbe required, however this analysis will not be undertaken in this study.The post-Cooperative supply curve is conjectured to be more inelastic althoughthe extent of the change is unknown. The calculated elasticity of supply through thesimulated price and quantity points will differ from the estimated supply elasticity since thelatter is associated with a different point. Since supply will be more inelastic at lowerprices (a consequence of assuming linear curves), the actual supply elasticity values at91the simulated points will be smaller. The large price decreases which result from a highlyinelastic supply curve overwhelms the effect of the increase in the slope coefficient.The corresponding derived supply curve may be viewed as the supply curve ofthe Cooperative, or any other type of marketing organization with no output restrictions.The marketing cost structure is assumed to be identical to the current structure. In actualfact, a non-cooperative marketing firm may very well have a different mark-up structure.Retaining an identical mark-up is a simplifying assumption which allows a more readycomparison with the current Cooperative marketing environment. Demand is assumedto be the same whether the Cooperative restricts output or not.Figure 11 illustrates the conceptual model of supply and demand under alternativesupply assumptions. The demand curve (Ow) is wholesale level residual demand forlocal lettuce, while the Cooperative and farm level supply curves are labelled Sc and Sfrespectively. The alternative, more elastic supply curves are labelled Sc, and Sr, whereSe, is the derived supply of any marketing firm with no output controls. As drawn,demand is inelastic. The intersection of Sc, with the demand curve yields a new marketequilibrium with an increase in output from Q to Q' and a large decrease in price from Pcto Pc'.92Figure 11: Model of supply and demand under alternative supply assumptionsTable 7 presents the changes in total revenues from incremental changes in theslope of the estimated farm supply curve that result in a flatter curve. At all price levels,the assumed supply is greater than that given by the base case observed supply. Asexpected, given the inelastic demand the new equilibrium points yield price decreases thatgreatly exceed the quantity increases. As a result total revenues decline with successivelymore elastic supply curves. At the farm level, with a one percent increase in the slopeof farm supply, the new equilibrium output is 0.94 percent larger while the new equilibriumprice has decreased by 3.5 percent resulting in a decline in farm revenues of 2.3 percent.93Table 7: Change in total revenues under alternative supply curve assumptionsWholesale demand elasticity: -0.26Change % % %in New Change Farm Change Coop. Change Farm ChangeSlope Slope Output in Price in Farm Price in Coop. Revenue in FarmValue Value (M.lb.) Output ($/lb.) Price ($/lb.) Price ($M) RevenueBase Supply0.6495 11.800 0.150 0.220 1.770Alternative Supply1.0% 0.6430 11.911 0.940 0.145 - 3.533 0.215 - 2.409 1.724 - 2.2622.0% 0.6365 12.024 1.898 0.139 - 7.133 0.209 - 4.864 1.675 - 5.3705.0% 0.6170 12.376 4.885 0.122 -18.356 0.192 -12.515 1.516 -14.3687.0% 0.6041 12.623 6.975 0.111 -26.210 0.181 -17.870 1.397 -21.06310.0% 0.5846 13.012 10.272 0.092 -38.597 0.162 -26.316 1.198 -32.290As discussed, the changes in total revenues depend as well on the behaviour ofthe wholesale demand curve. It may be recalled that a more elastic demand wasestimated in this study than those from other studies (see Table 6). On the basis onthese other elasticities, a value of -0.14 was used to derive an alternative demandequation to calculate changes in total revenues under an assumption of even moreinelastic demand. Alternative supply assumptions are as before. Table 8 shows theequilibrium quantities and prices, and changes in total revenues using the new demandelasticity. As expected, a more inelastic demand results in larger decreases in price andsmaller increases in quantity and consequently, larger declines in total revenue. The onepercent increase in the supply curve slope value now increases output by a mere 0.88percent while farm price and farm revenue fall by 6.2 and 5.4 percent respectively.94Table 8: Change in total revenues under alternative supply curve assumptionsWholesale demand elasticity: -0.14Changein New^ Change^Farm^Change^Coop.^Change Farm ChangeSlope Slope Output^in^Price in Farm^Price in Coop. Revenue in FarmValue Value^(M.lb.)^Output^(S/lb.)^Price (S/lb.)^Price (SM) RevenueBase Supply0.6495 11.800 0.150 0.220 1.770Alternative Supply1.0%^0.6430 11.905 0.887 0.141 - 6.210 0.211 - 4.234 1.675 - 5.3782.0% 0.6365 12.011 1.791 0.131 -12.513 0.201 - 8.544 1.576 -10.9665.0%^0.6170 12.343 4.600 0.102 -32.194 0.172 -21.950 1.255 -14.3687.0% 0.6041 12.574 6.561 0.081 -45.916 0.151 -31.306 1.020 -42.36810.0%^0.5846 12.938 9.643 0.049 -67.492 0.119 -46.017 .631 -65.357The results from the simulation of alternative supply and demand conditionssuggests that the output restriction by the Cloverdale Lettuce and Vegetable Cooperativehas been beneficial to its members. This conclusion hinges on our assumption that theCooperative, by controlling market output, has constrained supply response on the partof its members resulting in lower output levels than would be the case in the absence ofany market quotas. Given the inelastic nature of demand, farm prices fall dramatically forrelatively small increases in output. Therefore, the Cooperative's efforts to control supplyand members' willingness to comply with these market supply controls would appear tobe entirely rational.It is possible that supply response for lettuce would be unchanged even if nocontrols on output existed. The farms which grow lettuce also grow a variety of otherfresh market vegetables. Physically, substitution possibilities amongst the differentvegetables exist. However, many of these other crops are subject to the same marketoutput regulations that apply to lettuce. Hence rigidities in production may arise as95producers attempt to maintain market allocations for a number of crops. In the presenceof output price volatility, production decisions may be motivated by protecting thesemarket allocations.The possibility that producers faced monopsony power before the formation of theCooperative was discussed in Section 3.5. To show the effect of monopsony power weneed to calculate the marginal outlay (MO) curve facing the monopsonist wholesaler. Thequantity demanded by the monopsonist buyer is determined by the intersection of the MOcurve with the demand curve. The price is then determined according to the farmer'ssupply curve. Because our estimated supply curve is highly inelastic the intersection ofthe calculated MO curve and the estimated wholesale demand curve yields a negativesupply price. It is possible that even with the Cooperative wholesalers are able toexercise some residual monopsony power. In other words, the observed supply quantityand price is not on the supply curve where it intersects the wholesale demand curve butbelow that point. This could explain the negative supply price in our simulatedmonopsony scenario.The final simulation undertaken in our analysis is to examine the changes in totalrevenue from the sale of head lettuce when the quantity of lettuce imported is increased.As in the earlier analysis this increased quantity is assumed fixed at the amount in transitin the current period. In terms of Figure 6, an increase in the imported quantity has theeffect of shifting the residual wholesale demand curve for local lettuce in panel (a)downward. Recall that the horizontal distance between this curve and the total wholesaledemand represents the quantity of imported lettuce. An increase in the quantity of96imports, given that total demand remains unchanged will increase the horizontal distanceand shift the residual wholesale demand curve downward. The Cooperative and farmsupply curves are assumed to remain unchanged.For the simulation the quantity of imported head lettuce is increased by smallincrements to show the change in domestic product price, output quantity and farmrevenue. Table 9 shows the results of the simulation as well as the percent change fromthe base case when the import quantity is 5.863 M.Ib. (1990).A one percent increase in the quantity of imported lettuce leads to a very smalldecrease in the quantity of domestic output (less than one-half percent). However, sincethe domestic supply curve is highly inelastic this small change in quantity results in a 1.2percent decline• in farm price. Total revenues from the sale of head lettuce fall by 1.7percent due to the lower price and reduction in output. From the table it can be seen thatthe decline in farm price rather than the reduction in output is the source of the largedecreases in revenues.The results from this simulation suggests that in the short-run domestic producersare at considerable risk as a result of increases in imported lettuce. This may have beenthe situation for domestic producers during 1991 and early 1992 seasons when headlettuce imports increased. Given an almost vertical supply curve, the downward shift inthe wholesale demand for local lettuce would cause domestic prices to decline rapidly.97Table 9:ChangeinImports(%)Change in total revenues when imports increase (downward shift in wholesale demandcurve for local lettuce).New^Farm/^%^ %^ %^TotalImport^Coop. Change^Farm^Change^Farm^Change^Con-Qty. Output^in^Price in Farm^Revenue^in Farm^sumpt'n(M.lb.)^(M.lb.)^Output^($/lb.)^Price ($M)^Revenue^(M.lb.)Base Import Quantity5.8630 11.800 0.150 1.770 17.663Increased Import Quantity1.0% 5.9216 11.796 -0.034 0.147 - 1.770 1.739 - 1.805 17.7182.0% 5.9803 11.793 -0.068 0.145 - 3.603 1.707 - 3.674 17.7725.0% 6.1561 11.780 -0.170 0.137 - 9.524 1.613 - 9.710 17.9367.0% 6.2734 11.772 -0.239 0.132 -13.861 1.551 -14.133 18.04510.0% 6.4493 11.760 -0.342 0.124 -21.052 1.457 -21.466 18.20912.0% 6.5666 11.752 -0.410 0.119 -26.373 1.395 -26.891 18.31815.0% 6.7425 11.740 -0.513 0.111 -35.294 1.302 -35.988 18.48217.0% 6.8597 11.732 -0.582 0.106 -41.975 1.240 -42.801 18.59120.0% 7.0356 11.720 -0.685 0.098 -53.333 1.146 -54.384 18.755Although the members of the Cooperative have been benefitting from higherprices due to restricting supply, a longer run effect of the market quota scheme is that itmay also have reduced flexibility to price changes. As discussed above, with an inelasticsupply curve, price and not quantity does most of the adjusting to changes in demand.With the tariff protection form domestic head lettuce being reduced annually, it is possiblethat the quantity of head lettuce imported will continue to increase.33 Given the quotascheme, members may be less responsive to changes in market conditions, for example,members may continue to grow the same amount of lettuce to "protect" their quotainstead of reallocating land and other resources to alternative crops.33ltshould be noted that the exchange rate is also an important component in the wholesaler's calculationof the landed import price. The drop in the Canadian dollar relative to the U.S. dollar in the latter part ofthe 1992 season undoubtably afforded the Cooperative an extra degree of protection from U.S. imports.985.4^Summary and ConclusionsMarketing cooperatives play an important role in agricultural activities. Supportfor cooperatives is based on the idea that, collectively, farmers can achieve a beneficialmarket outcome that, individually, would be very difficult. Price related benefits dominatethe rationale for cooperative formation although risk reduction is important in some areasof agriculture. The most commonly stated ways a cooperative may influence priceinclude: 1) reducing the margin between farm and retail prices; 2) counterbalancingmarket power of the marketing sector firms, and; 3) increasing consumer price throughcooperative marketing either by restricting the quantity of farm product to the market orimproving the quality (physical or otherwise) of the farm product.The lettuce industry in the Lower Mainland of B.C. is organized into a producercooperative with exclusive marketing powers over all locally grown product (except fordirect retail sales at the farm gate). The Cloverdale Lettuce and Vegetable Cooperativeis also empowered to operate a market sharing scheme amongst its members. Althoughphysical production quotas are not a part of the scheme, the quantity of lettuce member-farmers are permitted to sell is based on a three-year moving average of historical sales.The Cooperative determines a selling price competitive with the landed price of importedCalifornia lettuce. As a "large country" exporter, the landed price of California lettuce isseen to be a ceiling for domestic lettuce prices. The quantity of domestic lettuce sold isdetermined by market demand together with the Cooperative's selling price and isallocated to members on the basis of their market shares.99A major objective of this study is to determine how and whether this system ofcooperative marketing benefits member-farmers. Restricting farm output to the marketis the classical way of raising prices, as epitomized by the monopolist seller. It ishypothesized that the Cloverdale Cooperative practices a limited form of output restrictionas a means to prevent the oversupply which, given an inelastic demand curve, wouldresult in rapid decreases in price and total revenues.In order to achieve this objective and test our hypothesis a supply and demandmodel of the B.C. lettuce market is constructed. Supply and demand elasticities areestimated with econometric equations using farm level and consumer level datarespectively. The estimated elasticities are used to formulate linear farm supply andconsumer demand equations using price and quantity observations for the year 1990.A proportional marketing margin is assumed to derive the wholesale level demand forlettuce. The supply curve for the Cooperative is derived by adding its fixed marketingcharges to the farm supply curve. The model shows that the B.C. lettuce market ischaracterized by inelastic demand and even more inelastic supply.To simulate a market with no controls on output, a counterfactual model isconstructed where it is assumed that the supply curve lies to the right of its presentposition. With no clear idea of what a non-restricted supply schedule should be, a rangeof alternative assumptions were tested. The results show that for very small increases inoutput (from imposing a flatter supply curve), large decreases in price occur.Consequently total revenues decline at every alternative supply assumption.This result supports the hypothesis that the objective of the Cooperative's output100restrictions is to prevent oversupply and the slide down the demand curve. The benefitsto members are the potential losses in total revenues. A key factor in the analysis is thedemand parameter. The more inelastic is demand the greater is the decrease in priceand total revenues. This is clearly shown in a second counterfactual model where a moreinelastic demand is tested with the original set of supply assumptions. The same slopechange in supply here more than doubles the loss in total revenues with the more elasticdemand.Regulated marketing by the Cloverdale Lettuce and Vegetable Cooperative wouldappear to be a response to the typical supply and demand characteristics of the marketfor many agriculture food commodities. The relatively price-inelastic demand for basicfresh vegetables, including head lettuce leads to a situation where, for a given demandcurve, increases in supply resulting from short-run favourable environmental growingconditions or longer run technological improvements, cause total revenues from the saleof the product to decrease. Total revenues can be expected to increase from reductionsin supply, thus planned output restriction is a way of controlling the output fluctuationswhich cause price and income variability.Although the Cooperative is the sole marketing agency for domestic lettuce itcannot act as a monopoly seller since imports are not restricted. The maximum to whichthe Cooperative may sell its product at is limited by the cost of transporting lettuce fromCalifornia plus the tariff protection. In the short run when the quantity and price ofimports is fixed, the Cooperative faces the residual wholesale demand curve and maycontrol the quantity of local product so as to keep domestic prices close to the level of101imports. For quality reasons, namely the longer storage life of California lettuce, domesticlettuce is priced, at best, slightly below that of imports. Total revenues are increased asa result of keeping domestic production at levels which yield this price.While maintaining higher price levels for producers, output restrictions havewelfare implications. If, as has been hypothesized, the observed farm supply curve hasbeen constrained by the market quota allocations, then the "true" supply curve mayinvolve a pivotal shift to the right. The area between the "true" and observed restrictedsupply curves below the demand curve represent a welfare loss due to allocativeinefficiency. The sources of these efficiency lqsses have been suggested to arise fromcost increases due to underutilization of existing productive capacity, non-capturedeconomies of scale and "x-efficiency" costs (Veeman, 1982). In the case of the B.C.lettuce industry it is difficult to establish whether allocative inefficiency is significant.Resources which may have been utilized in lettuce production may have merely beentransferred to other uses (i.e. other vegetable crops grown by lettuce producers). Non-captured economies of scale in lettuce production may be a more valid source ofallocative inefficiency, however crop diversification reduces farm income variability andhence the risk premium a farmer may require for specialization to a single farm output.There is no clear way of assessing these welfare losses since, ex post, there isno way to measure what "true" supply would be if the market quota allocations had neverbeen instated. Indeed, it is possible the "true" supply may not differ from what is currentlyobserved. However, the quota scheme which regulates the marketing of manyvegetables in B.C. remains a source of allocative inefficiency. The need for quota to102begin production and the incentive to maintain earned quota both render the farmer lessresponsive to price changes and delays the process of allocating resources to othercrops. As shown in the simulation, an increase in the quantity of imports shifts theresidual wholesale demand curve downward, but it is the price decrease rather thanquantity reduction which causes the decrease in revenues. In addition, the market quotascheme may also inhibit farmers from experimenting with new crops. As a result, the B.C.fresh vegetable industry may be in a less favourable position to respond to changingmarket conditions.The Cloverdale Lettuce and Vegetable Cooperative illustrates some of the benefitsof marketing through a central producer-owned selling agency. Members benefit fromscale economies for purchase of shipping cartons and materials; in the processing andmarketing of output and in the acquisition of market information. Standard grading oflettuce as well as post-harvest vacuum cooling have helped to bring the quality of thelocal product in line with that of imported lettuce. These services may not have beenprovided otherwise, or, at a higher cost to producers. More importantly, prices are keptas high as possible through member compliance with their market quotas. Membershave also benefitted from the lobbying activities of the Cooperative. For example, in June(1992) the federal government temporarily increased the duty on U.S. head lettuceentering Western Canada from the current 9 percent to 15 percent to protect producersfrom low import prices.34 This benefit may have been received without the Cooperativebut a coherent and united producer group undoubtably has some influence.34TheFinancial Post. June 25, 1992. p6.103The Cooperative and members view the marketing scheme as a system of deliveryto the marketplace to meet demand and to affect production to prevent short-term marketsurpluses. In addition, it prevents new growers and growers wanting to increaseproduction from "jumping into the market when it is already being supplied by existing(quota holding) growers".35 For current members the market quota scheme ensurestheir right to their earned market and guarantees them a share in any future marketgrowth. From a societal welfare point of view, such marketing schemes act as a barrierto entry for potentially more efficient producers. Furthermore, the scheme reduces theability of efficient producers to expand and may reduce the opportunity for innovativefarmers to explore new markets.Such a scheme requires the cooperation and support of participants, specifically,members must be committed to selling all their output through the Cooperative. Sincehead lettuce does not require processing growers can easily sell their output illegallythrough non-Cooperative channels. Anecdotal evidence suggests these side sales arefairly common but volumes are small relative to total sales. One reason for this may bethe difficulty in concealing production in field grown crops, especially within a small groupof producers such as the Cloverdale Cooperative. A second reason may be that sidesales are made at a discount from the Cooperative's price, therefore growers may engagein these sales mostly when their production exceeds their allocated market share. Todiscourage illegal sales, the Cooperative will withhold a member's market allocation forone or more days when transgressions are detected. Since market quotas in future years35BritishColumbia Vegetable Marketing Commission. Newsletter. Jan 14, 1987, p.7.104are dependent on current sales, there is an incentive for producers to protect their marketquota.At present, it appears producer compliance with the Cooperative's output controlsis successfully achieved, however, producers of head lettuce have no alternative but tomarket through the Cooperative. This raises the possibility that some growers, forexample, low cost producers willing to supply more output at all price levels could bebetter served by a different marketing structure. One option for producers limited by theirmarket quotas is production of non-regulated vegetables. Members may choose whetherto market these through the Cooperative and benefit from scale economies in thepurchase of inputs and marketing of output, or seek new market opportunities on theirown. For many small scale farms, the advantages of being within a large marketinginstitution may likely be greater than individual endeavour to market farm output.It is concluded that the Cloverdale Lettuce and Vegetable Cooperative has beeneffective in providing benefits to its members. As a unified group marketing through acentral selling agency, members benefit from economies of scale in input purchase,marketing and distribution, the lobbying activities of the Cooperative and potential priceincrease by collectively controlling domestic output to the market. The phasing out oftariff protection and the possibility of more aggressive marketing by U.S. exporters maywell increase the volume of imported lettuce to the B.C. market in the future. Given ahighly inelastic supply curve which results from the rigidity of the market quotas, the B.C.lettuce industry may be slow to respond to increased competition from U.S. imports.Two unintentional consequences of the market quota system have an adverse105effect on the industry's response to changes in market conditions. First, the quotas maybe keeping less cost efficient producers in the industry at the expense of more efficientproducers. Secondly, the quotas may reduce the flexibility of producers to switch toalternative crops. To address these issues, changes to the market quota scheme forregulated fresh vegetables will be needed. For example, the 3-year period on whichcurrent quota allocations are calculated could be modified to induce greater flexibility. Thiswould lessen the carry-over effects of past output on current farm production decisions.For the lettuce industry to survive in the face of increasing competition from U.S.imports, the marketing scheme will have to be changed to that the most efficientproducers are not constrained by their historic market shares to expand production. Thiswould inevitably require changes or retirement of the market quota scheme. The futurerole of the Cooperative will be less concerned with output control, and focused on theprovision of low-cost processing services, product development and more innovativemarketing.REFERENCESAgriculture Canada. Agri-Food Development Branch. Horticulture Section.Fruits, vegetable and honey crop and market report. 1969-1991.Barton, D. "What is a cooperative." in Cooperatives in agriculture,ed. David Corbia, pp 1-20. Englewood Cliffs, NJ: Prentiss Hall, 1989.Behrman, J.R. Supply response in underdeveloped agriculture: A case studyof four major annual crops in Thailand. 1937-1963.Amsterdam: North-Holland, 1968.British Columbia, Province of. Legislative Assembly. Select StandingCommittee on Agriculture. Marketing boards in British Columbia.Phase II Research Report. Vol II. September 1978.^ . Ministry of Agriculture and Fisheries.Production of vegetable crops together with an estimate of farm value.1969-1988.^ . Ministry of Agriculture and Food.British Columbia vegetable marketing guide. Rev. 1989.^ . Ministry of Agriculture, Fisheries and Food.Horticultural statistics 1989. 1990.Cotterill, Ronald W. "Agricultural cooperatives: a unified theory of pricing,finance and investment." in Cooperative theory: New approaches,ed. Jeffrey S. Royer, pp 171-258. Washington, DC: USDA, AgriculturalCooperative Service Report 18, July 1987.Corman, Jeff and Murray Fulton. "Patronage allocation, growth and memberwell-being in co-operatives." Canadian Journal of AgriculturalEconomics 38 (1990): 45-66.Curtin, L., D. Theoret and M. Zafiriou. Demand for foods in Canada: RecentEstimates. Agriculture Canada. Policy Branch. Working Paper 13/87.December, 1987.Deaton, Angus and John Meullbauer. Economics and consumer behaviour.Cambridge: Cambridge University Press, 1980.106Dunn, John R. "Basic cooperative principles and their relationship toselected practices." Journal of Agricultural Cooperation 3 (1988):83-93.Fulton, M.F. "Co-operatives in oligopolistic industries: The Western Canadianfertilizer industry." Journal of Agricultural Cooperation 4 (1989):1-19.Gujarati, Damodar N. Basic econometrics. 2nd ed. New York: McGraw-HillBook Company, 1988.Hammig, M.D. and R.C. Mittelhammer. "An imperfectly competitive market modelof the U.S. lettuce industry." Western Journal of AgriculturalEconomics 5 (1980): 1-12.Helmberger, Peter G. and Sidney Hoos. "Cooperative enterprise andorganizational theory." Journal of Farm Economics 44 (1962): 275-290.Helmberger, Peter G. "Cooperative enterprise as a structural dimension offarm markets." Journal of Farm Economics 46 (1964): 603-617.Howard, Wayne H. and George Klosler. An evaluation of market conditions andcritical success factors for agricultural marketing cooperatives.Guelph, Ontario: Dept. of Agricultural Economics and Business,University of Guelph, August, 1991.Huang, K.S. U.S. demand for food: A complete system of price and incomeeffects. Washington, DC: U.S.D.A. Economic Research Service.Technical Bulletin No. 1714. December, 1985.LeVay, Clare. "Agricultural cooperative theory: A review." Journal ofAgricultural Economics 34 (1983):1-44.^ • "Some problems of agricultural marketing cooperatives' price/output determination in imperfect competition." Canadian Journalof Agricultural Economics 31(1983): 105-110.Nourse, Edwin G. "The economic philosophy of co-operation." The AmericanEconomic Review 12 (1922): 577-597.Pauly, Mark and Michael Redisch. "The not-for-profit hospital as a physician'scooperative." The American Economic Review 63 (1973): 87-99.107Sexton, Richard J. "Cooperatives and the forces shaping agriculturalmarketing." American Journal of Agricultural Economics 68 (1986):1167-1172.^ . "Perspectives on the development of the economic theoryof co-operatives." Canadian Journal of Agricultural Economics 32(1984): 423-436.Sexton, Richard J. and Julie Iskow. Factors critical to the success orfailure of emerging agricultural cooperatives. Oakland, CA: Div.of Agriculture and Natural Resources, Giannini FoundationInformation Series No. 88-3, June, 1988.Sexton, R.J. and T.A. Sexton. "Cooperatives as entrants." The RAND Journalof Economics 18 (1987): 581-595.Staatz, John M. "The structural characteristics of farmer cooperatives andtheir behavioral consequences." in Cooperative Theory: NewApproaches, ed. Jeffrey S. Royer, pp 33-60. Washington, DC: USDA,Agricultural Cooperative Service Report 18, July 1987.^ . "Farmers' incentives to take collective action viacooperatives: a transactions cost approach." in Cooperative Theory:New Approaches, ed. Jeffrey S. Royer, pp 87-107.Washington, DC: USDA, Agricultural Cooperative Service Report 18,July 1987.Statistics Canada. Agriculture economic statistics. Catalogue No. 21-603E.^ Consumer prices and price indexes. Catalogue No. 62-010.^ Farm inputs price index. Catalogue No. 62-004.^ Industry selling prices. Catalogue No. 62-011.^ Provincial economic accounts, annual estimates.Catalogue No. 13-213.Taylor, Ryland A. 'The taxation of cooperatives: Some economic implications."Canadian Journal of Agricultural Economics 19 (1971): 13-24.Tomek, William G. and Kenneth L. Robinson. Agricultural Product Prices.Ithaca: Cornell University Press, 1990.108U.S. Department of Agriculture. Agricultural Marketing Service.Fresh fruits and vegetable shipments by commodities. states, andmonths 1989. FVAS Calendar Year 1988.Veeman, Michele M. "Social costs of supply-restricting marketing boards."Canadian Journal of Agricultural Economics 30 (1982): 21-36.109110APPENDIX: Data TablesTable 1:^Lettuce quantities and prices and leaf lettuce priceYearHeadLettuceAcresHead LettuceDomesticProduction(lb.)Head LettuceImports*(lb.)Head LettuceCooperativePrice($/lb.)Leaf LettuceCooperativePrice($/lb.)1969 482 11,424,200 1,900,000 .057 .0881970 548 14,716,640 1,000,000 .057 .1001971 545 16,231,468 1,750,000 .060 .0921972 551 14,254,000 2,918,000 .064 .0931973 571 21,801,000 945,000 .064 .1241974 572 19,214,400 1,435,000 .070 .1401975 576 16,480,000 1,927,000 .082 .1401976 560 17,491,000 2,563,000 .099 .1671977 563 16,925,000 3,364,000 .066 .1671978 551 16,543,000 4,028,000 .100 .2001979 565 20,676,000 1,781,000 .110 .2301980 516 16,984,000 3,126,000 .098 .2301981 497 10,364,000 5,605,000 .190 .3101982 520 14,032,000 3,551,000 .160 .2701983 511 16,138,000 2,222,000 .180 .3901984 500 15,665,000 2,686,000 .200 .3901985 500 17,139,000 2,298,000 .170 .3901986 530 14,240,000 4,369,000 .189 .2001987 510 14,643,000 3,084,000 .241 .2151988 515 12,551,000 4,055,000 .189 .1731989 525 12,600,000 4,485,000 .200 .2101990 510 11,800,000 5,863,000 .220 .2101991 N/A 8,388,000 4,879,000 .172 N/A* Imports for period June to September of the respective yearSource:^Agriculture Canada. Agri-Food Development Branch. Horticulture andSpecial Crops Division. Annual unload report: fresh fruit andvegetables. Various issues.B.C. Ministry of Agriculture, Fisheries and Food. Production ofvegetable crops together with an estimate of farm value.Victoria, B.C. 1969 through 1991.111Table 2:^Cooperative charges for head lettuce and leaf lettuceHead Lettuce*^Leaf Lettuce**YearAdminis-trationCharge($/ctn)CartonCost($/ctn)Adminis-trationCharge($/ctn)CartonCost($/ctn)VacuumCoolingCost***($/ctn)1969 .260 .390 .240 .351 .2161970 .303 .406 .283 .365 .2621971 .321 .420 .301 .388 .2871972 .316 .438 .297 .394 .3121973 .296 .469 .270 .422 .3371974 .253 .549 .233 .494 .3611975 .391 .594 .370 .535 .3861976 .433 .653 .405 .588 .4111977 .471 .676 .426 .608 .4361978 .531 .711 .481 .640 .4611979 .584 .878 .536 .708 .4861980 .636 .889 .590 .800 .5101981 .696 1.037 .647 .930 .5361982 .716 1.168 .641 1.051 .5601983 .762 1.242 .678 1.118 .5921984 .890 1.339 .810 1.205 .6231985 .745 1.402 .692 1.262 .6561986 .735 1.423 .676 1.281 .6871987 .764 1.544 .690 1.390 .7181988 .815 1.650 .720 1.485 .7501989 .835 1.350 .750 1.215 .7501990 .810 1.450 .720 1.305 .7501991 .815 1.250 .710 1.125 .750* Per lb. charges calculated by using 43 lb. per carton for head lettuce** Per lb. charges calculated by using 34 lb. per carton for leaf lettuce*** Vacuum cooling costs are the same for both head lettuce and leaf lettuce.Source:^See Section 4.2.2 for a description of this data112Table 3:^Total farm expenses (British Columbia)TotalTotal^Total^ Non-Labour Fertilizer^Total^MortgageYear^Wage^& Lime Pesticide^Interest($,000) ($,000)^($,000) ($,000)1969 22,713 5,923 2,597 10,8131970 26,341 5,281 2,319 18,8381971 27,859 6,382 2,313 18,1811972 28,976 5,996 2,789 18,2021973 36,743 5,186 4,002 22,0321974 44,476 9,880 5,413 33,7081975 51,852 16,330 6,633 39,1211976 58,969 14,402 6,491 46,4191977 70,449 14,657 6,770 50,1241978 73,998 21,678 5,399 70,1931979 80,553 19,300 6,087 102,8471980 100,571 21,939 8,502 121,3301981 113,351 25,050 9,232 189,6091982 131,838 27,473 10,613 177,3411983 142,333 27,790 11,568 120,0401984 155,962 30,627 13,385 118,5431985 168,918 36,529 14,606 115,6151986 180,303 35,407 13,690 112,4681987 189,972 35,313 12,766 103,5931988 199,595 34,058 12,000 110,4731989 212,035 37,029 12,165 124,6901990 206,729 36,902 14,236 129,4411991 N/A N/A N/A N/ASource:^Statistics Canada. Catalogue No. 21-603E.Table 4: Farm price indexes (Western Canada)*YearHiredFarmLabourMixedFert.& Lime Pest'd.Non-Mortg.Interest.FarmInputsPriceIndexFarmProductPriceIndex**1969 26.20 31.59 20.41 22.35 29.39 37.701970 26.82 30.31 21.30 24.33 29.57 36.901971 28.25 30.72 22.19 26.00 30.67 37.401972 30.68 33.73 23.90 26.15 32.76 39.801973 35.06 34.01 24.11 34.18 38.16 45.201974 41.90 40.86 37.74 46.63 44.26 51.001975 49.81 55.55 52.32 50.14 50.00 55.301976 57.46 65.32 53.91 58.17 53.65 58.401977 63.79 65.75 54.65 53.39 55.06 58.701978 67.77 65.04 56.22 63.53 62.27 67.201979 71.78 70.85 62.28 94.79 72.97 71.901980 76.93 94.20 72.66 108.83 80.27 85.001981 82.24 108.58 79.05 163.93 92.42 98.701982 87.42 105.54 88.14 150.00 95.56 96.201983 90.54 99.13 92.17 110.00 95.75 93.101984 94.33 101.73 94.55 115.25 98.15 99.301985 97.37 102.06 97.15 106.39 98.71 100.801986 100.00 100.00 100.00 100.00 100.00 100.001987 102.40 97.30 101.60 95.80 98.70 91.501988 107.00 98.90 101.50 107.30 101.70 92.501989 112.50 105.70 100.80 127.40 106.20 99.501990 116.30 93.70 101.90 135.90 108.60 98.401991 121.00 95.30 106.30 111.20 106.30 102.60* For all indexes, 1986=100** Farm product price index for vegetablesSource:^Statistics Canada. Catalogue No. 62-004.Statistics Canada. Catalogue No. 21-603E.113114Table 5:^Other price indexes*YearCorrugatedCartonsCanadaCPIAll ItemsVancouverPriceIndexof LettuceCanada**1969 29.20 30.30 20.601970 29.80 31.40 21.501971 30.50 32.40 22.301972 31.10 33.70 22.201973 33.40 35.90 36.801974 39.90 40.10 31.701975 43.20 45.40 28.801976 47.50 49.80 33.701977 49.10 53.30 36.701978 51.70 57.40 58.401979 57.20 61.90 50.101980 64.70 67.70 66.901981 75.40 77.30 82.701982 84.50 85.50 68.001983 89.90 90.20 85.101984 97.10 93.80 65.701985 102.00 96.80 70.501986 100.00 100.00 86.901987 108.60 103.10 85.701988 118.20 106.80 85.201989 120.30 111.50 82.001990 119.80 117.60 86.101991 113.70 123.70 114.40* For all indexes, 1986 =100** Median value for July to SeptemberSource:^Statistics Canada. Catalogue No. 62-010.Statistics Canada. Catalogue No. 62-011.Table 6: British Columbia Income and PopulationYearPopulation(,000)Total PersonalDisposable Income(million$)1969 2,060 5,5161970 2,128 5,9561971 2,185 6,6001972 2,241 7,7381973 2,302 9,3001974 2,376 11,1011975 2,433 13,1101976 2,467 14,9161977 2,499 16,7811978 2,542 18,9851979 2,589 21,7481980 2,666 25,5431981 2,744 30,1461982 2,788 32,9151983 2,814 33,6421984 2,848 35,6431985 2,870 37,8601986 2,889 39,3401987 2,925 41,8961988 2,980 45,5951989 3,048 51,2561990 3,127 54,8221991 3,218 57,679115Source:^Statistics Canada. Catalogue No. 13-213.

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0086336/manifest

Comment

Related Items