UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Marketing cooperatives and supply management Janmaat, Johannus Anthonius 1994

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

Item Metadata


831-ubc_1994-0460.pdf [ 4.19MB ]
JSON: 831-1.0087493.json
JSON-LD: 831-1.0087493-ld.json
RDF/XML (Pretty): 831-1.0087493-rdf.xml
RDF/JSON: 831-1.0087493-rdf.json
Turtle: 831-1.0087493-turtle.txt
N-Triples: 831-1.0087493-rdf-ntriples.txt
Original Record: 831-1.0087493-source.json
Full Text

Full Text

MARKETING COOPERATIVES AND SUPPLY MANAGEMENT: THE CASE OF THE BRITISH COLUMBIA DAIRY INDUSTRY by JOHANNUS ANThONTIJS JANMAAT B.Sc., The University of British Columbia, 1992  A THESIS SUBMITTED ]N PARTIAL FULFiLLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTERS OF SCIENCE in THE FACULTY OF GRADUATE STUDIES  DEPARTMENT OF AGRICULTURAL ECONOMICS  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA  August, 94 © Johannus Anthonious Janmaat, 1994  In presenting this thesis in partial fulfilment of the requirements for degree at the University of British Columbia, I agree that the Library freely available for reference and study. I further agree that permission copying of this thesis for scholarly purposes may be granted by the department  or  by  his  or  representatives.  her  It  is  understood  an advanced shall make it for extensive head of my  that copying  or  publication of this thesis for financial gain shall not be allowed without my written permission.  (Signature  /c5  Department of The University of British Columbia Vancouver, Canada Date  cI  DE-6 (2)88)  //  Abstract Cooperatives are commonplace in the dairy sector throughout the developed world. A cooperative is an organization whose patrons are those who contribute the capital. Two features that distinguish a cooperative are: profits are distributed by member patronage, and member control is democratic. In theory, this organizational form cannot sustainably capture economic rents. Members adjust their production until any captured rents are eliminated, restoring the competitive solution. In British Columbia, the dairy industry is regulated by supply management. Production quotas control output, while fanner returns are guaranteed by restricting imports and administering the price. All milk is pooled, and processors need not deal directly with dairy producers. A simple model of the BC dairy industry, with farm production or processor input as the only variable, shows that the ‘competitive yardstick’ is not maintained. The industry wide milk pool decouples the cooperative from its membership. When this cooperative maximizes its patronage dividend, supply management totally separates it from its members incentives. Given that the administrative price is not set to eliminate all processing rents, the positive patronage dividend is an incentive for all farmers to join the cooperative. Simultaneously, a competing IOF can capture rents because it is buying milk at the pooi price and does not compete with the cooperative for its input needs. The financial statements of the Fraser Valley Milk Producers Cooperative Association lend support to the model. Based on performance ratios, this cooperative is behaving similar to other firms in the dairy industry, and may be capturing rents on behalf of its members. The one area of discrepancy is in the source of financing, and this can be largely explained by changing member investment preferences. Our model predicts that in B.C. the price of quota should be dependent on the return generated by our theoretical cooperative. We find that the present perfonnance of the cooperative is not a useful predictor of the quota price. However, quota price appears to be closely linked to indicators of future economic performance, and the sign of this linkage is consistent with our model.  II  Table of Contents Abstract  ii  Table of Contents  iii  List of Tables  V  List of Figures  vi  Acknowledgments  Viii  Introduction  1  Chapter 1: The Structure of a Cooperative  3 3 10 19  1.1) What is a Cooperative 1.2) Consequences for Cooperative Behavior 1.3) Summary  Chapter 2: The British Columbia Dairy Industry  20 20 26 27 29 33 36 39  2.1) The History of the B.C. Dairy Industry 2.2) Goals of the System 2.2.1) The Federal Government 2.2.2) The British Columbia Government 2.2.3) The Farmer 2.3) Economic Consequences 2.4) Summary  40  Chapter 3: The Unrestricted Problem 3.1) Background 3.2) The For Profit Processor 3.3) The Independent Farmer 3.4) The Unregulated Market with Independent Agents 3.5) The Cooperative Processor 3.6) The Cooperative Member Farmer 3.7) The Unregulated Market with a Cooperative 3.8) The Unregulated Market with One Cooperative and one For Profit Firm 3.9) Summary  40 42 44 46 49 51 53 56 64  66  Chapter 4: A Quantity Restriction 4.1) The For Profit Processor 4.2) The Independent Farmer 4.3) The Supply Controlled Market with Independent Agents 4.4) The Cooperative Processor 4.5) The Cooperative Member Farmer 4.6) The Supply Controlled Market with a Cooperative and a For Profit Firm 4.7) The Quantity and Price Restricted Single Input Optimization Problem  ill  66 69 71 72 74 75 81  4.8) Sumniary.82  Chapter 5: The Pooled Milk Supply 5.1) 5.2) 5.3) 5.4)  83  A Cooperative and an IOF in Competition with Naive Conjectures A Cooperative and an IOF in Competition with Cournot Conjectures A Cooperative and an IOF in Competition with Unrestricted Conjectures Summary  83 91 93 98  Chapter 6: The Milk Market with a Supply Restriction, a Pooled Milk Supply, 99 and a Minimum Price 99 101 105 110  6.1) Problem Definition 6.2) Solution with Naive Conjectures 6.3) Solution with Open Conjectures 6.4) Summary  111  Chapter 7: A Linear Example The Unregulated Market in Isolation Open Economy in an Importing Country Closed Economy with an Aggregate Volume Restriction Closed Economy with a Floor Price Closed Economy with an Import Price Pool Summary  111 115 117 120 122 125  Chapter 8: The Financial Statements of the FVMPCA  128  7.1) 7.2) 7.3) 7.4) 7.5) 7.6)  129 132 134 139 141 143  8.1) Return on Total Assets 8.2) Current Ratio 8.3) Debt to Equity Ratio 8.4) Surplus Above Market Return 8.5) Percent of Surplus Retained 8.6) Summary  Chapter 9: Econometric Analysis  145 145 146 154 158  9.1) Theoretical Background 9.2) Data 9.3) Results 9.4) Summary  Conclusion  159  References  163  Iv  List of Tables Chapter 1: The Structure of a Cooperative 1.1) Rochdale Cooperative Principles 1.2) Cooperative Principles, International Cooperative Alliance 1.3) Cooperative Behavioral Principles, by Staatz  Chapter 2: The British Columbia Dairy Industry 2.1.1) Consequences of Supply Management  Summary statistics for the return on assets comparison Summary statistics for the current ratio analysis Summary statistics for the debt to equity ratio analysis Sununary statistics for the surplus above market return  20  128 131 133 138 140  145  Chapter 9: Econometric Analysis 9.1.1) 9.1.2) 9.2.1) 9.2.2) 9.3.1) 9.3.2) 9.4.1) 9.4.2)  5 6 8  38  Chapter 8: The Financial Statements of the FVMPCA 8.1) 8.2) 8.3) 8.4)  3  Summary statistics for quarterly series Summary statistics for annual series Quarterly regression results Annual regression results Quarterly data auxiliary regressions Annual data auxiliary regressions Quarterly regression corrected for serial correlation Annual regression corrected for serial correlation  V  153 153 154 154 156 156 157 157  List of Figures Chapter 1: The Structure of a Cooperative 1.1) 1.2) 1.3) 1.4)  Annual Transfers to the FVMPCA Number of Plants and Average per Plant Production Number of Producers and Average Production Changes in the Method of Financing  Chapter 2: The British Columbia Dairy Industry 2.1.1) Free Market Situation 2.1.2) Supply Managed Market  3 9 12 15 18  20 37 37  40  Chapter 3: The Unrestricted Problem 3.1.1) Firm’s total revenue and total cost as a function of total milk received 3.1.2) Firm’s marginal value product as a function of total milk received 3.2.1) Farmer’s total revenue and total cost 3.2.2) Farmer’s marginal revenue and marginal cost 3.3) Supply and demand 3.4.1) Supply curve for raw milk 3.4.2) Monopolistic processor total 3.5.1) Cooperative total revenue 3.5.2) Cooperative marginal value product and average value product 3.6.1) Cooperative member total revenue and total cost 3.6.2) Cooperative member marginal revenue and marginal cost 3.7.1) Marginal and average value product against member supply 3.7.2) Aggregate marginal revenue and supply curves 3.8.1) IOF Oligopolist in imperfectly competitive market 3.8.2) Cooperative in imperfectly competitive market 3.8.3) Aggregate market supply curve  44 44 45 45 47 48 48 50 50 53 53 55 55 60 60 60  66  Chapter 4: A Quantity Restriction 4.1) Marginal value product, free market price, and shadow value 4.2) Farmer’s marginal cost with and without supply management 4.3) Aggregate processor MVP and aggregate processor MC 4.4) Cooperative MVP, AVP, and adjusted AVP 4.5) Oligopolist’s marginal value products 4.6) Cooperative’s marginal value products  68 70 72 74 77 78  83  Chapter 5: The Pooled Milk Supply 5.1) The rents of the cooperative when it has been decoupled from its membership 5.2) The supply effect of the patronage dividend  88 89  Chapter 6: The Milk Market with a Supply Restriction, a Pooled Milk Supply, 99 and a Minimum Price 6.1) Interaction between the decoupled cooperative and an IOF 6.2) Patronage dividend effect on farmers  vi  104 105  111  Chapter 7: A Linear Example Unregulated competitive market Unregulated market with imperfect competition Unregulated competitive market in an open economy Oligopolist in an open economy for the final output Quantity restricted competitive market Competitive market with a floor price Pool price as a function of total milk brought to the pooi  112 113 115 117 118 121 125  Chapter 8: The Financial Statements of the FVMPCA  128  7.1) 7.2) 7.3) 7.4) 7.5) 7.6) 7.7)  8.1) 8.2) 8.3) 8.4) 8.5) 8.6)  Return on assets comparison Current ratio comparison Composition of member funds in the FVMPCA Debt to equity comparison Percent surplus above market return Percent of surplus retained as a share of total surplus earned  Chapter 9: Econometric Analysis  130 132 135 137 139 142  145  9.1) Quota price with fresh fluid price and blend price 9.2) Quota price and economic indicators 9.3) Quota price against surplus earned and TSE 300 composite  VII  147 148 150  Acknowledgments This project grew out of what was initially intended to be a simple graduating essay, to be competed over a four month summer. Clearly it has grown somewhat. Many people could be acknowledged for their involvement in the preparation of this document, and in providing me with the information that helped construct this analysis. In particular the Social Sciences and Humanities Research Council and the University of British Columbia must be acknowledged for funding support. I am particularly grateful to the people at Agrifoods International Cooperative Limited, in particular Carol Paulson and Cliff Denning, for the information they helped me gather, and the feedback they gave me on the pieces of the work as it progressed. I am also grateful to the members of this cooperative, in particular my parents, for their opinions and comments which were invaluable for the insights it provided into how the industry and the organization operates. I thank Mary Bohman, my supervisor and the initiator of this project, for her direction, insights, and encouragement. Lastly, I must thank my wife Suzie for her constant support, and Charolait and Skunkie, the cats, for their company while the project took shape.  vi”  Introduction Cooperatives are a prevalent feature in the dairy sector throughout the developed world. In almost all countries one would consider, at least fifty percent of the milk produced is received first by a cooperative [Grant, 1991]. The behavior and success of these cooperatives varies between countries. In the United States dairy cooperatives tend to be first receivers of the milk. Many then sell it to processing plants, or if there is an excess, process it themselves. That portion processed by the cooperative often makes its way to the stocks of the Commodity Credit Corporation. In the European Community cooperatives are more involved in the direct marketing of the milk. In Britain, there is one dairy cooperative, which has a legislated monopoly on the receipt of milk. It is also the largest marketer of milk through its own ‘DairyCrest’ label. A key factor that may determine what role cooperatives take in the marketing of milk is their interaction with the regulatory structure where they operate. This interaction has been little studied. In Canada, cooperatives are also directly involved in the marketing of milk. The degree of involvement varies between provinces. In Ontario, they are not very prominent. In this province all milk is sold through the milk board, with rights determined in part by an auction system. No processor, cooperative or otherwise, has a direct claim any of the milk. Producers receive the same basic return, irrespective of the processor they first ship to, and so see little need for a cooperative. In BC the situation is a little different. Here the milk marketing board may require milk transfers between processors, but the receiving processor has first option, given the same class of use. BC also had one of the most successful dairy cooperatives in Canada, the Fraser Valley Milk Producers Cooperative Association (FVMPCA). Before its merger with two other cooperatives, the FVMPCA was a successful dairy cooperative based in the lower mainland region of BC. Interviews with industry participants indicate that it was an aggressive competitor, and an innovator in the industry. In 1992 the FVMPCA merged with two Alberta dairy cooperatives, the Northern Alberta Dairy Pool (NADP) and the Central Alberta Dairy Pool (CADP), forming Agrifoods International Cooperative Association. The organization saw this merger as an essential move to maintain its ability to serve increasing demands from its clients.  1  In any industry, the behavior of the agent firms, be they cooperative or investor owned firms (IOFs), is governed by the interplay between the incentives facing the decision makers in the agent firm, and the constraints imposed by the regulatory environment. Partly encouraged by the prominence of the FVMPCA, this work investigates how the regulatory structure in BC may have affected the evolution and success of a cooperative. The first chapter will consist of a description of a cooperative, and some of the general implications of a cooperative’s structure on its behavior. Chapter two describes supply management and summarizes the general effects of this policy on the market. In chapters three through six we build a model of the interaction between a cooperative, an investor owned firm, and the regulatory environment. Chapter seven explores a simple mathematical model of these results. Chapter eight presents information taken from the financial statements of the FVMPCA, and in the last chapter we look at the results of a simple econometric analysis of the price of dairy quota as a function of the cooperative’s expected returns. We conclude the paper with a summary of our results.  2  Chapter 1, The Structure of a Cooperative. This chapter begins by presenting a definition of a cooperative. In the process we explore some of the history of the cooperative movement. The development of this movement has helped to identify some key features of a cooperative, as defined by an economist. We then show how these features appear in the FVMPCA. A cooperative’s structure has implications for its behavior. Following the definition of a cooperative, we present a summary of these implications. The structural implications produce characteristic patterns in the cooperative’s organization and relation to its membership. The financial information provided in the annual reports of the FVMPCA, among other sources, shows how these implications may have affected the behavior of this cooperative.  1.1) What is a Cooperative? A cooperative is a form of industrial organization where those that receive the services of the organization are also the suppliers of the capital. In agriculture, cooperatives are a particularly common form of organization. This is probably a function of the unique problems faced by farmers. The input supply and output purchasing markets are normally occupied by a small number of buyers and sellers. Farmers are caught between these oligopsonistic and oligopolistic players with little market power, facing high input prices and low prices for the products they produce. With such unfavourable market conditions, farmers could clearly benefit from collective action. As with any situation of economic crisis, ideologies have become strongly interwoven in the cooperative movement. Fundamentally, through cooperative action, farmers can gain a degree of bargaining power, bringing more competition to the industry, and in so doing improve their own welfare [Staatz, 1987a; Shaffer, 1987]. This has generated two lines of thought in the literature on cooperative. One approach has been to model the cooperative in consideration of the behavioral consequences of its structure. Another is devoted to promoting the growth of the cooperative movement, and spreading the ideology and philosophy of cooperation. For example, the International Labour Office defines a cooperative as follows:  3  A co-operative [...] is an association ofpersons who voluntarily joined together to achieve a common end through the formation ofa democratically controlled organization, making equitable contributions to the capital required and accepting a fair share of the risks and benefits of the undertaking in which the members actively participate [ILO, 1988 p6].  This definition embodies most of the essential features of the cooperative, but in a highly normative way. The ideas of a ‘common end,’ of ‘democratic control,’ and of ‘fairness’ have strong emotional appeal, but contain few specifics and are of limited analytical value. To the people involved in cooperatives, this form of industrial organization is not only a way of countering disproportionate market power, but is also a way of living. Many authors have written and continue to write on the importance of the normative side of cooperatives [Bogardus, 1952; Bailey, 1955; ILO 1988;  ...].  Cooperatives must therefore not be analyzed simply as economic institutions affecting the balance  between demand and supply, but the analysis must be completed with an understanding of these subjective valuations. Cooperatives have been both an economic agent and an ideological movement for most of their existence. Most cooperatives are based on the famed Rochdale principals. These were laid down by a group of renegade weavers, the ‘Society for Equitable Pioneers’ in Rochdale, England in 1844. The Rochdale cooperative was not the first attempt at cooperative activity, but it was the first success, and the principles that the Rochdale Weavers adopted were chosen with an awareness of what had befallen previous attempts at cooperative action. The Rochdale cooperatives were set up in response to perceived market failures, but these failures were seen as tied to a general societal malaise, and not to disproportionate market power [Bailey, 1955]. The goals of cooperation were set down both as an ideology and a practical attempt to change the structures affecting the welfare of the people involved [Rhodes, 1987]. In this way, they contributed to rearranging the balance of market power. The Rochdale principles are listed in table 1.  4  The Rochdale Cooperative Principles 1.  Open membership to all, regardless of sex, race, politics, or religious creed;  2.  One vote per member;  3.  Any capital needed should be provided by members, and should earn a limited rate of return;  4.  Any net margins should be returned to members in proportion to patronage.  5.  Cooperatives should allocate some funds for education in the principles and techniques of cooperation;  6.  Market prices should always be charged, i.e., no price cutting to pass on cooperative savings directly;  7.  Cash trading: no credit given or asked;  8.  Products should be accurately formulated and labeled;  9.  Full weight and measure should be given;  10.  Management should be under the control of elected officers and committees; and  11.  Accounting reports of financial health should be presented frequently to members.  {reference: Cooperative Theory, Agricultural Cooperatives: A Unified Theory of Pricing, Finance, and Investment. Ronald W. Cotterill, ppl7l-258.} Table 1.1: Rochdale Cooperative Principles  These are the foundation upon which the structure of the modern cooperative is based. Most modern cooperatives fundamentally adhere to the first five principles listed here. The ideas of open membership to all, equal democratic power, little return on equity, surpluses distributed according to patronage, and allocating resources for education in cooperative ideals. The further Rochdale principles are more specific to the market that the Rochdale weavers faced. The charging of the market price and the use of cash trading were instituted to prevent the financial crises that had confounded previous attempts at cooperation. The remaining principles represent the promise by the members to conduct business honestly and openly with each other, one of the main motivations for their union in the first place. Modern cooperatives have modified or discarded many of the more specific Rochdale principles. Today there are institutions in place to prevent corruption in the conduct of business affairs; ensuring ethical  5  The Rochdale Principles of Cooperation Established by the 1966 Congress of the International Cooperative Alliance 1.  Membership of a cooperative society should be voluntary and available, without artificial restriction or any social, political, racial, or religious discrimination, to all persons who can make use of its services and are willing to accept the responsibilities of membership.  2.  Cooperative societies are democratic organizations. Their affairs should be administered by persons elected or appointed in a manner agreed by the members and accountable to them. Members of primary societies should enjoy equal rights of voting (one member/one vote) and participation in decisions affecting their societies. In other then primary societies the administration should be conducted on a democratic basis in a suitable form.  3.  Share capital should only receive a strictly limited rate of interest.  4.  The economic results arising out of the operations of a society belonging to the members of that society and should be distributed in such a manner as would avoid one member gaining at the expense of others. This may be done by decision of the members as follows: (a) by provision for development of the business of the cooperative; (b) by provision of common services; or (c) by distribution among members in proportion to their transactions with the society.  5.  All cooperative societies should make provision for the education of their members, officers, and employees and of the general public in the principles and techniques of cooperation, both economic and democratic.  6.  All cooperative organizations, in order to serve the interest of their members and their communities, should actively cooperate in every practical way with other cooperatives at local, national, and international levels.  Cooperative Theory, Agricultural Cooperatives: A Unified Theory of Pricing, Finance, and Investment. Ronald W. Cotterill, ppl7l-258. Table 1.2: Cooperative principles, International Cooperative Alliance.  behavior in the business dealings between members, suppliers, and customers is no longer a driving force for organization. However, market power is still a concern, as is the issue of alienation caused by large institutions and distant decision making. The principles that today’s cooperatives adhere to reflect this modern reality. The cooperative principles adopted by the International Cooperative Alliance (ICA) are listed in table 2.  6  The first principle is a direct extension of the first Rochdale principle. The second ICA principle is a modification of the second Rochdale principle. It acknowledges the different structure of federated cooperatives. In such cooperatives, the ‘primary’ or local cooperative is democratically structured. However, the federation is often administered based on the patronage of the member coops [Hyadu, 1988]. The third principles are virtually identical. The fourth is an extension of the fourth Rochdale principle, which recognizes that cooperatives need not necessarily return all earnings to the membership directly, but may return earnings in the form of services or future returns. An example would be general information or extension services provided by the cooperative, or investment for future sales growth. The fifth principles are again virtually identical. The sixth ICA principle is not presented in the Rochdale list. It further recognizes and promotes federated cooperatives, along with a loyalty to the ideology. Economic analysts try to avoid normative criterion. They attempt to identify behavioral rules or laws based on the interaction of the stated principles and the theoretically predicted or actually observed behavior of the actors involved. These “economist’s principals of cooperation” are meant to represent the main factors that determine a cooperative’s behavior, rather than those that determine its ideological position. In this regard the values and principles embodied in the ideology of the cooperative are not ignored. One attempts to explain behavior that at first seems to be economically irrational by putting it in the context of an economic ‘value’ and using other measurable values as references. Many authors have compiled lists of cooperative behavioral principles. When these authors are dealing with agriculture, they are usually similar those in table 3, presented by Staatz [198Th]: In the case of the FVMPCA, these principles have been embodied as follows. The service provided by the cooperative, the processing and marketing of milk, is available primarily to members. Some non-member milk would be processed. In BC the FVMPCA acted as a ‘processor of last resort,’ receiving all production when other plants were closed. Further milk could be redirected according to processor demand. Between 1970 and 1991, the largest net receipt of milk accounted for about 15% of production, with the average being less than 4.5% of member production [FVMPCA, Annual Reports]. Figure 1.1 shows the amounts of milk received by the cooperative. The peak in 1981 occurred immediately before the Fraser Valley Milk  7  Cooperative Behavioral Principles 1.  The stockholders, who are farmers, are the major users of the firm’s services.  2.  The benefits a stockholder receives from committing capital to a cooperative are tied largely to patronage. There are three reasons for this: (a)  The business pays a strictly limited dividend on equity capital invested in the organization.  (b)  Net margins are distributed among stockholders in proportion to their patronage with the business rather then in proportion to their equity ownership in the firm.  (c)  Stock of cooperative firms does not appreciate because there is a very limited or nonexistent secondary market for it. Therefore, capital gains are not a major benefit of stock ownership in cooperatives, in contrast to IOFs.  3.  The formal governance of the business by the stockholders is structured “democratically” in the sense that: (a)  Voting power is not proportional to equity investment. The limitation on “voting one’s equity” may be in the form of one-member/one-vote, or voting may be proportional to patronage or stock ownership but subject to some limit such as restricting any one member from having more than 5 percent of the total votes.  (b)  There are strict limitations on the number of non stockholders who may serve on the board of directors.  Table 1.3: Cooperative behavioral principles, by Staatz [198Th].  Producers Association (FVMPA) merged with the Shushwap-Okanagan Dairy Industries Cooperative Association (SODICA) to form the FVMPCA. The peak could indicate market growth for the FVMPCA, some of which was likely at the expense of SODICA. The benefits of membership accrued as a ‘final payment’ on the milk production, not as a return on equity. The rules of the association stated that, “No member shall receive any dividend or interest on his shares” [FVMPCA, 1983, p11]. This is consistent with the cooperative principle, “share capital should  S  16.0%  50.0 Transfers  45.0  14.0%  Percent Transferred  • 4O0  lao% 10.0%  3O0  .  0  25.0 .  20.0  U,  0  15.0 h..  I—  10.0  CD  .  2.0%  5.0  0.0%  0.0 —  _  —  —  e  e  —  —  —  —  —  —  —  —  —  —  —  —  Figure 1.1: Annual transfers to the FVMPCA  only receive a strictly limited rate of interest.” The board of directors decided how much of the receipts from sales would be retained to cover losses, costs, borrowing charges and expenses incurred by the Association in carrying on the business of its membership, and a reasonable allowance for depreciation on all plant and equipment. [FVMPCA, 1983, i S] 2  The association could deduct up to 5% of total sales from what remained for the purpose of investment in plant, equipment, or other activities. Investments that were larger than this 5% limit could not be proceeded upon without a special resolution of the membership. Capital invested was redeemed as a bond, known as a ‘loan certificate,’ which paid a fixed rate of return. Loan certificates were paid out after 15 years, and were freely tradable between members and others who did not patronize the cooperative. The interest rate was decided by the members in a vote at the annual meeting. Capital held as shares conferred voting rights only. A farmer who wanted to become a member of the cooperative had to buy a minimum number of shares. Before 1983, members could accumulate shares. The retained portion of the final payment was rolled into shares and loan certificates, normally 10% as shares, the remainder as loan certificates. The shares were redeemed at par when the last of the member’s  9  loan certificates was paid out. However, following the merger in 1983, the rules were changed. All voting members would hold exactly five shares. Voting power was held by members, as distinct from shippers. Family members who worked primarily on the farm of someone holding a shipping account, and shareholders who worked primarily in an incorporated farm business, could be voting members. The board of directors was chosen by vote among the members [FVMPCA, 1983]. This organization shows most of the characteristics of the typical cooperative, as required by law. The access to capital seems to have provided a high degree of flexibility on the financial side, providing the firm with a strong competitive position. Simultaneously, the loan certificate system, multiple memberships, and the level of management and board accountability, added to or facilitated a high degree of member participation in the governance of the organization.  1.2) Consequences for Cooperative Behavior It is in light of the unique behavioral rules of a cooperative that one must evaluate the differences between a cooperative and an investor owned firm (IOF). One should not consider a cooperative as guided by the same objectives and constraints as a traditional IOF. In general the cooperative does not have the same objectives, a consequence of the democratic control by the member patrons. Some of these differences may put the cooperative at a competitive disadvantage, and others may give it an advantage.  Multiple Objectives and Efficiency: The ‘broader scope for optimization’ [Staatz, 1987a] derives from the integration of the behavior of the cooperative and that of the member farms. Farmers may see the cooperative as ‘an extension of the farm.’ The cooperative members will promote decisions by the cooperative to support the objective of the farm, and the farmer will modify the decisions at the farm level to integrate with the behavior of the cooperative [LeVay, 1983]. This behavior may be disadvantageous from a competitive perspective. For example, in the case of an agricultural marketing cooperative with members dispersed over a wide geographic area, isolated members will have an interest in keeping remote facilities operating. Such small facilities are likely less efficient, due to from diseconomies associated with their size. Members are concerned about paying  10  higher transportation costs to get their production to the cooperative’s facilities. At the same time, the farmer, as a member of the local community, is concerned about the community reaction if the plant is closed. Farmers see their acceptance in the community as a positive ‘good’ and the effects of closing the plant as a ‘cost’ that is equivalent to a certain level of income. An IOF will look only at the profitability of the plant, and the effects of its closure on overall firm profits. The concerns of the individual shipper at this isolated location will not be a factor in the decision, and would be ignored. As the cooperative becomes larger, and covers a large geographic area, some degree of subsidization of the less profitable plants by the more profitable plants will probably occur. This subsidization will distort the decision that would have been made had the local cooperative been independent. It may have chosen to close the local plant in isolation, and either dissolved the cooperative or simply pooled transportation costs to a distant processor. With the possibility of being subsidized, and likely a history of having received subsidization, there will be greater pressure to keep the plant open. If the cooperative has a number of such isolated plants and producer members, there will be a large political lobby to keep all the small plants open. This political pressure may result in business decisions being compromised, and the profitability of the cooperative being jeopardized. Producers who ship to more profitable plants have an incentive to leave the cooperative. They will remain until the difference between the price they can receive from IOF buyers and that from the cooperative exceeds the value of the pro-competitive effect of the cooperative and the other non-market values such as “fairness” and “cooperation.” The history of the FVMPCA shows that this pressure may be affecting its behavior. It has the largest number of plants of any single processor in the British Columbia, and has been slower to rationalize production than the industry as a whole. Figure 1.2 shows the number of FVMPCA plants and provincial processing licenses and the normalized average throughput per plant or license. The milk board requires all processors to hold a license if they are actively processing milk. License numbers are not restricted, but a license does require an annual fee be paid to the board if it is not to be revoked. Throughput was normalized by indexing it to the relevant 1974 production level. A normalization was used to highlight changes between FVMPCAs and non-FVMPCA throughput. The number of cooperatively owned plants  11  250  60  50  200  I4IA[ii -  1:  :1  I I I J.LiiikuiIIIIIItIlIIth :  1974 1975 1976 1977 1978 1979 1980 1981  FVMPCA’  ‘All Coops  1982 1983 1984 1985 1986 1987 1988  BC Totals —4—— FVMPCA —4— BC Average  Figure 1.2: Number of plants and average per plant production  has remained relatively constant over the period considered. This contrasts with a generally downward trend in the total number of processing licenses. The high number of licenses in 1988 occurred while the price of milk in BC was higher than in the rest of the country. The number of licenses issued may reflect an attempt by processors to enter and take advantage of the market opportunity. The constant number of cooperative plants, when compared to the total number of processors would be uninteresting if cooperative plant efficiency was keeping pace with the rest of the industry. The average throughput of the plants was calculated by dividing the total production in the FVMPCA and in the province as a whole by the respective number of plants. This number was then scaled by using 1974 as a base year, which allows both data sets to be shown on the same chart. The average throughput for FVMPCA plants in 1974 was 39.33 million litres per year, which was more than ten times the average among nou-FVMPCA processors of 2.92 million litres. By 1986, the FVMPCA throughput had increased to 42.50 million litres per year, while the throughput of non-FVMPCA processors had increased to 7.11 million litres per year. The total processing licenses in the province includes cottage industries, which accounts for the much lower level of production in total. Many of these small processors are highly specialized, which may endow them with unique avantages. Efficiency measurements based on  12  throughput need to be interpreted carefully. However, the rationalization that has occurred in the industry as a whole does not seem to have occurred in the FVMPCA. An interesting fact is the jump in the number of processing licenses in 1987 and 1988. At this time, the relative price for milk in BC was higher than the rest of the country. The jump in the number of processing licenses issued may be a response by entrepreneurs to this opportunity. The increase in the number of licenses is primarily responsible for the drop in the throughput level. Historically, many of the cooperatives in the dairy industry arose specifically in response to the rationalization taking place in the early part of the  th 20  century. The large companies that controlled the  industry were closing small, unprofitable plants in rural areas, leaving farmers with a supply of cream and no one to sell it to [Hazlett, 1992]. Because of fixed assets in the form of diary cattle and dedicated equipment—more a factor in recent times as dairy farms specialize—the farmer had a high sunk cost in the production of milk [Manchester, 1983]. The ability to market the milk and gather some return is a benefit to the farmer, even if the profit is not as high as an IOF would demand. The local farmers are in fact better off ‘subsidizing’ their local cooperative processor by accepting lower returns on capital for the plant, since it maintains a higher return to the assets on the farm. A small local cooperative may have a competitive advantage locally. The cooperative provides a marketing service where none would otherwise exist [Sexton and Iskow, 1988]. Democratic Control and Organizational Direction: The democratic principles that are the hallmark of the cooperative also present potential problems. In a cooperative where all members are the same size, and have uniform interests, the democratic method of making decisions generates the best decision for all. However, as presented above, producers at different locations may have different interests. Similarly, producers with different sizes may also have different interests. Larger producers often are more ‘businesslike’ in their approach to agriculture. They have reduced their own costs and increased scale, and have reduced their cost to the cooperative by shipping a large volume of product at one time. These members benefit when shipping and inspection costs are allocated to individual producers. Smaller  13  producers prefer to pool these costs. Overall, large efficient producers and producers close to plants subsidize those who are smaller, and those located farther away. In most industries there are more small firms than large ones, while large firms are together usually responsible for the largest share of output. This is also the case within most cooperatives, where we find far more small member producers than large ones, even though the large volume producers collectively account for the majority of the cooperative patronage. The democratic rules of a cooperative make it difficult for the interests of the largest producers to be represented in accordance with their impact on the cooperative’s revenues. Figure 1.3 shows the average production of all producers in BC, of FVMPCA members, and of producers who are not members. The most relevant comparison is between FVMPCA members and non members, since cooperative members are about 80% of the total number of producers. The average production of members is lower, but not by all that much. Before 1980, the average production of the members of the FVMPA was almost the same as that of the rest of the producers in the province. In 1982 the FVMPA merged with the SODICA to form the FVMPCA. Following the merger, a number of larger shippers (although actually non-members) began their own processing facility in response to an FVMPCA policy of accepting production from members only. The merger meant that the FVMPCA acquired more facilities and a number of smaller producers, causing average member production to fall. The FVMPCA responded to the risk of loosing large producers by offering a volume discount. This move was made after a number threatened to leave [personal interview]. The tendency for the interests of large producers to be ignored in a cooperative may be lessened by the supply management system. Supply management makes it more difficult for producers to expand, reducing the size range of cooperative members. It also provides a more secure return to small producers, reducing their reliance on the cooperative’s patronage dividend. In 1977, it was estimated that the minimum efficient processor in North America would be supplied with the milk of between 10,000 and 20,000 cows [Sexton and Iskow, 1988]. In the U.S., where there are no supply restrictions, herds close to or above these sizes do exist, and are often closely tied to a processing plant. In Canada, the rules of  14  1600  600  1400  550  1200 C) .1000  500  0  (0 ‘1  2  5  -‘  800  400  0  3 -U  600  j::..:.:...iMPCAMembers  E  g  Producers  400  200 I  0  m  —  —  —  —  a —  a  —  o  —  c  —  0 —  o —  o —  _  o —  o —  FVMPCA Producers • Non-FVMPCA Producers • —A——All Producers —r —i- ---i —I o  —  o —  o  —  —  —  —  —  0  I  250  200  —  Figure 1.3: Number of producers and average production  supply management often restrict the overall market share a singe producer may hold, keeping farm size small. Another often cited effect of the democratic structure of the cooperative is the difficulty in bringing expertise to the board of directors from outside the organization. It is felt that member farmers lack the business skill required to run a large organization. It is also felt that management can take advantage of this possible confusion of the board of directors and pursue its own objectives. This is seen as being in contrast to IOFs, which are felt to be free to pursue profit maximization as the sole objective, with the board of directors perfectly representing the interests of the stockholders. There is in fact no reason to suppose that the objectives of the board of an IOF are that uniform [Rhodes, 1987]. The board of directors of the FVMPCA has contained members with a diversity of experience to draw on. All are required to be producing dairy fanners, but board members have had backgrounds in banking and agricultural economics. This indicates that expertise has been available within the organization. In Canada an entrant into dairy production requires a large commitment of capital. New entrants who are not coming into the industry by succession will often have had non-farming, and often white-collar,  15  experience before entering. This generates a pool of diverse skills which one would not expect if it was easier to enter. Board membership has tended to remain fairly stable over time, suggesting that the board members could become fairly well informed about how the processing industry operates. However, static board membership could mean that the management has successfully convinced the board and the membership of the complexity in the industry and the fact that ‘nonnal’ farmers could not run the organization. Some members have expressed sentiments along these lines [personal interviews]. Considering the high degree of member involvement during most of the life of the FVMPCA, member apathy and/or domination of the board by management interests seems not to have occurred.  Moral Hazard and Financing: Members have a tendency to under finance the cooperative, relative to the optimal level of financing for an IOF [Lerman and Parliament, 1991; Parliament eta!, 1991;  ...].  This  tendency arises in part from member risk aversion. Farmer patrons have the choice to invest capital in the cooperative or in other opportunities. These farmers already have a lot of capital invested in this industry in their farms. To avoid risk, the farmer should diversify outside the industry. The inability of share capital to appreciate also reduces the incentive to invest. Since shares cannot be sold, but are redeemed after the member retires, there is no vehicle to capture the future value of the cooperative. The benefits of investment in the cooperative accrues to members only during their involvement with the cooperative. Members choose their optimal level of investment based in part on this finite discounted income stream. The level of investment members make is based on the individual’s perceived income risk should the cooperative fail. As the cooperative ages, becoming a more established player in the market, the portion of the membership that experienced the economic conditions at the time when the cooperative was formed declines. The perceived severity the cooperative’s dissolution becomes smaller, reducing the incentive for members to invest. The free rider problem, where members seek to minimize their own investment relative to others who have invested, will also lead to a tendency to under finance. New members and young members have an  16  inclination to try to minimize their investment, while the investment of older members remains in the organization as the financing capital. They can have equal control through the democratic process, but lower the risk of their position. The FVMPCA has experienced a reduction in the level of member equity. Figure 1.4 shows how the distribution of financing has been arranged in the cooperative. In 1970 over two percent of the surplus was retained. The percent retained was calculated as the ratio of the retained earnings to the total available for distribution. As the years passed, this ratio was reduced until in 1991 it was just below one percent. As a consequence, the debt to equity ratio basically during this period. The debt to equity ratio was calculated as the liabilities of the cooperative to the sum of loan certificate capital and share capital. The reduction in the share of the profit retained was seen as a positive step by younger producers interested in expanding their own operations. They appear to be less concerned about the historical events than older members. These older producers often speak against these changes, and also show a greater philosophical link to the cooperative principles that were seen as critical to breaking the monopsonistic pricing they faced in the past. Factors in Favour of Cooperative Organization: There also are some potential advantages that the cooperative has over the private firm. Unfortunately these differences are difficult to quantify and analyze in anything approaching an objective manner. Perhaps the greatest evidence that there are benefits to cooperative organization is the fact that they continue to exist in spite of clearly identifiable structural shortcomings. The cooperative structure offers the opportunity for more communication between management and the member patrons. Management is able to better respond to the concerns of the member patrons in how services are delivered. It also allows membership to be better informed of the market conditions that determine the success of the cooperative, and may provide the member farms with the information to more rapidly respond to these changes. The cooperative structure offers a potentially highly effective mechanism for the dissemination of information on production techniques and industry direction.  17  2.50% .U  m  -I  •2.00% (0 0 • 1.50% 0  U)  0.50% 0 0. 0.00% —  —  —  —  —  I0  D  F:  —  —  —  —  —  —  —  _  _  —  0 —  Q  —  O  —  m —  —  —  —  Figure 1.4: Changes in the method of financing  Cooperatives, by their nature and history, embody a philosophy as well as a set of economic incentives. This philosophy often brings with it a sense of loyalty that goes beyond that associated with the risk of loss should there be no cooperative in the industry. This fact can be a significant competitive advantage. Members are willing to accept losses at present to guarantee the long term success of the cooperative. Depending on the degree to which the ‘philosophy’ of cooperation is adhered to, this fact may partially offset or even override the tendency to under finance the cooperative. The role that the agricultural cooperative plays is dependent on many factors. In the beginning, cooperatives were formed to address a wide range of perceived failures in the marketing system. Today many of these market failures have been contained by government authorities. Modern cooperatives are agencies that allow the farmer to vertically integrate into the marketing system, without having themselves to establish the scale for an efficient processing operation [Sexton, 1988]. Today’s farmers are also managing much larger and more specialized operations, requiring operators that are much technically skilled. The benefits of cooperation are different today. They allow farmers to specialize in production, while the cooperative takes care of marketing. The question is whether this organization still  18  serves the best interests of the producer, or if the main interests being served are those of management, and personal goals of individuals on the board of directors.  1.3) Summary Cooperatives are a unique form of industrial organization that is common in production agriculture. A cooperative is characterized by a number of features, including democratic control by the members of the organization, return of profits to the members on the basis of patronage, and a limited return on equity, all of which is provided by the membership. This form of organization has a number of potential disadvantages, stenmiing primarily from the individually opportunistic actions of members. Members do not fully internalize the cooperative objective, as a result, the organization as a whole is constrained in the actions it can take. The information contained in the financial statements of the FVMPCA indicates that some of these problems were making themselves evident. However, there are also some potential advantages to cooperative organization. These are hard to identify, and even harder to quantify. Many stem from the interaction between economics and social philosophy which a cooperative represents. The very fact of their prevalence indicates that there must be some advantages to membership.  Throughout the remainder of this work we explore the interactions of a cooperative with the regulations in the B.C. dairy industry. In the next chapter we outline the development of the supply management system, with a special focus on B.C. Following this we build a simplified model of a dairy marketing cooperative, and explore how the regulations might impact on its behavior. Our main concern will be with the clearly identifiable costs and benefits of this form of organization, so we will largely ignore the wider range of features that characterizes a cooperative.  19  Chapter 2, The British Columbia Dairy Industry The dairy industry has a history of govenunent intervention that stretches back almost one hundred years. Many feel that the dairy industry faces unique challenges, such as specialized capital investment on the farm which generates the opportunities for opportunistic behavior by the monopsonistic processors who purchase the milk. Late last century dairying underwent a stressful transition from producing primarily cheese, butter, and cream to the production of fluid milk. The hardship faced by farmers eventually gave government few politically acceptable options that did not involve extensive involvement in the dairy sector. Canadian governments chose to enact a supply management system, which supports the fanner by guaranteeing that a specific amount of the farmer’s production will have access to the market at a guaranteed price. In return the farmer is required to accurately regulate production, and must pay for the disposal of any surpluses. In B.C. the program is administered by the provincial milk board, which allocates quotas for fluid milk production, and administers the federally allocated quota on industrial products. The supply management system accomplishes a number of objectives, in particular bringing stability to the industry. However, this stability comes at a cost, which the entire economy pays. In this chapter we characterize the regulations that make up the supply management system.  2.1) The History of the B. C. Dairy Industry The Canadian dairy industry, like that in most of the industrialized economies, has been heavily regulated for many decades. One objective of these regulations has been protection of farmers from what are seen to be unpredictable and often devastating market forces. This attitude arose in the later part of the nineteenth century. The dairy industry was adjusting from producing primarily solid milk products, such as cheese and butter, to one supplying fluid milk to urban centers. The large capital investment in processing technologies, the perishability of the product, and the strict sanitary regulations, generated a  20  highly asymmetric bargaining position between the farmer producing the milk and the processor who was purchasing it. This bargaining relationship was further aggravated by the production technology. Solid milk products were to this point made primarily on the farm, and were sold to the merchants in town on a per unit basis. If the fanner could not sell the day’s production on a particular day, it could be stored to the next day. On top of this, the farm was usually very diversified, and the skim milk that was a byproduct when the cream was removed was usually fed to calves or hogs. The storable manufactured product was treated as a unit quantity, and the trading mechanisms were appropriate to this. However, fluid milk cannot be practically stored, so that the farmer had to find a buyer for a certain quantity of milk every day, irrespective of the market price that day. The problem now became one of properly managing and pricing a flow rather than buying a unit of milk, a problem to which the trading institutions of the day were unaccustomed. The traditional trading relationships relied on a per unit style of pricing, and was unable to properly price a situation where the short run marginal cost of production was almost zero [Manchester, 1983]. The farmer received a significant premium for the milk that was sold for fluid consumption, but could never be certain that the milk produced would receive this price. Many farmers expanded production in response to the price, only to find themselves forced to sell milk at a loss. Processors often turned away milk during the spring and summer months, and were searching far afield during the low production fall and winter. For many years as this adjustment proceeded, milk prices were highly volatile and generally trending downwards. The continuing adjustment forced many farmers from their farms. As the shakedown continued, farmers in many areas organized into cooperatives and associations to gain some market power. They also aggressively lobbied their governments to provide some relief and impose some control in the industry. The added hardship of the depression further aggravated the situation, and where governments had not acted earlier, they intervened now. In British Columbia the situation was much like that elsewhere in the industrialized world. After many years of aggressive competition, governmental involvement began in 1929 with the passage of the Piii Products Sales Adjustment Act. This act created a board to equalize the returns between producers who  21  managed to ship milk to the fluid market and those who were shipping to the lower priced manufacturing milk market. This was the first time that pooling had been seen in BC. Before this, severe competition was forcing down the price of milk, and heavily impacting on the livelihood of the dairy farmers [SSCA, 1978, p’7]. Two years later began the legal challenges, the traditional way that Canadians arrive at a stable division of power between the federal and provincial levels of government. The part of the Canadian constitution affecting this division of powers, derived from the British North America Act (BNA), does not clearly delineate the level of authority. Parliament has the authority to regulate trade and commerce [BNA Act, 91.2], but the provincial Legislature has authority over  “...  all Matters of a merely local or private Nature  in the Province [BNA Act, 92.16].” Industrial milk is generally considered to fall under trade and commerce, as industrial products can move across provincial borders. However, when the regulations were introduced, inter-provincial trade of fluid milk was nonexistent making its regulation a matter of a local nature. Traditionally the courts have been the mechanism through which the legally proper level of authority was identified [Jackson, 1990]. In 1931 the Dairy Products Sales Adjustment Act was declared ultra-vires  --  beyond the authority of—the  provincial government. In response the federal and provincial governments simultaneously introduced market control legislation. Shortly thereafter, the federal law was found to be ultra-vires the federal govermnent. However, the provincial legislation survived until 1941, ten years more [SSCA 1978]. During this period, the federal government introduced subsidies to assist the dairy industry. Initially these were put in place to alleviate the conditions farmers faced with the depression. Later, as the war began, subsidies were again introduced to encourage production in the face of wartime shortages [Barichello, 1981]. In 1941 the provincial board was found to be exercising powers beyond those provided by the enacting legislation. Once again the industry was without regulation. However, in 1942 the marketing of milk came under the control of the War Time Prices and Trade Board. All prices were controlled to prevent inflation. Coupled with this a subsidy was provided to encourage production. These measures remained  22  in place until 1946. In 1947 the BC Milk Board was established by amending the provincial Public Utilities Act. The board was empowered to set both retail and producer prices, and subsequently raised these prices to provide the producers with compensation for the wartime subsidy that had been lost [SSCA 1978]. At the same time, the government began to support farm prices with deficiency payments and offers to purchase. These were intended to be temporary measures to alleviate unusual hardships, but they heralded the beginning of the active involvement by the Canadian government in the trading and storage of dairy products [Barichello, 1981]. In response to lobbying efforts by Canada Safeway, consumer price controls were abandoned by the milk board in 1953. This change lead to heavy competition between processors attempting to dominate the market. At the same time, in a bid for market superiority, the Fraser Valley Milk Producers Association (FVMPA), a producer cooperative formed in 1913, followed a policy of aggressive pricing that lowered producer returns. The FVMYA argued that as a cooperative it was not bound by the producer price that the board dictated. By 1955 another court case concluded with the finding that the board had no power to control the price paid to producers [SSCA 1978]. Subsequently a royal commission was struck to deal with the crisis. The findings lead to the passage of the Milk Industry Act in 1956. This act placed the control of production and marketing of milk under a reestablished Milk Board. It gave the board the power to: 1.  license producers and processors;  2.  audit farm and dairy business record.s, and;  3.  set standards for the production, processing, distribution and sale offluid milk within the province [SSCA 1978, p9].  It also established a quota system covering fluid milk production. The quota system gave the board the authority to regulate the total amount of fluid milk produced, as well as control the farm level of production. This was seen as a way of apportioning the returns from this market fairly between the farmers.  23  Shortly after this, in 1958, the federal govermnent introduced the Agriculture Stabilization Act. This act did not affect the dairy industry immediately, as the federal authorities continued to interact with the industry on an ad hoc basis. In the early 1960s the dairy farmers, disturbed with this piecemeal approach, lobbied for a central authority to administer the federal dairy policy. In response, in 1967 the Canadian Dairy Commission (CDC) began operation. The CDC was vested with a broad range of powers. It could issue offers to purchase, provide for storage, for processing, and for disposition of product. It could apply import controls, provide deficiency payments, conduct dairy product promotion, deduct levies from individual producers, and conduct investigations into production, processing, and marketing activities of any dairy product. Participation by the provinces was voluntary. [Barichello, 1981]. The CDC developed the federal Market Sharing Quota (MSQ) program. Under this program, all the farmers in each participating province were allocated an annual quota guaranteeing them a share of the market. This move brought the regulation of all of the production of industrial milk in Canada under one consistent set of rules. Farmers would receive a guaranteed price for the milk produced within the MSQ amount, and paid a penalty levy on any excess amounts. The initial allocation of MSQ between the provinces was based on the historic distribution of production. Provinces such as Ontario and Quebec, which had a long history of dairy production, received the largest shares of the MSQ. In 1973 the province of British Columbia joined the federal Comprehensive Milk Marketing Plan. At the time, the provincial milk board was given the authority to administer the federal Market Sharing Quota (MSQ), and collect the levies that were due on overproduction [SSCA, 1978]. The Milk Board could not issue MSQ, and as the historically based allocation did not anticipate future population shifts, B.C. saw its share of the local market for industrial milk fall over time [SSCA 1979]. In an attempt to acquire more MSQ, B.C. withdrew from the national scheme in 1982-83. The next year B.C. won an agreement on levies and MSQ allocation, and reentered the system [Barichello, 1987]. The agreement, known as the  ‘65/35 agreement’ guaranteed B.C. a level of MSQ equal to 35 percent of the total milk produced in the province. Although this arrangement maintains industrial production at a level that is below the national  24  average, it has managed to insulate the B.C. dairy industry from changes in the level of consumption of industrial milk products. At present the industry is facing several important challenges. The first challenge comes as a result of the price differential between the Canadian and American milk prices. Consumers have become aware of how large the price differential is, and as the global economy continues to sputter along, consumers are becoming more and more conscious of where their income is going. The amount of cross border shopping for dairy products has become a significant issue over the last decade [Shelford, 1988a] and has been estimated as up to five percent of the B.C. domestic market. If this cost gap does not close, either through price change or exchange rate movement, imports will continue to reduce the domestic demand. ,  A persistent case involving a number of milk shippers who have opted to produce milk without quota has been an ongoing irritant to the B.C. Milk Board for almost a decade. In August 1993 the final court decision was handed down, and the ruling threatened the dairy industry throughout Canada. The ruling handed down by Madam Justice Newburry held that the enabling legislation at the federal level defined milk as the product that was regulated, and not dairy products [BCJ, 1993]. As such, any producer could produce any amount of milk, provided that the milk was processed into a further product while still in the legal ownership of the producer. This ruling means that the grounds for the federal market sharing system are in question unless the enacting legislation is changed. As of 1988 there were over thirty producers who were producing without quota, and in the uncertainty that exists now, this number is bound to grow [Shelford, 1988b]. The last challenge to the supply management system is a result of the General Agreement on Tariffs and Trade (GAiT). A GATT appeal decision ruled that ice cream and yogurt are not primary products, and thus cannot be given the same protection as primary dairy products. Under article eleven of the GATT, import restrictions were allowed if there was a domestic supply control program in place. This article has been removed. .Tariffs must replace the import restrictions, relaxing the precise control government had over import volumes. All indications are that the tariff levels will be very high, giving domestic producers significant protection. However, this change invariably limits the degree of control that the domestic  25  authorities have over the domestic consumption. Further, the system of levies to offset export losses is considered an export subsidy, and is also prohibited by the GATT. Supply management will have to undergo some significant restructuring to survive over the long term.  2.2) Goals of the System The goals of the Canadian Dairy programs follow from the turmoil that was present in the industry before the programs began. Following Barichello [1981], the federal policy objectives can be stated as: 1.  to ensure a reasonable degree ofself-sufficiency in processed daiiy product supplies,  2.  to procure price stability for both producers and consumers,  3.  to ensure that efficient Canadian industrial milk producers receive a reasonable return on their resources,  4.  to provide Canadian consumers with adequate and continuous year-round supplies of high quality processed dairy products at reasonable realprice levels.  For the most part, these goals apply at both the federal and provincial levels. These goals have been achieved, more or less, through the use of the supply management system. Self sufficiency is ensured by default through the elimination of legal imports of dairy products. Stability is achieved through an administered price and supply management, limiting the opportunity for supply and price shocks on both the consumer’s and producer’s side of the market relationship. A ‘reasonable’ return is guaranteed by an administered price partly derived from a formula which suggests a price in accordance with milk’s ‘cost of production.’ In fact, the rapid rate of technological advance in agriculture probably means that the administered price is higher than it ‘should’ be. And lastly the ‘adequate’ supply and high  26  quality is ensured through active regulation of how much must be produced, and under what conditions this production takes place.  2.2.1) The Federal Government The Canadian Dairy Commission The federal Comprehensive Milk Marketing Plan is controlled by the Canadian Dairy Commission (CDC). The objective of this body, created under the Agricultural Stabilization Act of 1957-58 and amended in 1975, is to provide efficientproducers ofmilk and cream with the opportunity of obtaining a  fair return for their labour and investment, and to provide consumers of dairy products with a continuous and adequate supply of dairy products ofhigh quality [Barichello, 1981, p19]. To accomplish its goals, the CDC operates the Market Sharing Quota (MSQ) system. The MSQ is measured on the basis of a farmer’s butterfat production over the year. The amount of MSQ that is required is decided every year by a panel of representatives from each province that is part of the program. This quota is administered by the provincial Milk Board, along with the provincial fluid quota [SSCA, 1979]. Thus, it is at the provincial level that one finds the regulations that govern the allocation and transfer of quota between producers, and the regulations that govern processors. To make the supply management system effective, imports must be tightly controlled. The regulatory body cannot accurately control how much milk is brought to the market if consumers and buyers are able to import at will, and producers can sell outside the supply regulations. The Export and Import Permits Act was used to establish import restrictions, which have been extended from only butter in 1951 to cover almost all dairy products today. These regulations have been admissible in an era of growing free trade through the presence of the article eleven exemption for agriculture in the GATT.  27  Until 1988, the government calculated the milk price through a “Returns Adjustment Formula” (RAp) that tried to balance the farmers’ cost of living and the cost of production. On top of this there is still a high degree of ad hoc adjustment possible by the administrative authority. The RAF was established in 1975, and was intended to be in effect for only a couple of years. However, it was not replaced until 1988. In 1988 a system was established to estimate production costs based on provincial surveys. However, the government has continued to be actively involved in the price determination process, with significant opportunities for political involvement [McKinley, 1990]. This has meant that the price is still not reflective of the costs. In general it is believed that the price which is paid to the farmer is significantly higher than their cost of production. The fact that farmers are willing to pay substantial amounts of money to purchase quota rights supports this idea. When a farm price has been decided upon, this value is then used in another formula to arrive at a price that reflects a fair return to the processor as well. The formula is based on a support price for butter and skim milk powder, and a guaranteed processing margin. At this price, the CDC buys any surplus that exists, establishing a price floor. With the MSQ program the stocks are usually quite small; the CDC essentially stores the product for resale when supplies are short. This price floor then becomes the price against which all other products are compared, and the returns generated by other dairy products will be at least as great as that possible on butter and skim milk powder. Target prices are also established for other processed dairy products, and have usually been slightly above those for butter and skim milk. A subsidy is added to the milk price that the farmer receives to make up any difference between the support price floor established for the processor, and the target return to the producer. From the price that must be paid to producers, a levy is deducted to cover the cost of disposal of the surplus milk. This ‘within quota’ surplus is the result of the ‘sleeve’ that is built into the system to allow for year to year fluctuations in price and demand [Barichello, 1981]. The disposal is usually accomplished by exporting the milk products at a loss. This levy is distinct from the producers individual levy that must be paid on that part of the milk that is produced in excess of the quota amount.  28  The actual price that is paid to the farmer is determined at the provincial level, where the processor price is set. On a practical administrative level, the dairy industry in each province is regulated by marketing boards that sit in those provinces. The federal legislation, the Agricultural Products Marketing Act, specifically authorizes the transfer of the federal regulatory power to the provincial board for the purpose of administration. The complicated nature of this system means the price paid to producers is almost totally unrelated to the consumer’s demand in the market. The combination of the price control and the market size regulation ensures that the milk price will maintain a roughly constant relationship with the cost of production and the cost of living. The guaranteed market and the security of ‘sufficient’ returns accomplish the main goals as seen from the farmers’ perspective. From the consumers’ perspective, one sees a stable price, and a constant supply of dairy products from year to year. Whether or not this is in the consumers’ interests depends on what alternative sources of the product exist, and at what cost these can be had. The pricing system does appear to be biased in the farmer’s favour. The formula includes a profit margin that is meant to be ‘fair’ and price changes tend to lag behind the cost of production.  22.2) The British Columbia Government The Marketing Boards The provincial authority in BC is derived from the British Columbia Natural Products Marketing Act, which provides for the delegation of a broad range of powers to the designated board, the British Columbia Milk Marketing Board (BCMMB).  29  the board is vested with the power within the Province to promote, regulate, and control in any and all respects the production, transportation, packing, storing, and marketing, or in any of them, of a regulated produc4 including the prohibition of production, transportation, packing, storing and marketing, or any of them, in whole or in par4 One of the powers that is specifically designated to the board is the ability to establish production quotas and regulate their transfer. The act specifically prevents the board from assigning a value to the quota it issues, but there is nothing in this act that prevents producers from paying each other for the transfer of these rights. The crown appoints three people to the BCMMB, one of whom must not be involved in the dairy industry. Beyond one required annual meeting, they hold meetings whenever necessary to deal with any issues that come up. A small administrative staff handles the day to day responsibilities, which include:  1.  Licensing ofvendors and producers;  2.  establishing the classification ofqualifying milk on a basis of utilization;  3.  establishing values at which vendors shall account to the Board for qualifying milk used in each class;  4.  establishing daily milk quotas for producers  5.  and other miscellaneous duties.  The regulatory specific are contained in two ‘general’ board orders, General Order 31 and General Order 133. General Order 133 covers the specifics of the regulation of the provincial fluid milk market. General Order 31 governs the way in which the federal powers are administered locally by the local board. The federal program is administered by dovetailing them with the fluid milk regulations, for the most  30  part. Until 1991 the quota that governed industrial milk was not even traded independently. It was transferred in the same proportion to the total held as the holder transferred fluid quota. Several principal features of General Order 133 are the formula price, the pooling of the milk supply, and the volume constraint on producer output. The BCMMB establishes a basic accounting value for all ‘qualifying’ milk using a specific price formula. The formula is established in recognition of the additional cost of producing a constant year-round supply. The producer price has been established at a level which adequately compensates for the higher cost ofproducing a consistent year round supply, available when and where the consumer desires it [SSCA, 1978, p.32]. The formula uses an average price for a ten year period beginning with the previous year. This base is adjusted by a weighed average of three general economic indices and four agriculture specific indices. This accounting value is used to generate a blended pool price according to the actual utilization of milk within the board’s administrative area. Milk is divided into seven different classes under the BCMMB orders. Class 1 milk is used in ‘fresh’ fluid form in the local British Columbia market, and is allocated the full accounting value. The remaining milk is divided into six classes covering different combinations of final use and final market. These classes are allocated a lower price than the class one price. The final price the farmer receives is the sum of the milk prices weighted according to the share of the final use of all milk, the utilization adjusted price. Processors must pay the utilization adjusted formula price to the farmer. The specifics of the utilization of the individual plant or processor are irrelevant from the farmer’s perspective. If a processor utilizes milk for a lower average class value than it pays the farmers, the milk board makes up the difference. If it uses its milk in a higher class, it must pay the difference to the milk board. The result is that each processor pays the same price, ignoring administrative costs involved in the adjustments, for all the milk it uses from a particular class.  31  To make the program effective, the Milk Board must have a tight control on all aspects of the production, distribution, and utilization of the milk produced in and brought into the province. To accomplish this, the Milk Board collects extensive details from all ‘vendors’ of dairy products in the province. These details include all volumes of milk received, final utilizations of all milk, i.e., classes of product produced, and detailed financial information including statements of profit and loss [BCMMB, 1990]. With these details, the board can determine if all the milk is going to its best use, and take punitive action if milk is being channeled inappropriately. The aggregate milk supply is regulated through the quota system. Each licensed producer is allocated a share of the milk market that remains with the producer until a valid transfer takes place. The total quota in the province is determined by the milk board through a projection based on factors such as population growth, demographic changes, demand shifts, etceteras. The quota amount is allocated to existing quota holders on the basis of their current share of the total quota.  The Milk Board ofFluid Quota is a daily production quota which is issued to producers by the BC Milk Board. The quota represents a producer’s share of the fluid milk market and is based upon his[sic] total daily production during the four month period in which aggregate milk production in each Milk Board Region is at a minimum. The aggregate allocation of milk quotas to each region, in any year, is equivalent to 120 percent offluid milk sales during the quota earning period of the previous year. Because milk quotas are based upon daily production, this 20 percent margin provides for seasonal and transitional fluctuations in production an4 also, for fluctuations in day-to-day demand [SSCA, 1978, p.32]. Farmers who supply more than their quota allocation are charged a levy, and farmers who consistently ship less than their quota allocation will have their allocation reduced. Reductions are distributed to those who are consistently overproducing. The combination of the overproduction levees and the threat of quota loss through underproduction ensures that the total supply on the market remains very close to the amount specified by the milk board.  32  The pooling of the milk is accomplished through the authorization of processors to request milk from other processors. The BCMMB orders establish a priority of use for class one milk. If a processor requests milk for class one use from another processor that would use it for a lower class, the milk must be transferred. The board may  ...  direct that excess quantities of qualifying milk receive4 or to be  received by a vendor shall be transferred by that vendor to another vendor requiring the millç provided that Class 1 utilization has the “highest priority” Below class one transfers, the hierarchy is based on the accounting value of the milk in that class. Most processors generally regard supplying the fluid market as the highest value use for milk, so this provision generates a situation much like an anonymous pool from which the individual processors draw milk. They are not specifically tied to any producer. The mitigating factor in the orders is that the receiving processor must compensate the supplying processor for the cost of handling and transporting the milk.  2.2.3) The Farmer When the program was introduced, quota was assigned on the basis of the producers share of the aggregate production during that part of the year when the production level was the lowest. Since then, the quota that a farmer holds can change in two ways. Firstly, if the farmer is consistently below the quota level over two “quota months,” the amount of quota held can be reduced to the fanner’s average production level during these periods. This has seldom happened, indicating that there is likely a shadow value associated with the supply restriction. In the second place, the farmer’s individual quota allocation can be increased as a result of increases in the overall demand for the product [SSCA, 1978,  p33].  The  overall demand for milk in British Columbia has been growing steadily for almost as long as the program has been in place. The main driver in this trend has been the continuing growth in the population of the province.  33  The benefit that accrues to the farmer from the system comes in several parts. Two general classes of benefit can be identified. The first class of benefits are those that arise from the difference between the supply managed market and the unregulated market. These benefits include the security the system generates and a price level that is higher than that which would prevail in a free market. Security is important to the farmer since the production level is planned at least one year in advance. The farmer must make the production decision in the face of market risk, and the greater this risk, the more resources are going to be devoted towards protecting against it. The higher price required as ‘adequate compensation for year round production.’ guarantees the farmer a level of profitability from operations on the farm. As far as the price is above the cost of production and the opportunity cost of the assets committed, the farmer will be receiving a benefit. The second class of benefits are those that accrue as a result of environmental and structural changes while the system is in place. These benefits include the growth in quota that accrues to existing dairy farmers and the degree to which the price more than ‘adequately’ compensates the farmer for what is produced. Quota growth comes from the increase in the volume of milk represented by the quota that results from growth in the market. This benefit accrues unequally across Canada. In regions such as the Maritimes and Saskatchewan which are facing a declining population, the farmers quota allocation will be falling as the market contracts. This compares to provinces like Alberta and B.C. that are experiencing population growth. The extraordinary compensation results from factors such as the lethargy of the price formula to reflect changes in cost structure, and the ability for political lobbying to generate and protect economic rents that are accruing to the farmers. These extraordinary rents are believed to be largely capitalized into the price of the quota, and the high price which this scarce ‘input’ commands indicates how large these rents might be. For those producers who have been in production for more than five years, the quota becomes a tradable asset. Subject to certain volume requirements, farmers can freely trade quota between themselves. In 1978, any producer who was transferring Fluid Milk Quota also had to surrender an equal percentage of  34  MSQ to the Milk Board. There is no restriction on whether the quota is to be sold with cows, or with a farm. The province is split up into a number of quota regions, with quota being freely tradable within the region, but no transfers possible between regions. The increases in quota that accompany increases in demand are also tied to the quota region within which the demand increases. This means that the benefit which accrues to the quota holder associated with increases in the demand is tied to the region where the quota applies. The fluid quota technically remains the property of the Milk Board, and as such the milk board does not recognize the existence of a price for the quota. However, farmers are willing to pay each other to acquire the right to ship milk that goes to the person who holds the quota. Quota is the most scarce resource in the industry, there are no substitutes. We expect that with many potential buyers, much of the will be capitalized into the quota price. The size of this price will reflect the security benefit, the expected demand change, the expected difference between the cost of production and the amount of revenue generated by producing milk, and any benefits generated through the tax system. The forecast income stream will be discounted at a rate that reflects the opportunity cost of placing capital in quota and the risk that the program will collapse. Trading of quota is carried out either by direct trade between farmers, or through agents such as livestock traders. The farmer is restricted to hold an amount no more than one percent of the total provincial production [Barichello, 1987]. At present there are no farms in the province even close to this size. In 1991, the average dairy farm in BC was milking close to 80 cows. To have one percent of the provincial milk production would imply a herd of more than 700 cows with the average per cow production. The largest herds in the province are only about half of this size. If we look to the US, we see herds of over 1000 animals. The difference in herd size can be explained in part by different production conditions that favour large size. However, the presence of much larger operations in parts of the US that are geographically similar to areas of Canada would indicate that there is also a policy effect.  35  Beyond brokerage costs, there is a tax of between 9.1 and 20 percent transfer assessment, dependent on the nature of the transfer, which returns quota to the milk board. Through this process, the milk board acquires a stock of quota which it uses in its ‘Building Program,’ providing fluid quota for farmers entering the industry. Because of the attractiveness of the industry, there is a long waiting list of people desiring to enter [Barichello, 1987].  2.3) The Economic Consequences The economics of supply management are quite straightforward. Figure 2-la shows the unregulated market outcome. The quantity of milk is detennined by the intersection of the supply curve and the demand curve. An amount qi of milk is produced and sold at a price p. Under supply management, the total quantity available is restricted by a central authority. In figure 2-lb the supply has been restricted to  . 8 q  Consumers are willing to pay Pc for this quantity of milk, but it has a  marginal cost to the producers of only Pp. In the absence of a program, producers would produce qf q -  more milk, all of which would be consumed at the market price of p. In terms of consumer surplus, the area represented by A is transferred from the consumers to the producers. Areas C and D are deadweight loss. The quota price determined by the shadow value to the producer of another unit of production. In the diagram, the supply curve represents the industry average marginal cost to the producer of producing milk. Consistent with controlled pricing, producers take the price as exogenous. At the margin, the producer would receive the difference Pc  -  p, if another unit of output could be produced. Since another  unit worth of quota would entitle the farmer to ship this unit of milk, the price of the quota is going to have an upper limit given by the present value of this price difference.  36  S  $  $  PC  /  A  Pt Pp D  ND  q  Q  0  Figure 2.1.1: Free market situation  Figure 2.1.2: Supply managed market  Following Barichello [1981, 1987] we can identify some further consequences. Table 2-1 presents a list of the positive and negative effects of the dairy industry regulation in Canada taken from Barichello [1987]. Since the degree to which the programs have changed is minimal, these consequences are basically unchanged. A couple of subtle benefits can be included. The most obvious one is the fact that the supply management program involves little direct burden on the government treasury. However, it has been argued that it would be more efficient if government were to subsidize the farmers directly [Barichello, 1981; Grubel, 1977]. It is also difficult to measure efficiency gains because when less resources are needed to protect against risk. The stability and predictability of the milk price has meant that the assorted insurance programs and periodic cash infusions which characterize other sectors of Canadian agriculture are absent. Further, since resources are not being devoted to risk reducing activities, the resources in place in the industry are likely to be employed more efficiently.  37  SUMMARY OF POSITIVE AND NEGATIVE EFFECTS OF CANADIAN DAiRY REGULATION  A: Program Goals 1. Increased dairy farmer income  -  -  2. Size of dairy sector  3. Stability  temporary income enhancement  production levels preserved compared to free trade more dairy products in Canada are “made in Canada” due to self sufficiency policy for fluid milk and butterfat  5. Equitable across producers.  smaller producers and recent entrants share few of these benefits. level of total milk production slowly falling. number of milk producers continuing to fall at a steady pace.  -  -  -  -  production and price patterns continue to be stable production and price patterns are more predictable  this stability may be only partly due to the present regulation. policy instability in early years of the program.  -  -  4. Continuous, adequate supply of milk products at reasonable prices.  -  capital gains to established producers, mostly to larger producers.  -  supply is continuous  -  -  -  -  -  industrial milk program rules apply to all provinces  -  -  -  B. Efficiency Effects 1. Farm level welfare costs  -  product choice narrowed. consumer price not reasonable compared to alternatives. unfair to entering and expanding producers unfair to fluid milk producers unfair to producers in more efficient milk producing provinces. inefficient, arbitrary allocation of industrial milk production across  provinces. -  -  -  -  2. Processing, Distribution, Retailing Sector.  -  -  -  Table 2.1: Consequences of supply management.  38  inefficient, arbitrary allocation of industrial milk production within provinces where MSQ is not traded. reduced future supply to dairy farming of best entrepreneurial talent. biased genetic selection in dairy cattle breeding. institutional policy risk of unexpected changes in policy rules. inefficient milk allocation system to processing plants. reduction in competition. distorted advantage to butter production  2A) Summary  Govermuent involvement and regulation has been a constant feature of the dairy industry for almost one hundred years. Regulations were initially adopted to counteract the hardship that farmers were facing as the industry shifted from producing solid milk products with only a limited perishability hazard to the production of fluid milk for the growing urban market. The experiences of this transition period are still the paradigms against which the industry participants struggle, although it is unclear whether they are still true. Canada has adopted a supply management system to deal with the perceived problems of the dairy industry. Supply management does provide stability, a continuous flow of safe, high quality product, a degree of self sufficiency, with little effect on the government treasury. These positive factors must be weighed against the costs, which include reduced total milk consumption, less efficient resource allocation, stagnation against changes in demand and comparative advantage, and a loss of entrepreneurship. At present, supply management faces a number of challenges. Consumers are unwilling to accept the high price that the program generates. Legal challenges have resulted in a vacuum in the legislation, and the General Agreement on Tariffs and Trade has rendered several of the policy tools on which supply management relies inappropriate. In light of these challenges, it is time that the supply managed sectors reevaluate their goals, and consider various alternatives to accomplish them. Canadian dairy cooperatives developed within this policy environment, and it is under the present regulations that they must operate. The subtle differences in competitive conditions and policy environments have lead to pronounced differences in the degree and competitive strength of dairy cooperatives between the provinces in Canada. In Ontario we find a virtual absence of marketing cooperatives for dairy products, while in Quebec they hold by far the largest share of the market. In the following chapters we will explore how these interactions are manifested.  39  Chapter 3: The Unrestricted Problem In this chapter we begin to develop a mathematical model of a cooperative. The principle objective of the analysis is to see how supply management impacts on the relationship between a cooperative and its membership the interaction between a cooperative and an IOF competitor. Trade will be ignored for the duration of this analysis The impact of trade is only on the magnitude of the interactions we are exploring here. This chapter begins by building the model without restrictions. We define the technology and the objective functions of the economic actors, and see how they interact. In the next two chapters we explore the effects of the volume restriction and the pool price in isolation. Chapter six pulls all the elements of supply management together.  3.1) Background. The theoretical analysis of cooperatives first developed around the ‘Illyrian’ firm, a producer cooperative, the most quoted work being that of Ward [1958]. As reviewed by Bonin et a! [1993], the objective of Ward’s producer cooperative is to maximize the patronage dividend that can be returned to the labour input. Ward finds that the equilibrium solution is not Pareto-efficient, with the market employing more labour than competitive firms would. When the labour input is allowed to move between the producer cooperative and its competitive rivals, this resource moves until it generates the same marginal return with all employers. The producer cooperative restores the competitive solution. If the producer cooperative is allowed to restrict membership, then the competitive yardstick result does not occur [Bonin et a!, 1993; Cotterill, 1987]. Restricting membership allows the cooperative to short the market as an IOF would, generating a higher than average return for its membership. The membership will not allow new members to join, protecting their higher return. This form of cooperative behaves much like its IOF rivals.  40  The first analysis of marketing cooperatives to be widely quoted was that of Helmberger and Hoos [1962]. Helmberger and Hoos modeled a cooperative that marketed a homogeneous product on behalf of its member finns. Their analysis also chose dividend maximization as the objective of the cooperative. They demonstrate that the optimal point of operation is consistent with the minimum cost point for those inputs not supplied by the members. In the Helmberger-Hoos cooperative, the member firms choose their own production, which they market collectively. The equilibrium outcome occurs where the marginal production cost of the members is equal to the net revenue—surplus per unit received—of the cooperative. A competitive yardstick result is generated analogous to the labour managed case. Hardie [1968] developed the analysis of the marketing cooperative to the multi-input and multi-product case. He used a linear programming approach to show that the shadow value associated with each input can be thought of as an appropriate member return. Hardie’s model was developed from the single input case built by Helmberger and Hoos, showing that a shadow value can represent member returns in the single input case when the production level is taken as parametric. Imperfect competition between cooperatives and between cooperatives and IOFs is only beginning to be addressed [Ireland 1987]. The majority of the research has dealt with the labour managed firm, in particular as the former communist countries are liberalizing. Law and Stewart [1983] show that when a labour managed firm and a profit maximizing firm are in competition, the profit maximizing firm will assume the Stackleberg leader role, and the cooperative the follower role, and neither will want to change position. In general, an IOF prefers to have a cooperative as a rival in place of another IOF. Objectives different from dividend maximization have been proposed [Cotterill, 1987; Ireland, 1987; LeVay, 1983]. Principle among these alternative objectives is maximizing farmer welfare. Overall farmer welfare is not maximized at the maximum dividend point. It is also not maximized at the intersection of the supply function of the member agents and the net revenue product of the cooperative. Typically the welfare maximizing point still involves some market shorting, and is unstable while members are free to choose their own production level.  41  Sertel [1991] investigated the relationship between a workers enterprise and a profit maximizing firm in imperfect competition. Sertel’s model moves from a standard capital and labour technology to a labour only production function by demonstrating that capital will be chosen optimally and can therefore be ignored. Sertel then focuses on the labour input, which is supplied by the members. The labour managed firm has a restricted membership by demanding compensation for workers who are replaced. This ‘market’ for membership rights mimics a capital market, so that the behavior of the cooperative and the TOF become indistinguishable. The point where the membership rights have the largest value generates the same labour input level as an IOF with the same technology chooses. Our analysis follows a construction very similar to that of Helmberger and Hoos, and our conclusions reflect for a marketing cooperative many of the results that Sertel finds for a worker enterprise. We assume that there is only one homogeneous, undifferentiable product which the members deliver to their cooperative. All other inputs are ignored. The stochastic nature of member production, differences between members, and variations in output price are also ignored. These factors are important, but would complicate the many features that are already part of this analysis. The objective of our model is to investigate the effects of the various features—price pooling, price control, and aggregate volume restriction—on the interaction between a cooperative, its membership, and a competing IOF.  3.2) The For Profit Processor. There is only one input in the processing sector, the raw milk. All processors have the same technology. For the investor owned firm (IOF) we write the profit function as:  3.1  In this equation,  t  is profit, the difference between revenues, the product of the output price  amount of output produced using the production process input price  f (v,,), and the costs, the product of the raw  m and the amount of raw product used by the firm,  42  p and the  V.  This is a somewhat atypical way of modeling the behavior of a firm. Normally one would minimize cost subject to a production constraint, and then optimize over the amount of output the firm chooses to produce. We are modeling the input decision of the processor, and therefore ignore the output market. It is easier to see the effect of the supply management restrictions on the input side. We are modeling the demand behavior in the milk market, where our consumers are the raw milk processors, and their utility is maximized by maximizing their profits.  The firm controls the amount of input,  3.2  v,, used. The first derivative of the firm’s profit function is: ,  Tht/öv =p(af/av)_m  To locate the profit maximizing input level, we set this equation equal to zero. As expected, we find that the profit maximizing solution is to set the marginal value product equal to the price of the input.  3.3  p(af/avj=m  We assume that the production function is concave and decreasing over the relevant range, guaranteeing a diminishing marginal product. We assume the functions are smooth, monotonic, and continuous, ensuring only one solution Finally we assume that the second order conditions are satisfied, so that our solution is a maximum. Figure 3.1.1 shows what the total revenue and total cost curves might look like with respect to the amount of milk the processor receives. We suppose that at low production levels, the plant would be running inefficiently and little of the milk received would be transformed into the final product. Efficiency would increase until some optimal level was reached, after which it might decline. Total revenue is a function of both the technical efficiency of the production process and the effect of the firm’s production on the output price. Our functional specification has so far contained a fixed output price. With a fixed output price, the shape of the revenue curve will be determined by the production technology alone. As the profit  43  $  $  Total Revenue MW  Total  Cost m  v,  v,  milk  Figure 3.1.1: Total revenue and total cost as a function of total milk received.  milk  Figure 3.1.2: Marginal value product as a function of total milk received.  function has been defined, the total cost is going to be the product of the milk price and the amount of milk used, a straight line through the origin. The marginal value product for this relation is shown in figure 3.1.2. It is initially climbing, and crosses the line representing the milk price once before reaching a peak. From its high point, it falls, crossing the milk price line once again. The milk price line is the marginal cost curve that corresponds to the total cost curve in the first figure. There are two intersection points. The left point is an unstable equilibrium. At this point the last unit produced is finally breaking even, and no profits are being made. The right point corresponds to the profit maximizing production level for the firm.  3.3) The Independent Farmei The independent farmer’s profit is not linked to the processor’s profit. This farmer takes the offered price as given. For this fanner, our model assumes a well behaved cost function, convex and increasing in the amount of milk produced. It is probably not too unrealistic since there is generally felt to be an optimal production level for most of the assets held by the farm, with increasing costs away from this point.  44  $  $  yr  vf  milk  Figure 3.2.1: Total revenue and total cost as a function of total milk produced.  milk  Figure 3.2.2: Marginal value product as a function of total milk produced.  Relying on these assumptions, we propose the following objective function:  3.4  The profit t  rtj’=mvf_c(vr)  f  is equal to the revenues  delivery to the plant, less the cost  3.5  mv[ generated by fanner i producing  yr  units of milk for  c(vf) of producing it. The first derivative of equation 3.4 is:  mtr/ovr =m_ac(vf)/ovf  Setting 3.5 equal to zero, we get the traditional relationship where marginal cost is equal to marginal revenue, the price.  3.6  m  =  ö c(vr )/ovr  Figure 3.2.1 shows the total cost and total revenue curves for a hypothetical farm. On this diagram, the vertical distance between TR and TC is greatest at the production level given by  v/’. Profit is represented  by the distance between these two lines. In figure 3.2.2 we have the marginal conditions. The convex and  45  monotonically increasing marginal cost is consistent with the assumptions that it is well behaved. The price m which the fanner receives represents the marginal revenue. To locate the optimal production level, we set the marginal revenue equal to the marginal cost, and reproduce the profit maximizing point given by the pmduction level  vI.  3.4) The unregulated market with independent agents. Now that we have defined the objective functions for the actors in this market, we need to study their interaction. The objective functions for the JOF processor and the independent fanner shipping milk to this processor are: 3.1  (v  m  max{mv! vi,  In this construction,  —  )  —  mv  C(V  }  IOF processor optimization problem  )}  Independent fanner’s objective function.  v is the amount of milk the processor chooses to purchase. It is purchased at a  price of m, is processed using a technology represented by  fanner produces yr of output at a cost of  f(v), and the output is sold forp. The  c(vf), and sells it to the processing market for m. The model  has been built under the assumptions of an unregulated competitive market. The agents treat the prices as predetermined, and believe that their own actions have no effect on the actions of the other agents in the market. The first order conditions we generated are:  3.3  The for profit processor.  p(af/av)= m 3.6  The independent fanner. m  =  a c(Vf )/avf  46  $  Supply  Demand  milk  Q Figure 3.3: Supply and demand derived from aggregate MW and MC curves.  This result is precisely what we expect between atomistic agents in a competitive market. The IOF processor chooses ifs input level such that the marginal value product equals the marginal cost of the input, the milk price. The farmer chooses to produce such that marginal cost equals marginal revenue, again the milk price. The market interactions between these agents balances their individual optimization problems to clear the market. If we assume a large number of processors and a large number of farmers all acting independently, then all will take the price as given, and we get the competitive outcome, the intersection of the market supply and the market demand curves. Figure 3.3 shows the unregulated market when there are a large number of buyers and suppliers. The demand curve D is the aggregation of the marginal value products curves for all the processors buying milk. The supply curve S is the horizontal aggregation of the marginal cost curves for all the producers. The intersection of these two curves locates the competitive solution. At this point, total farmer production is Q, which is purchased by the processors for the price m. There are no economic rents available to anyone.  47  $  $  1 TC 0 TC  mW’1  Q  milk  Figure 3.4.1: Supply curve for raw milk. total revenues.  1Q Q 0 Figure 3.4.2: Monopolistic processor total  milk cost and  If there are a small number of processors the competitive outcome is less likely. If we go to the monopolistic extreme with only one processor, then the processor has the power to set the price along the producers’ supply curve. The total cost that the monopsonist must pay is equal to the product of the price and the amount of milk purchased. In the competitive market the price is exogenous to any agents decision. With one purchaser, the price that needs to be paid to generate a particular level of output is given by the supply curve. At the margin, this means that the monopsonist faces an upward sloping marginal input cost curve rather than the horizontal one that characterizes the competitive case. Graphically, the total cost that the processor pays, at any given level of output, is equal to the area of the rectangle bounded by the output level on the right, the price line above, and the axes (figure 3.4.1). If the price is taken as given, the area of this rectangle increases only in response to increases in quantity. The top of the rectangle is fixed at the market clearing price. For the monopolist, the area of the rectangle increases both as a response to the rightward movement of the amount of milk purchased, and with the upward shift of the price line. This  48  upward shift occurs because for any particular level of milk purchased, the monopsonist need only cover the marginal costs of the producer represented by the supply curve. When we construct the total cost curve, we see it starting at the origin, but initially increasing at a slower rate than the linear total cost curve of the fixed input price (figure 3.4.2). The TC 1 curve increases in slope, and crosses the TC 0 curve at the competitive market solution. At the competitive solution the 1 curve is steeper than market price is also on the supply curve. However, at this point the slope of the TC the TC 0 curve, and therefore also steeper than the TR curve. The maximum profit point must lie to the left of the competitive solution.  3.5) The Cooperative Processor. The cooperative processor is a little different. We assume the cooperative distributes all of the profits generated to the members, and that there is no member capital within the cooperative to worry about. This cooperative is also unable to carry losses or profits between periods, allowing us to avoid the issue of distinguishing between farmer cash receipts and cooperative performance. A cooperative must decide on other fixed and variable input levels. We assume these inputs are chosen consistent with our chosen objective, and are therefore ignored. All production costs consist of the cost of the milk input supplied by the farmers, making it easy to model. We first take the independent processor’s objective function and set the profit equal to zero. The total amount that can be distributed to members,  mv, is equal to the total  revenues generated.  3.7  mv =pf(v)  There are a variety of objectives the cooperative could pursue. It might maximize aggregate farmer welfare by trying to force the competitive outcome. It could maximize member return as a vertically integrated firm, including the costs at the farm level. Or it might maximize the patronage dividend, ignoring the costs at the farm level. We model only dividend maximization because it is the simplest case.  49  $  $  milk  0 1 Q  Qi Qo  milk  Figure 3.5.2: Cooperative marginal value product and average value product.  Figure 3.5.1: Cooperative total revenue  Some conclusiom are dependent on this specification. However, provided that the cooperative is pursuing an objective which offers the members an effective price above the competitive market price, our general result continues to hold; the competitive yardstick is not guaranteed under supply management. In its simplest form, maximizing the patronage dividend is equivalent to maximizing the price the cooperative pays for the member supplied input.  m = pf(v )/v  3.8  To find the amount of input associated with the maximum price that can be paid to the fanner, we can take the derivative of this relation with respect to the quantity of milk processed. This derivative is:  3.9  ôm  1  —=  avc  —pof (vj/ov ----pf(v) vc  Setting 3.9 equal to zero and eliminating terms we simplify it to the following:  3.10  paf(v )/av  =  50  (1/va )pf(v)  The left hand side of this equation is the marginal value product. The right hand side is the average value product, the total revenue, divided by the total amount of input used. Graphically the total revenue curve is the same as that for the IOF, as shown in figure 3.5.1. However, it does not make sense to include a cost curve, since the profits are totally distributed against the milk used in production. The marginal curves show where the cooperative would operate, if it was able to maximize its objective function. The maximum price that the cooperative can pay for the milk received is the maximum of the average value product. This point occurs where the MVP curve intersects the AVP curve in figure 3.5.2, at an output of Q. For the given curves, this point is below the free market equilibrium which would occur at the intersection of the price line and the MVP, at Qo.  3.6) The Cooperative Member Farmer. The cooperative member farmer is assumed to have the same objective function as the non-member. The only differences are that the ‘price’ the farmer receives is the patronage dividend, the total revenues of the cooperative divided by all member production.  =[pf(vj/v]v —c(v:)  3.11  The amount of milk that the cooperative processes, v, is equal to the sum of all the milk the members choose to produce,  2 v:.  This implies that öv /öv  is equal to one. For the moment we are ignoring  the effect this farmer may have on the production of other farmers. Evaluating the derivative of 3.11, it can be written as:  3.12  ör /Ov  =  -—pf(v)÷  .i{[p 13f(Vc)]  —  -_pf(v  Setting this relation equal to zero, and rearranging, we get the following:  51  )}  —  oc(v:)/av:  pf(vjv 3.13  vc  Patronage Dividend  vc  pf(v)  v  ôf(v)  vc  vc  t3v  Dilution  Output +  Effect  Effect  öc(v) Marginal  =  Cost  In equation 3.13 we have split the left hand side of the equation, the cooperative member’s marginal revenue, into three terms. The patronage dividend is the per unit revenue the member receives. The dilution effect is the impact that a change in this member’s production level will have on the share of the patronage dividend this farmer receives. If this member chooses to increase production, then the existing cooperative revenues will be diluted over a larger patronage, reducing the share captured by this member. The output effect is the amount by which revenues increase in response to an increase in output. If the extra production of the cooperative increases efficiency, then revenues available for distribution will be larger, and the patronage dividend will increase accordingly. The member’s optimal production point sets the sum of these effects equal to the marginal cost of attaining the farm’s output level. If we were strictly correct, we would need to acknowledge that the dilution effect also affects the other members, so they would be expected to reduce their own production. Therefore, the absolute magnitude of the dilution effect will be less than shown here. 3.13 is easiest to interpret at the extremes of one infinite membership and a single member. If there are an infinite number of member farmers in the cooperative, then the ratio V’ /v is equal to zero. The only factor the farmer considers is the size of the patronage dividend. This is typically assumed [Cotterill, 1987  ...], with the result that the farmer acts as if the patronage dividend is a predetermined price.  If we have only one only supplier to the cooperative,  v’ /v equals one. The only factor that the farmer  considers is the marginal value product of the milk. The farmer’s own marginal revenue is equal to the marginal value product to the cooperative of the milk shipped, and the farmers decision will completely internalize the final market. This corresponds to vertical integration by a single farmer. A typical member faces is a linear combination of these two extremes.  52  $  1 v  v  milk  Figure 3.6.1: Cooperative member total revenue and total cost,  milk  Figure 3.6.2: Cooperative member marginal revenue and marginal cost.  The farmer’s individual marginal revenue is changed when he/she belongs to a cooperative. Since we are assuming that the industry is earning some rents, the cooperative is able to distribute a return to the fanner above the market price for milk. This generates a steeper total revenue curve as shown in figure 3.6.1. However, the total revenue curve is no longer linear. As the farmer increases production, the average value product of the cooperative falls. Since the price that the farmer receives for the milk is drawn directly from the average value product, the slope of the total revenue curve falls as the farmers individual production increases. The degree to which the AVP falls depends on the scale of the farmers change in production relative to the overall product the cooperative receives.  37) The Unregulated Market with a Cooperative As was done for the unconnected agents, we now interact the objectives of the cooperative and its members. The objective functions we specified are:  53  3.8  max {pf (v )/v V  3.11  }  max{[pf(v )/v }‘“ vc  3.14  Cooperative processor objective function. V = choice of input level. —  ” 1 c(v  )}  Member farmer’s objective function. =  member’s production level.  Equilibrium condition  v,  The objective of the cooperative is to maximize the size of the patronage payment. The objective of the member is to maximize profit. To make things balance, the cooperative is required to process all of the milk that is delivered to it by its members, as expressed in the equilibrium condition shown in equation 3.14. We assume throughout that membership and member production is unrestricted. As for the simple IOF case, we are assuming that the output price is fixed. The first order conditions we developed for this problem are: 3.10  The cooperative processor.  paf(v)/av 3.13  =  pf(vj/v  The member farmer.  v  v  v  öv  v  ac(v:) ôv  The objective of the cooperative is satisfied when the average value product of the cooperative is set equal to the marginal value product. The member solves its optimization problem by choosing a marginal value product that is equal to a farm adjusted patronage dividend. The farmer’s have the ultimate power over the level of input the cooperative will process. If we have a labour managed firm, then membership could be adjusted or restricted until the highest possible wage is being paid [Bonin, 1993]. However, this violates the principle of open membership, and could clearly not be maintained if new labourers can freely enter. We are studying a marketing cooperative, where members are free to choose their own production level, making it very difficult to be stable at the maximum of the AVP. Farmers are assumed to the MR and MC they see on the farm. If there are many farmers, this occurs where the marginal cost of the average farmer is equal to the average value product of  54  $  $  Qo  Qi  Figure 3.7.1: Marginal and average value product against member supply.  milk  Qo  1 Q  milk  Figure 3.7.2: Aggregate marginal revenue and supply curves.  the cooperative. This is beyond the patronage dividend maximizing point the cooperative would like to be operating at, assuming it is trying to maximize the patronage dividend it can pay. As we build the model, we introduce a milk pooling system that breaks the direct link between member production and the amount the cooperative processes. This policy instrument breaks the competitive yardstick result we find at this level. Graphically, figure 3.7.1 combines the marginal value product and the average value product with the member’s supply curve. The market solution is shown as Qo in this figure. If we assume that the farmer does not internalize the cooperative incentives, and that there are profits available in the industry, then the average value product represents the farmer’s marginal revenue. Fanners will continue to join the cooperative, expanding the member supply curve. This will continue until the supply curve has expanded 0 to S from S 1 has been reached, the average value product has been lowered to the point where . When S 1 it is equal to the market price at the input level that is finally established. At this point, the farmers are indifferent between belonging to the cooperative and no belonging.  55  However, if the number of farmers is small enough that the individual fanner somewhat internalizes the cooperatives problem, the average value product no longer represents the price that the farmer optimizes against. The fanner’s marginal revenue will instead lie along a line that is a linear combination of the MVP and AVP curves. This line is labeled MR in figure 3.7.2. The amount by which the farmer’s join the cooperative will be somewhat lessened by this effect.  3.8) The unregulated market with one cooperative and one for profit firm. The interactions when there is a cooperative in a market can range from one cooperative and one for profit firm, to one cooperative among a very large number of for profit firms. In the later case, we would expect the free market outcome to occur. Processors will compete for the output of the farmers, to the point that all rents are exhausted. It is questionable if a cooperative would ever appear in a competitive market, but if there was one, it would not create any economic benefit for its members relative to the free market price. This is not be particularly interesting to explore. The case that we explore has two processors, an IOF and a cooperative. In parallel we develop the traditional duopsony case to illustrate the differences. As above, we assume a simple revenue function for the processors, depending only on the amount of input used. The farmer’s individual actions are aggregated into a supply curve. 3.15  3.16  3.17  max{f (vt) rnax{f (v m  =  m(V)  —  ‘)  m(v + VP )v  —  V  m(v + i  =  }  First oligopsonist’s objective function.  }  Second oligopsonist’s objective function.  Input price relationship.  v,, +  We first consider the two agent oligopsony, two processors with identical technology in the same market, facing an upward sloping input supply.  56  With Cournot assumptions each firm believes that its rival will hold output fixed. The firm optimizes against the slope of the supply curve. To find the firm’s profit maximizing point, we take the first derivative of the objective function.  Mathematically, the oligopsonist believes that  Ov, ‘/Ov  =  0. We impose this condition, and set the  derivative to zero. After rearranging we get:  of Om p———v “ =m Ov OV,,  3.19  The oligopsonist chooses an input level where the marginal value product, adjusted by the effect of the firm’s input decision on the input price, equals the marginal input cost, the price itself. With functional forms, this relation is used to generate a reaction function. The equilibrium solution for the competition between these two firms would be the intersection of these reaction functions. The derivative of the revenue function is assumed to be negative, while the derivative of the supply function is assumed positive. Relative to the free market, the oligopsonist purchases less of the input, so that farmers receive a lower price. Next we introduce a cooperative in the place of one of the oligopsonists. We assume Cournot conjectures for the processors, and atomistic and opportunistic actions for the individual farmers agents and members of the cooperative. Cooperative members assume that the patronage payment from the cooperative is fixed and independent of their actions. The objective functions for the agents in this market are:  3.20  max {f(v) m(v —  3.21  max’mv” vip I.  —  c(v”  +  ,  ).i  Oligopsonist’s objective function.  }  )}  Independent farmer’s objective function.  57  }  Cooperative’s objective function.  3.22  max {f(v )/v  3.23  max {[pf(v )/v ]v’  3.24  m = m(v);  v  —  +  =  c(v:  )}  Member farmer’s objective function.  Input price relationship.  v  We have four agents, the IOF, the independent farmer, the cooperative processor, and the member farmer. The IOF believes its actions have some effect on the milk price, while the independent fanner feels that the farm’s production is too small to have any effect. The objective of the cooperative is to maximize the patronage dividend. However, it has no control over how much it processes. The actual production decision of the cooperative is made by its members, who choose this level by choosing how much to ship to the cooperative. Strictly speaking, the cooperative does not have an objective function. The market must clear, and we apply the law of one price. The first order conditions that derive from this problem are: 3.25  The oligopsonist’s objective. /  ‘  v  +v)—-—— aV  3.26  The independent farmer. at’/avf  3.27  =m_ac(vr)/avr  The member farmer.  or/av;  =  m÷!L{[t’i  I  We have ignored the first order conditions of the cooperative. The important condition that determines what happens in this market is the one price condition. The cooperative distributes all of its surplus to its members, so that if one price is to prevail, the following must hold:  3.28  f(v )/v  =  m(v + v)  Effectively this states that the cooperative patronage dividend must equal the market price.  58  This solution is usually used to argue that the presence of a cooperative in the market will lead to the restoration of a competitive result. The cooperative pays out all its returns to its members, as a result of the zero profit assumption. The members will earn a profit as long as there are rents available in the industry. The IOF must pay a price equal to the return the member receives if it is to acquire any input. New farmers will enter the industry until all these rents are exhausted. At this point, the free market solution has been restored. This need not be the case. The cooperative’s production point occurs where the AVP is equal to the aggregate supply price. However the IOF can choose an input level where it is earning rents against this price. As long as it has an average revenue that is above the market determined price for the milk input, the IOF processor will earn a profit. If we maintain the same Cournot conjectures that we investigated by comparing the two oligopsonists, we get the following first order conditions:  3.29  The oligopsonist’s objective. 3m of /  p—=mv +v)+—v,, (3V Ov  3.30  The independent farmer.  a c(v” )/av[ 3.31  =  The member farmer.  pf(vjifpf(v) vc  vc  vc  Of(v)v’ Ovc  Oc(v) Ov  vc  This is almost identical to the first order conditions for the situation where two IOFs are competing, since under the Cournot conjectures the IOF takes the production of the cooperative as fixed with respect to its own actions. The independent farmer and the cooperative member have not changed their actions. The cooperative does not deal with the milk price, so this is not internalized into the member’s actions.  59  $  $  MYP  m  Vp  milk  VC  milk  Figure 3.8.2: Cooperative in imperfectly competitive market.  Figure 3.8.1: IOF Oligopsonist.  $ Supply  m  v-t-vp  Vp  milk  Figure 3.8.3: Aggregate market supply curve.  Figure 3.8 shows the relations presented above. Figure 3.8.3 is the total supply of milk produced by all the producers, the sum of the member and independent farmer supply curves. In 3.8.2 the intersection of the member supply curve and the average value product determines how much the cooperative will process. Figure 3.8.1 shows the situation for the IOF. If the IOF produces at the point where the marginal revenue intersects the price line, its average revenue is greater than its marginal revenue, and it earns positive profits. This outcome relies on the identical revenue curves for the cooperative and the  60  IOF. The IOF can do even better. It can act strategically to manipulate how the cooperative and its members interact to determine the input market price. As it reduces its purchases of inputs, the price of milk falls. It shorts the market until it reaches its maxhnum profits. Since the IOF is shorting the market, and since we have assumed one price, this shorting translates into a situation where the farmers collectively are receiving a price that is below the competitive result. The sustainability of this solution is a function of the entry barriers. As in most simple introductions to imperfect competition, we just assume that these barriers to entry exist. In the dairy industry returns to scale and market structure may generate entry barriers. The efficiency of larger dairy processing operations, and the fact that most processors have moved to small numbers of plants for their processing, indicates that there are returns to scale over some range of production sizes. This fact is compounded by the ease of market saturation. In BC there are in effect two markets for milk. The large volume market is occupied by large buyers, such as grocery chains and convenience stores, which translates into a small number of customers purchasing large amounts of product. The smaller market of home delivery and small retailers is easily saturated. These facts make it difficult for new firms to start up. We can take this analysis one step further, to the realm of the Stackleberg leader-follower model. In this model one of the firms, the leader, internalizes the reaction of the other finn, the follower. The leader anticipates how the follower will react to its own production decision, and chooses its optimum taking this into account. The objective function for both firms are still the same. However, the first order conditions are different. The leader is the only firm that changes its conjectures. The first order conditions that result from these effects are:  61  The Stackleberg leader knows that its own actions affect how the follower acts. The derivative  öv ‘/ôv is no longer equal to zero. Using the technique of implicit differentiation on equation 3.33 we can sign this derivative. The partial derivative of 3.33 with respect to  334  ,’ 1 f öv 2 ö 2 ãv’  If we rearrange this, isolating  c3i’,,  öm öv  aV ‘/aV  ôV,ôV  9v  aV  V  is:  ôm  ôVöV’  2 ôV  ôv  öV  ,we generate the following:  -1  f  335 öV  ôv  2 ) av’ 2 av  ömôm ÔV  2 cJv  From our assumptions about the supply curve, we know that the first and second derivative of the supply function are positive. This makes the numerator a positive quantity. The second derivative of the revenue function is negative by assumption, which is demanded by the downward slope of the marginal value product. The resulting negative denominator makes the right hand side of 3.35 negative. The optimization condition for the leader is:  3.36  ôv am / 1+— 1÷— v “ —=mlv +vC 1 ôv,  With the condition outlined above, we see that  ôv,, ‘/ôv will be less than zero. This means that the  term on the far right will obey  3.37  t3m v —v  —  av  aV,,  The optimization condition for the leader will require a smaller marginal value product than with Cournot conjectures. Along with the decreasing marginal value product assumption, the Stackleberg leader will  62  purchase more input than a Cournot oligopsonist with the same technology. The leader is capturing a larger share of the profits by anticipating its rivals reaction. In markets with a cooperative, we often see the cooperative accounting for the largest single share of the production received. This is the case in the BC dairy industry, and is similar elsewhere. Superficially the cooperative appears to be the logical choice as the leader firm. This conclusion is a little premature. The cooperative is bound by the actions of its members. It cannot autonomously determine its level of output. It is more logical to conclude that the IOF will be the leader, in spite of the fact that we see the IOF capturing a smaller market share. We can develop the Stackleberg model with the cooperative as one of the firms. The objective function for the IOF is again  max{f(vp)_m(vp +v)v}  3.38  while the equilibrium condition between the cooperative and its members is  f(vj/v =m(v +v).  3.39  The first order condition for the IOF can be defined as above, and is  0v  am / —=miv \p +v C)1+— p p  3.40  Vp p  Now we must again determine the sign of the conjectural variation, in this case what will be the production response of the cooperative to the decision of the oligopsonist. We again proceed by implicit differentiation to get the sign of  .  Completely differentiating 3.37 with respect to V we get:  iafaVif(jf3vam 3.41 v öv öv,  V  VC  ÔVp  63  öV  We can rearrange this and isolate  öv/öv  to get:  avam  of  Ov  öv,  3.42  f(v)o,n C  v  _1v  Ov  As above, we assume that the supply curve is upward sloping, which translates into a positive numerator on the right. The denominator is equal to the difference between the marginal value product of the cooperative and the average value product, less a scaled slope term. Since we are assuming that at the production point for the cooperative is in the range where the AVP is above the MVP, the denominator is going to be negative. This result indicates that the IOF, when in competition with a cooperative, can respond strategically in anticipation of the behavior of the cooperative. The profit maximizing point for the IOF is at an input level above that of the equivalent oligopsonist with Cournot conjectures. However, it is still less than the point where the intersection of the IOF’s marginal value product and the price line occurs. The IOF anticipates that if it increases the price it offers, relative to the Cournot situation, it will attract farmers away from the cooperative. The higher price will attract new farmers to the industry, so that the overall price increase in the input market is not as large as the firm with Cóurnot conjectures believes. The input supply effect felt by the IOF is modified by the interaction between the cooperative and its members.  3.9) Summary In this chapter we defined the objective functions and technology for our model. We have a single commodity, milk, being traded between the farmers who produce it an the processors who purchase it and sell it into a further market. With no market regulatory constraints, and ignoring the potential strategic interactions, we reproduce the standard ‘competitive yardstick’ outcome. If we have entry barriers or downward sloping marginal and average product curves, then an IOF can act strategically against a cooperative. It uses the fact that the cooperative dividend must equal the aggregate supply price to choose its output level, such  64  that its input cost is reduced. Oligopoly profits are not earned, but the competitive yardstick is not complete.  65  Chapter 4: A Quantity Restriction In the last chapter we introduced the problem and solved it for the unregulated case. We found that the intersection of the members’ incentives with the cooperative structure generate the competitive yardstick effect. We also found that with the right technological constraints we can soften the competitive yardstick a little. In this chapter we continue our analysis by introducing our first restriction, an upper bound on the aggregate production of milk. This restriction simulates the effect of the supply management quota on the dairy sector. We first investigate the effect of the restriction on the processor-producer and cooperativemember relationship in isolation. We then locate an equilibrium between two IOFs and an IOF and cooperative in competition.  4.1) The for profit processor We now add the first restriction, constraining the amount of raw milk available. Under supply management, the total amount of the raw product available to all processors is established by an external authority, the milk board. From the perspective of an individual processor, the production that other processors are carrying out is beyond their control. The processor’s production must be less than or equal to the difference between the total amount of production allowed and the production of all other processors. The problem is defined as: 4.1  rnax{pf(v) mv —  4.2  maximv”  4.3  =  ,  —  c (yr  }  IOF objective function.  )}  Independent farmer’s objective function.  Market clearing condition.  +  In this definition, the IOF is maximizing the profits over the choice of input volume  Vi,.  The  independent farmer is maximizing the difference between the market revenues and the production costs over the choice of  vf. The market clearing condition requires that the sum of the production of all the  66  producers is equal to the amount of milk used by all the processors. As written,  v_ is the amount of  input used by all the other processors. However, we now have a restriction attached to the market clearing condition which must be obeyed. The sum of the production of all farmers must equal the milk used by all the processors, and this amount must be less than the upper limit given by V. Since we have a restriction to contend with, we solve this problem using lagrangians. The lagrangian functions for this problem are: 4.4  The IOF processor.  +[(v_v)_v] 4.5  The independent farmer  L(vr,)= mvr  —  c(v”)÷  M[(v  vf]  —  We will first look at the processor’s problem. We are maximizing the amount of input chosen and the shadow value of the constraint. To be completely general, we would allow the possibility that the restrictions would not be binding. To allow this, we search for a solution by way of the Kuhn-Tucker conditions. The Kuhn-Tucker conditions we derive from this optimization are:  4.6.1  aL(v, )/ov  =  p[af (v)/ov]_ m  0, —  .,  0, oL(V,.)/oA.  4.6.2  =  (v  [aL(v ,  =  ]i  0  0,  v,,  —  )/av  o,  =0  These Kuhn-Tucker conditions allow for corner solutions and binding constraints. Normally one tests all the possible combinations of binding and free constraints until the best solution is found. In this particular case there would be four possible combinations,  {(v  =  0,  =  o),(v 0, ,  =  o),(v  =  o),(v 0,  0,  67  o)}.  We can make a couple of simplifying assumptions to eliminate all but one of these choices. Our first assumption is that the constraint on the amount of input used is binding. All the firms in the industry collectively use all the raw product available. With this constraint, the shadow value of another unit of raw product, were it available, might be positive. This reduces our set of conditions to  {(v  =  o,, o),(v  o,.., o)}.  The next condition we impose is that the finn we are  investigating is using input. This constraint eliminates the situation where the input level is restricted to be zero. With these restrictions we can generate two equalities that reveal the details of the solution. These are:  4.7.1  4af(v)/av]= m+  4.7.2  v=V—v  Equations 4.7.1 and 4.7.2 characterize the solution. In 4.7.1 we see that the marginal value product is equal to the sum of the marginal cost of the input, the price of the raw product and a shadow value. 4.7.2 shows that the sum of all the raw product used is equal to the total available, reproducing the constraint. Figure 4.1 shows the solution graphically. Without restrictions, the firm would purchase Q 0 units of input.  $  ::::::::::::  0 1 Q  /  Shadow Value  milk  Figure 4.1: Marginal value product, free market price and shadow value.  68  At this point, profits are maximized since the value to the firm of the next unit of input used is less than the free market price m. This firm is unable to reach Q , for example, when the other firms have secured 0 enough of the input so that this firm can only get Q . Under this constraint, the last unit of input is worth 1 m’ to the firm. The difference between m and m’ is the shadow value for the last unit of input, extra profit this firm would earn if it could get another unit of milk. The price m’ is the maximum that the firm is willing to pay to get that extra unit.  4.2) The independent farmer. The farmer’s objective function is unchanged, except that all farmers together face an upward bound on the amount of milk that they can produce. The farmer can produce no more than the difference between the maximum allowable production, V, and the amount that is being produced by other producers in the industry. We assume that the farmer is producing a positive amount of output, and that at the margin it is not unprofitable to produce the last unit of output, eliminating the Kuhn-Tucker problem. In effect we have made the output restriction binding. No farmer would choose to produce less than the amount they are entitled to produce. SuppJy management changes the nature of the production decision. Changes in the level of production involve a cost, in whichever direction one may deviate. To escape these costs, one buys or sells quota. However, this is an investment decision, not a production decision, and is not an explicit part of this model. For simplicity, we assume that the farmer has chosen the profit maximizing quota amount, and all farmers are identical to the point that if the constraint is binding on one farmer, it is binding on all. In this way we have changed what is accurately an inequality constraint to an equality constraint for this model. Why are we evaluating a single variable optimization problem with an equality constraint? The solution identifies the cost, at the margin, of this production constraint. This cost is the source of the farmer’s willingness to pay for quota. In chapter nine we test our model using the price of quota. The theoretical foundation for this test lies with the assumption that farmers will capitalize economic rents into the quota  69  $  Shadow  m  v’ v 1  milk  Figure 4.2: Fanner’s marginal cost with and without supply management.  price. Economic rents are identified by a positive shadow value, which this analysis allows us to demonstrate. The first order condition comes out as:  4.8  t3L(vr, . )/öv”  =  m  —  oc(vf )/av[ +  Equating this to zero we have:  4.9  m— oc(vr)/ovr  —  =  o  And rearranging we get:  4.10  m= ac(vr)/avr  +p  We see the fanner again equates marginal revenue, the price they receive, with marginal cost. However, the restriction adds a shadow value to the marginal cost, reflecting the extra profits the farmer could earn if another unit could be produced.  70  ’ represents the share of the total output that this farmer produces. The institutional 1 In figure 4.2. v shorting of the market pushes the price up to m’. The shadow value is the distance between the MC and m’ at the limit of production. The supply restriction is equivalent to an upward shift of the supply curve. The forgone revenues on the next unit of production are represented by the shadow value, the amount by which the supply shifts up. For a fixed milk price, the individual farmer, is now supplying only v’ units of milk to the market, instead of the desired production of vg.  4.3) The supply controlled market with independent agents. We now explore the interaction of the IOF processor and the independent farmers. The first order conditions that we generated above are: 4.7  The IOF processor. p[af(v)/ov]= m÷  4.10  The independent farmer m=ac(vf)/ovf  ÷‘  These conditions indicate that at the equilibrium outcome, at least one of the players will face a shadow cost at the margin. Who gets the rents is unclear, but the market equilibrium condition that there be only one price must be satisfied. The lack of any formal mechanism for distributing these rents in the model means that the main way these rents are distributed is through bargaining. Graphically, figure 4.3 shows what is going on. The supply restriction holds the production of the industry at the level V, which is below the free market solution. At this production level, there are rents available to the total industry, represented by  ?. + 4. How these rents are distributed between the  processors and producers depends on the negotiation power of the agents. As drawn, the restriction has allowed farmers to receive enough of the rents to push their effective price above the free market solution. However, this need not be the case, and since processors are usually more concentrated than producers, one would expect that under a system like this, they probably have the majority of the bargaining power.  71  $ Supply  V  milk  Figure 4.3: Aggregate processor MPV and aggregate producer MC.  4.4) The Cooperative Processor. The cooperative processor must also operate under the restrictions that govern the industry. We add the volume restriction to the cooperative’s problem. The objectives will now be:  }  Cooperative objective function.  4.11  rnax{pf(v )/v  4.12  max {[pf(v )/v }c  4.13  =  +  —  C(yc  Member’s objective function.  )}  v  Market clearing condition.  v_ represents the amount of input used by all the other processors. As before, the flexibility for the cooperative is limited, but knowing where the objective is maximized allows us to see how the incentives of the members change this. We again evaluate this problem using lagrangians:  4.14  The cooperative processor.  L(v, 4.15  )  =  pf(v )/v  [(v  +  —  v) v] —  The cooperative member  L(v(,)  =  pf(v )v/v  —  ..v;) v:]  c(v:) +  —  —  72  If we assume that the restrictions are binding, then the first derivatives are: 4.16  The cooperative processor.  i3L(v, ) )/ôv 4.17  =  p[of(v )/av J/v  —  pf(v )/v2  —  The cooperative member  aL(vr,r)/avr + [am(vr)/av ]vr 1  —  ac(vr)/avr  —  We have evaluated this derivative without replacing the member’s price with the patronage dividend. The aggregate supply restriction generates a shadow value at both the processor level and the producer level. We can see that the cooperative structure transfers the effect of the restriction to the members. We locate the cooperative’s solution by setting equation 4.16 equal to zero. After rearranging we find:  4.18  p{af(v)/vJ/v  =  2 pf(v)/v  +  Multiplying through by the volume of milk received as above, we get the following:  4.19  4of(vj/av]= pf(v)/v  +  Like the for profit processor, there is a shadow value resulting from the industry wide constraint on the production level. This shadow value takes on a somewhat different form than it does for the IOF solution. It appears that the size of this shadow value varies with the amount of milk that the cooperative receives from its members. The structure of this shadow value is an artifact of the first order condition. We are looking for the maximum of an average value product. The optimization is per unit of input used rather than for total profit. At the margin, the shadow value is equal to the sum of the shadow value over every unit, multiplied by the totul number of units being taken.  73  $  +  m  1 Q  milk  Figure 4.4: Cooperative MVP, AVP and adjusted AVP.  In figure 4.4, the scaled shadow value is given by the increasing linear function  This function is  an artifact of the optimization, which is evaluated on a per unit basis. When we adjust this to get the MVP and AVP for the entire cooperative, we get a linear shadow value. We add the shadow value to the AVP to generate an adjusted AVP curve. To find an optimum, we locate the intersection of the MVP curve with the adjusted AVP curve. To maximize the patronage dividend that the cooperative can pay its members, it should operate at Q , left of the unregulated market solution. 1  4.5) The Cooperative Member Farmer The cooperative member faces the same optimization as the independent farmer, except that the price the farmer receives is tied to the return generated by the cooperative. Since the return generated by the cooperative is a function of the amount of milk that is processed, subject to the restriction imposed by the regulations, there is again an endogenizing of the price received that is not the case for the independent fanner. Rearranging equation 4-17 into the standard form, we get:  4.23  m + [om(vr )/avr ]vI  =  oc(vr )/ovf +  74  T  =  Into this we need to substitute the marginal impact that the fanner would have on the cooperative, if the fanner could ship another unit of milk. We take this from the lagrangian that defines the cooperative’s behavior, equation 4.14 above. At the point where the average value product of the cooperative is maximized, this number is zero. However, where this occurs the fanner’s own marginal cost might be below the average value product, which generates an incentive for the farmer to ship more milk to the cooperative. With the assumption that the cooperative must take and process all of the production of its members, this moves the cooperative away from its maximum average value product. The behavioral equation will look like:  2 m+ {p[of(v)/ovJ/v —pf(v)/v  4.24  —  =  oc(vr)/avr  +iq’  =  o  Multiplying out, rearranging, and replacing m with the average value product, we get:  4.25  (i  —  vf /v, )pf (v )/v  +  (v[ /v, )p{af (v )/av]  ac(vf )/ivf  +  r  +  =  0  Here we see that the marginal revenue the fanner faces is again a combination of the marginal value product and the average value product. The marginal cost is now adjusted by two different shadow values. The shadow value represented by represented by the term  is the farmers individual shadow value. The shadow value  ?.vf is the farmers individual share of the cooperatives shadow cost of  This shows that we can interpret the shadow value which is experienced by the cooperative as a rotation of the member’s marginal revenue curve, or a rotation of the member supply curve faced by the cooperative.  4.6) The Supply Controlled Market with a Cooperative and a For Profit Firm. Having looked at the individual agents, we now turn to the effect of these controls on the interaction between the cooperative and the for profit finn. As in chapter three, we assume that there are two milk processors with identical revenue functions. The revenue function is continuously differentiable and assumes the standard three stage shape. We also assume that the restriction on the total quantity of raw  75  milk on the market is binding. This assumption decouples the IOF’s input cost from the supply curve. The two oligopsonists are competing against each other to secure a share of the fixed size input market. The general form of the supply relationship on the farmer’s side is taken to be unaffected. They take the price offered as given. However, due to the restriction, there may be a shadow value related to the next unit of production. The structure of this problem is: 4.26  rnax{pf(V) m(v)v —  4.27  } }  First oligopsonist’s objective function.  rnax{pf(v’) m(v)v  Second oligopsonist’s objective function.  4.28  m = m(v);  Input price condition.  4.29  v,,  —  +  v,,’  v  =  v,  +  Market clearing condition.  V  As we have been doing all along, we will construct the lagrangians to represent the problem of the oligposonists. The problem is symmetric, so we only need to construct one lagrangian.  4.30  L(v,?)=  f (v)_mv +[(v—v’)—v]  Alter setting the first derivative to zero, the first order condition is:  4.31  af(v)/av =m+  This result is very similar to the competitive result, which it resembles. The firm is no longer able to use the supply curve to optimize its profit. We must investigate the stability of the solution 4.31 represents. The most obvious potential equilibrium point is at the equality of the marginal value products of both firms. Identical firms are expected to share the market equally. The equilibrium condition can be written as:  4.32  af  )/öv,,  =  ôf(v ‘)/av’  76  $ of(v )/o v , 1 S. S. S.  af(V- v)/o(v  Supply  -  V  v,  milk  Figure 4.6: Oligopsonist’s marginal value products.  We can use the fact that the market size is fixed to write everything in terms of v:  4.33  of(v)/ov  =  of(v v)/o(v —  —  v)  Relying on this makes it easy to continue with a graphical analysis. We reflect the marginal value product of the rival finn, moving its origin to the total market size at V. This is shown in figure 4.6. The symmetry inherent in the definition being used means that each firm will use exactly half of the available input, under this equilibrium assumption. If the finn holds Cournot conjectures, it believes that its rival will keep production fixed. The firm cannot increase its production, since the total input supply is fixed. However, it does not need to keep its own price fixed. Given that the rival finn holds its production fixed, this firm should reduce the price it offers, down to the point where the aggregate farmer’s supply curve intersects the limit on the aggregate volume. If both firms hold strictly to the Cournot conjectures, we would expect the market price to fall to the minimum price where all of the supply would be brought to market. However, if we propose that the atomistic suppliers to this market are indifferent with respect to who they supply their product to, then the situation becomes a little different. If both firms are offering the minimum price, mm, then one firm could  77  $  V  milk  Figure 4.7: Cooperative’s marginal value products.  achieve a larger share of the market by raising its price a little. The firm will then be able to process at its optimal point, the intersection of its new price offer with its marginal value product. The rival firm has exactly the same options available to it, and it will therefore raise its own price, and the price will ratchet up to a maximum at m. With these modified Cournot conjectures, the milk price will be unstable and lie somewhere between m, the intersection of the two firms marginal value product curves, and mm, the minimum price to have V brought to market. If we introduce a cooperative, we modify the equilibrium condition slightly. The cooperative must obey the condition that its average value product is equal to the market price, which follows from the assumption that the members behave opportunistically. This can be written as:  4.34  f(v)/v =m  At the equilibrium point between the cooperative and the oligopsonist, both firms face the same market price. Relying on the fact that the IOF’s production is equal to V-va, the equilibrium condition is:  4.35  f(v)/v  af(v vj/a(v —  78  —  v)  We can explore this relation graphically as we did for the two oligopsonists. The equilibrium condition is satisfied where the cooperative’s average value product intersects the marginal value product of the oligopsonist. This intersection occurs at a higher price, and a lower production level for the oligopsonist, than for the two oligopsonist equilibrium. If the oligopsonist holds Cournot conjectures, it will again be the profit maximizing decision for the oligopsonist to reduce the price it offers. However, assuming indifference on the part of the fanner suppliers, this will result in more members joining the cooperative. To respond to this, the oligopsonist will offer a higher price, with the highest price it will pay being that price where its marginal value product intersects the average value product for the cooperative. This becomes the only stable outcome, since the cooperative does not price discriminate against its members. We have restored a result that is much like the competitive yardstick. We can also build a Stackleberg model. The input purchase identity is still V  -  =  V.  Assuming that  the follower chooses to operate where its marginal value product is equal to the price, we get the following profit function for the leader:  4.36  =  f(v)—[öf(V—v)/ö(V--v)Jv  The leader firm knows that the follower will choose to operate at the point where the marginal value product is equal to the marginal input cost. It can maximize its own profit against this fact. The first derivative of this relation is:  4.37 ô/öv =Of (v)/öv  If we set this condition equal to zero and rearrange it, we get:  4.38  af(v)/ov =af(V_v)/a(V_vp)_[o2f(V_vp)/a(V_v)2]vp  The optimal point of operation for the leader incorporates the second order effect on the leader’s profits of a change in the followers point of production. To interpret this we compare it to the equally split market  79  explored above. At the equal split point the equality does not hold. The slope of the MVP is negative here, so the far right term is greater than zero. The entire right hand term is therefore less than the left hand side. The optimal point for the Stackleberg leader is to choose a production level that is less than half of the market. It maximizes its profits by allowing the follower to take a larger share of the market, generating a lower price for the input. We can conduct the same analysis for the combination of the cooperative and the oligopsonist. Here the factor that the oligopsonist internalizes is the price identity of 4.34. The cooperative’s input choice, and the resulting market price, are determined by the atomistic actions of the members. The profit function for the oligopsonist is now:  4.39  =  f(v)—[f(V—v)/(V—v)]v  The rival IOF controls the market price, through the control of its own production level. This control is effected by offering a price that gets it the amount of input it needs. It can use its offer price to control v, and thereby move the market price around to maximize its own profit. When we take the first  derivative of this function with respect to the oligopsonists production level, we get:  4.40  aof(v)f(V_v)  öf(V—v)f(V—v)  av  o(v_v)  V—vs  v—vs  v,  v—vs  If we set the above equal to zero and rearrange terms, we get:  4.41  af(v)f(V-v) ovp  V—v,  of(V_v)f(V_v) a(V—V)  V—Vs  V  , 1 V—v  We can inspect this equation the same way that the for profit oligopsony was looked at. We need to investigate the sign of the far right term. This term is equal to the cooperative’s marginal value product less the average value product, scaled by the share of the total production conducted by the IOF. This term is going to be negative. The IOF’s profit maximizing point is below that when the IOF holds  80  Cournot conjectures. The IOF maximizes its profit by allowing the cooperative to take a larger share of the market, and allowing the price to fall as a result. If we think in terms of the residual supply available to the IOF, we see that it is downward sloping. The IOF is essentially pricing its input against the AVP of the cooperative, rather than the MC of the farmers.  4.7) The Quantity and Price Restricted Single Input Optimization Problem Adding a minimum price to the problem does not change the situation presented above. If the minimum price is set above the maximum price which the oligopsony models generate, then we will restore a competitive situation between the competing firms. On the farm side, the high minimum price will mean that there are farmers who are willing to produce milk, milk which fits in the upper limit imposed by the government, which cannot be sold. Effectively the high milk price has shifted the rents resulting from the output restriction entirely to the farmers, and in fact ‘cheated’ some farmers out of some of these rents. If the minimum price is below the price required to get farmers to produce the legislated volume, it will have no effect at all. The competing oligopsonists do not offer a price below the price which will result in the total volume being produced. As such, the minimum price will be ignored altogether. If the minimum price falls between these ranges, it will act as a floor for the oligopsonist’s price offering. As such, it will reduce some of the variability in the distribution of rents. Altogether, imposition of a minimum price condition adds little to the analysis. The principle impact of a minimum price is fixing how part of the rents will be distributed. If the minimum price is set above the supply price the farmer’s will accept, then the minimum price guarantees that the IOF cannot generate as much profit from the market as it might.  81  4.8) Summary If we add a volume restriction to the problem presented in chapter three, we force a gap between the processors’ MVP and the farmers’ MC. This gap translates into rents. If the input market is occupied by two IOF firms, we find that the input price becomes unstable. We have reproduced the Bertrand result from Cournot conjectures. The volume restriction means that the IOF can act strategically when facing a cooperative. The lower bound for the input price is given by the cooperative’s AVP. It is against this downward sloping curve, rather than the upward sloping supply curve, that the IOF optimizes. The farmer price will be higher, but the IOF can still secure some rents.  82  Chapter 5: The Pooled Milk Supply We have now solved our model without regulations, and with an aggregate volume restriction. In this chapter we explore the implications of a price pooi for all the input produced in the market. A price pool is a policy instrument that guarantees all producers of the input receive the same price for their production. It is as if all the milk produced by the farmers is thrown into a giant vat, and all the processors purchase their input form this vat. The proceeds of these sale to the processors is collected, and distributed to the farmers according to how much they delivered to the pool. If a cooperative is operating in isolation, or if there are no cooperatives present, the introduction of a pooling scheme adds little to our analysis. Since we already assume all the producers are identical, it is of no advantage to the processors to be able to differentiate between them. Further we are supposing that all the processors are the same and that one price must prevail in the market, so in effect we are acting as if we have a pool in place. However, the introduction of a milk pool when there is a cooperative in competition with an IOF is a little more complicated. In this chapter we explore the interaction between a cooperative and an IOF if both are securing their input from a common pool.  5.1) A Cooperative and an IOF In Competition with Naive Conjectures. As in the previous chapters, this situation will be modeled with a large number of farmers producing milk for the raw milk input market. There are two processors with identical technologies. One is investor owned, while the other is a cooperative. All milk produced is made available to both processors at the market price, irrespective of whether or not the farmer is a member of the cooperative. There will be no volume restriction and no price floor. The milk pool guarantees all farmer’s who ship milk the same price for their production, and allows all processors to purchase milk at the same price. The basic problem is set up below. We take the first derivatives of these functions, set them equal to zero, and use the equilibrium conditions to solve.  83  5.1  p  p(y);  y  =  =  5.2  m  =  m(v)  v  5.3  w  =  W(V)  V  A  max Iw(v)v” vf  5.5  1.  —  Supply function for raw milk.  +  c(vr  )}  Independent farmer’s objective fUnction.  i F p(y)f( ) v max .J  5.7  [  [  ÷ w(V)v:  max {p(y )f (vi,)  —  —  C  c(V:)  w(v)v  [p(y)f(v)— w(v)v max V  5.8  +  “ =  Cooperative member’s objective function.  1  —  5.6  Pool price function  V + V  1  V  Output market demand.  f(v)+ f(v)  j [ j V  }  IOF objective function.  }  Cooperative’s objective function.  Market clearing condition  + V,  We have introduced a new function,  w(v), to the problem definition.  This function represents the price  that the IOF and cooperative must pay to get milk from the pool, and the price that independent farmers and cooperative members get from the pool for the milk they deliver. This fact has changed the objective function of the cooperative and of the cooperative member in particular. In 5.7 we see that the cooperative now has an objective function that is almost identical to 5.6, the objective of the IOF. The principle difference between the two lies in the denominator in 5.7, the shipments to the cooperative from its members. In 5.5 we see that the objective of the cooperative member is equal to the objective of the independent farmer, shown in 5.4, with the addition of the patronage dividend term. Belonging to the cooperative entitles the farmer to a share of any dividend that it is able to earn. Both farmers receive the same basic price for the milk that they deliver to the pool. The first case we explore involves what we will call naive conjectures. The naive conjectures we postulate follow the competitive assumptions. We assume that the individual agents believe that their own effect on the market is zero. Mathematically, these conjectures can be written as:  84  , 0 a(’)=  0 aw(v)_  and  öv  O,  —=  öv”  [1,  j, k c, p, 1 c, p i=j, k=c,p, l=c,p i  =  =  All the actors believe that their own actions will not have any effect on the input or output prices. The cooperative processor feels that its own input use decision will not affect the input use decision of the IOF processor. Similarly the IOF processor believes that its actions will not affect the input choice of the cooperative. Finally, the individual farmer believes that any changes in this farm’s production level will not affect the production decision of other farmers in the milk industry. These simplifying assumptions are typically made for a basic analysis, and we have used them in previous chapters without identifying them. Alone they add little to the analysis, but we have not yet explored all the conjectures that exist. With a milk pool in place, the cooperative is free to choose its input level independently of the decisions of its members. The only interaction that exists between the cooperative and the membership is by way of conjectures about the responses. The conjectures that we are going to use for this stage of the analysis are:  —-=0 and  L=0 av  The cooperative believes that its own input use decision will not affect the production decisions of its members, and the individual members believe that their decision will not affect the input use of the cooperative. These conjectural assumptions are critical, and identify the principle effect of the milk pool. The cooperative is decoupled from the incentives of the members who ship milk to it. With some rearranging, the first order conditions for the model under naive assumptions are:  85  5.9  The independent producer  w(v)=  ac(vr) ovr  5.10  The cooperative member farmer  p(y)f(v  )  —  v: 5.11  w(v)v ) v.c  p(y)f(v  w(v)v —  —  v:  ae(v:)  —  ÷  ‘%  ‘  The for profit processor , , ..of(v,,) =WIV PI%Y) ôv,  5.12  The cooperative processor  of(v) p’y)  i3v  1  c=  1  cY  The first result is that for the independent producer. The producer chooses to operate at a point where the price that is received for the milk that is delivered to the pooi is set equal to the marginal cost of producing the milk. From the perspective of the independent producer, the fact that the milk is delivered to an anonymous pool is no different from if the milk was sold to a specific processor. Now we turn to the problem of the cooperative member. The cooperative member delivers the farm’s milk production to an anonymous milk pool, but receives a patronage dividend from the cooperative according to how much milk is delivered. Unlike the previous case, the cooperative is not bound to process the milk which its member’s produce.  p(y)f(v)— w(v)v  —  p(y)f(v)—  =  W(V)V  +  w(v) av:  5.13  Patronage Dividend  -  Dilution Effect  +  Pnce =  Marginal Cost  The marginal revenue which the cooperative member faces includes three factors which represent the effect of the farmer’s production on the revenue generated. The patronage dividend is the amount which the farmer receives, per unit of milk produced, of the surplus on operations generated by the cooperative. The dilution effect is the amount by which the member’s production changes the size of the patronage  86  dividend. These two terms are added to the price which is prevalent for the milk that is delivered to the pooi to determine the marginal effect on the member’s revenue of a change in production. The investor owned firm’s optimal input use level is identified by the standard simple relationship. This relationship is the equating of the marginal value product with the marginal input cost, here represented by the price that must be paid for the milk purchased from the pool. This result follows logically since the processor must pay the going milk price when making the input decision, and profits are maximized when the processor chooses an input level where the marginal value product of the last unit of production is equal to the marginal cost of the last unit of input. The optimal point for the cooperative processor is similarly straight forward. If we eliminate unnecessary terms, the optimal input level is determined by the relation:  5.14  af(v) 0  =  w(v)  c3v This is identical to the relation that identifies the IOF’s optimal input choice. When the cooperative and the IOF are purchasing their input from a common anonymous pool, and paying the same price, the fact that the cooperative is distributing its earnings to its members on the basis of how much milk they produce while the IOF is generating a return on investment is irrelevant. Given these naive assumptions about the conjectures of the agents, and accepting that the objective of the cooperative is to maximize the patronage dividend, the behavior of the cooperative and the IOF should be indistinguishable. Figure 5.1 shows the interaction between the cooperative and the IOF in the input market. For the moment we ignore the effect of the cooperative’s patronage dividend on the pool price. The two firms must share the market, so the cost curves of both finns have been drawn side by side. The curves for the IOF begin at the point where the cooperative chooses to procure its inputs, and fills the gap to the intersection of the supply curve and the price line. Both firms pay the same price, w, to the pool for the milk that they use. The two processors are splitting the market evenly between them, which follows from their use of identical technology.  87  $  IOF’s MVP  Co-op’s MVP  ients w IOF’s AVP  milk Figure 5.1: The rents of the cooperative when it has been decoupled from its membership.  The behavioral difference that occurs in this market is a result of the actions of the farmers. The IOF and the cooperative behave in an analogous way. However, the patronage dividend generated by the cooperative is like a price bonus, which affects the profitability of the cooperative member and the independent fanner differently. Under these conjectures, and given the assumption that the fanner is an opportunistic profit maximize, it is optimal for all fanners to belong to the cooperative, so long as the cooperative is generating a positive profit. The cooperative and the independent fanner face the same price for the milk that they contribute to the pool. However, the cooperative member receives an extra payment on top of this. Given that profit maximization is the only objective of the farmer, this extra payment will mean that all farmers will want to be members of the cooperative. The distribution of the patronage dividend to the member farmers, while the cooperative does not itself have to handle all of the product that its members produce, has a supply effect. At the equilibrium the farmers must be operating at a zero profit point. Assuming all farmers are members of the cooperative, and farmer produces where marginal cost equals marginal revenue—the sum of the patronage dividend and the pool price—the pool price will be less than the supply price. The rents generated by the cooperative are distributed to its members. The cooperative distributes these rents to the farmers in the form of a patronage dividend. The farmer sees an effective price that is above the pool price. As a result,  *8  $  Co-op’s MVP  IOF’s MVP  \  /  m w  ;;;;;;N%%__ / /  ‘  1 +x+x+x+x.r.ww”.—  ++:y  /  S... ..%.S.—’...—._—  I  7  ‘\  Patronage  v  milk  Figure 5.2: The supply effect of the patronage dividend.  the fanner is willing to accept a pool price which is below the supply price. The IOF buys milk at this pooi price, not at the higher supply price. In competition with a cooperative, the IOF faces a lower input cost than if it was in competition with another IOF. Investigating this result in detail, we find that all the rents captured by the cooperative are transferred to the cooperative. We assume that farmers are atomistic and opportunistic. No farmer is large enough to affect the actions of any other agent in the model, and all act to maximize their individual welfare. With a large number of farmers, their own interactions dictate that the effective price they receive lies on their aggregate supply curve. Entry, exit, and adjustment of production level must by assumption force the farmers’ economic profits to zero. This fact allows us to write a pool price identity:  5.15  {P(Y)f(t’c)_ w(v)v  }  +  w(v) = m(v)  This identity follows directly from the assumptions stated above. The sum of the pool price and the patronage dividend, the effective price the farmer sees, must be equal to the supply price. Farmers cannot earn economic rents through their membership in the cooperative.  89  We can rearrange this relationship to isolate the pool price function. The identity which defines the pool price is:  vm(v) p(y)f(v) —  w(v)  5.16  V  —  The price farmers are willing to receive from the pool is equal to the product of the supply price and the total milk produced—total farm revenues—less the cooperative’s earnings, divided by the share of the industry output not handled by the cooperative. We can use the identity in 5.16 to evaluate the profits of the cooperative. If we insert 5.16 into 5.6, and do a little rearranging, we find:  p(y)f(v  —  5.17  If we recognize that V  = ,  —  V,,  =  p(y)f(v,,)  V and  )}  )  f(v + f(v) = y then we can immediately rearrange 5.17 to  be:  5.18  =  yp(y)— vm(v)  All the profits in the industry accrue to the IOF which is competing with the cooperative. This result is not unexpected. The cooperative redistributes any rents it captures to its members. The members compete to be part of the industry and receive these rents. They succeed by offering to accept a lower price for the milk they ship to the pooi. This process continues until the farmers are again earning zero profits, while the IOF captures all the benefit of the reduced pool price.  90  5.2) A Cooperative and an IOF In Competition with Cournot Conjectures. The second set of conjectures we imagine have the Cournot form. Under Cournot assumptions, the agents assume that their rivals, or comrades, hold their production fixed. We can summarize these conjectures as:  0 öm(v)  , 0 a&) #3y  öv  —=0 and  We assume that the supply  t3v  ôv  i  f, k c,p, 1 c,p i=j, k=c,p, l=c,p  0,  —=  [1,  öv  =  curve for the input market is upward sloping from the perspective of the  processor. We also assume that the demand curve for the final processed product is downward sloping. All the agents still believe view that their actions will not affect the actions of their rivals. If we apply these conjectures to the derivatives derived above, we get the following first order conditions. 5.19  The independent producer ac(vr)  w(v)= 5.20  övf  The cooperative member farmer  p(y)f(vj— w(v)v  v’  v: 5.21  (  p(y)f(vj— w(v)v  ãc(v) +v)=  v)  The for profit processor aP(Y)af(vP)f() 3y  5.22  —  “i  öv  +p(y)  ôf(v)  öw(v)  The cooperative processor 1  2  IöP(Y)öf(VC) v ay Ov  c  ôm(v)1  1  )‘1=Jm(v)  av  J [  +  2  For the farmer, ptimization with Cournot conjectures does not change the result we found above. Independent farmer, and the cooperative member, are both small producers with respect to the total  91  market size. They maintain their belief that they have no impact individually on the market price. As above, the optimal decision for all producers is to belong to the cooperative. The processors results are also similar to the finding above. When we simplify the first order conditions for the cooperative, we reproduce the same optimization condition. The IOF and the cooperative will still behave identically if the IOF is maximizing its profit, and the cooperative its patronage dividend. The equation presented in 5.21 and 5.22 is the standard first order condition for a Cournot oligopsonist operating between a supply curve and a demand curve. On the left we have the marginal revenue. The oligopsonist believes that its rival will keep its production fixed. As a result, this firm faces the slope of the demand curve as its price effect on the revenue side. On the right we have the marginal cost. Again the firm assumes its rival’s production level is fixed. It therefore faces a similar price effect from the input supply slope as from the output demand slope. These price effects scale the marginal value produce and marginal input cost that we equated under the naive conjectures. The next case is the Stackleberg leader-follower model. With the conjectures that identify this model, we propose that one of the agents internalizes the optimization problem of the other, who has Cournot conjectures. Normally one considers firms that are characterized by the same objective, so that it is irrelevant which agent is assumed to be the leader. Before this point the cooperative has been bound by the actions and could not assume the leader’s role. However, we have broke this tie, so that there is no longer any reason to assume that the IOF will be the leader. We will avoid the mathematics and explore this relationship intuitively. The basic result in the traditional Stackleberg model is that the leader produces more than the follower. The total production is greater than in the Cournot case, so that the market price is lower. However, the leader gels a sufficiently large share of this to generate greater profits than if the Cournot solution. The follower is left to ‘take up the slack.’ It has a lower production level, and usually much lower profits, than in the Cournot case. However, since it makes its production choice second in this sequential game, it cannot do any better for itself.  92  In isolation from the actions of the members, the Stackleberg case is identically repeated here. However, we cannot really look at it in isolation of the fanner. If the cooperative takes the larger share of the input market, it generates a greater patronage dividend. However, as we pointed out earlier, the farmers will adjust so that the profit is still zero even with this patronage dividend. This allows a lower price to prevail for the pooled milk, actually lowering the cost for the IOF. If we reverse the situation, giving the IOF the largest share of the market, the cooperative will generate a smaller patronage dividend, and the zero profit outcome at the farm level will require a higher price. With a larger market share, the IOF actually increases its input costs relative to the Cournot case and when it has a smaller market share. The total effect on profitability is of course dependent on the revenue effects, but it is clear that the IOF is not as well off capturing a large share of the market as when it has a smaller share.  5.3) A Cooperative and an IOF In Competition with Unrestricted Conjectures. With this model, the remaining step is to explore the relations when everything is allowed to vary. If this is the case, and after setting the first derivatives equal to zero, the first order conditions are:  5.23  The independent producer  w(v)=  ac(vr) avr  5.24  The cooperative member farmer  1 af(vjOw(v) 3v FoP(y)Fof(viaf(vP)avPlf() (__L+1’L +p(y) I., öv ) öv öv avc 11  [  p(y)f(vj—w(v)v  j  v:  övc +  p(y)f(v)— w(v)v  tv:’  93  +  öw(v) c3v Ov av:  v: av: av  We look first to the independent farmer. The optimization condition here is identical to the previous case. We maintain the assumption that this farmers impact on the market is too small to be considered. The picture for the IOF, equation 5.25, is a little more complicated. The IOF now takes into account the effect it has on the output market. Labeling this relation we get:  ap(y)  öf(v)  +  f(v  y  t3v  #3v  +  ()af(vP)  öf(v)öv  )+  äw(v)  öv  = w(v)  ovp  +  öv  5.27  Input Market  Input Output Market Effect  MVP  +  =  .  Pnce  ÷  Effect  The output and input effects have grown a little more complicated than for the Cournot case. Now the IOF must take into consideration the response its rival, the cooperative, will make to its actions. On the output side, the cooperative might change its production. This will affect the size of the price response the IOF’s change in production. On the input side, the cooperative’s input purchase response will affect how the input price moves. Overall, one would expect the cooperative to act in the opposite direction to that taken by the IOF. This will result in a mitigation of the effects that the IOF’s own actions are having, tending to increase the amount of input it will utilize. The cooperative will now also take into consideration the responses of its rival, the IOF.  94  ap(y) af(v) ôv, öy  ãf(v)öv av,,  ôv,  C  Effect  +p(y)  i3v  ôv’  Effect  —  af(v)  w(v)v —  Member Response  Market Price 528  p(y)f(v)  f(v)  ãw(v) ôv, —+1 t3v ôv, Pool Price  +  MVP  =MIC+  Effect  The market price effect is analogous to the output market effect in the IOF case. As for the IOF, it includes a component for the response of the cooperative’s rival, the JOF. This effect is also present in the pool price effect. As for the IOF, the cooperative will probably expect the IOF to respond to its production decision in an opposite direction, thus mitigating some of the own effect on the result. The member response effect is a new component here. It is equal to the negative of the patronage dividend, multiplied by the total member production effect of a change in the cooperative’s input level. When the conjectures are wide open, the cooperative will expect a response from its members when it changes the size of the patronage dividend. The effect on the overall objective function of a change in the amount of milk  produced by the membership will be equal to the current size of the patronage dividend, multiplied by the size of the change. All these factors need to be considered when trying to identify the optimal point for the cooperative to operate. When the conjectures are wide open, we finally find a breakdown of the earlier result that the IOF and cooperative will behave in the same way. The member response effect reduces the  impact of the Market price effect and the MVP, thus reducing the size of response required to market changes. The last step is to look at the response of the members. The cooperative member must now consider the effect of the cooperative, its rival, and the other member farmers on this one member’s revenues.  95  ap(y) öf(v) + ö’ f(v 9y öv,, öv ôv -  [  )  öf(V)  —  +  öw(v) (avg öv  +  t3v  v: öv, v. öv  1  2  Patronage Dividend Effect  5.29 —  p(y)f (vj.... w(v)v  2 Dilution Effect  v  2 +  ôv.  p(y)f(v)_ w(v)v  2  L’i tv:  Patronage Dividend  +  +  v)  öc(v) av:  Pnce  Marginal =  Cost  The patronage dividend effect is the term that internalizes the impact of the member on the cooperatives objective function. It is equal to the cooperative’s marginal condition, scaled according to how much of the total milk received by the cooperative is the result of this member, and then multiplied by the amount which this member thinks the cooperative will respond to the member’s change in production. If the member believes the cooperative will not change its input decision, then this term becomes zero. The dilution effect has been modified a little. The term  2• 0v  /ôvc incorporates the effect of a change in  this member’s production on all the other members of the cooperative. The value of this term will usually be less than one, reducing the impact of the dilution effect to a certain extent. The patronage dividend is the same as presented above. If the member is small relative to the total production of the cooperative membership, then the combination of the decoupling and the member’s size will reduce this relation to the earlier result. In most cases that is what one would expect. However, these have been included to show that with the right conditions, the member will internalize some of the cooperative’s problem, and the resulting optimization will not be the same as for the naive case. However, it is still optimal for all farmers to join the cooperative. Why do we find that there are farmers who do not ship their milk to a cooperative? There are several further factors which can explain this. In the first place, milk in BC is not truely delivered to an anonymous pool. Rather it is shipped to a particular processor, and if that processor is unable to use it for as high a class value, it will be shipped to a processor who can. This transaction has a cost, and the processor who transfers the milk away is compensated for the hauling cost. One would expect that the  96  JOF will want to keep a few shippers whose milk is first shipped to them so that they will not have to rely on the highest class of milk being available from another processor. On top of this, if the compensation rate for interplant transfers of milk is higher than the hauling cost to collect milk from the individual farmers, then it may be optimal for the processor to receive some of its milk directly. The farmers may also have an interest in not belonging to the cooperative. All the farmers in this model are making a profit at the margin. The government determined floor price is such that all the production costs of the fanner are fixed. At the margin, the farmer would still make a profit on a further unit of production, should they be able to produce more. The critical factor for the farmer is then the opportunity cost of not belonging to the cooperative. Various non-market factors may for some fanners balance this opportunity cost. Some farmers place a very high value on their independence, and see the ‘collective’ nature of a cooperative as conflicting with their individual values. The opportunity cost of not joining the cooperative would need to be very high before they would join. Other farmers may be unable to join the cooperative for reasons pertaining to the bylaws of the organization. One important way that cooperatives maintain member cooperation is by having strict bylaws with eviction as a punishment for violation. There may be farmers who are banned from being members. At the market equilibrium point, assuming there are a fairly large number of farmers with diverse beliefs and attitudes, one would expect most farmers to belong to the cooperative, but some to choose to remain outside it. If this distribution is going to be stable over time, the profitability of the member fanner must be the same as for the non-member. This can only be assured if the IOF pays a rate that is equivalent to the return that the members are receiving to those farmers shipping to it. Over the long term, one would expect the profitability of the producer’s farms to follow quite closely the profitability of the cooperative, regardless of whether or not the producer was a member. In a market where there are rents being generated for one party, these rents are usually capitalized into the scarcest input required for production. In the BC dairy industry the scarcest input is the right to produce. The farmer must hold a quota if they are legally entitled to produce milk. This quota is tradable between farmers, and fetches a price according to how much farmers are willing to pay. Since the rents that the  97  farmer receives are expected to follow the profitability of the cooperative, one would also expect the price that farmers are willing to pay to follow the profitability of the cooperative. An empirical test of this model could involve looking for a relationship between the price of fluid milk quota and the patronage dividend or pmfitability of the cooperative being considered.  5.4) Summary In this chapter we have explored how an input price pooling scheme affects the relationship between a cooperative and its membership. Under price pooling we see that the cooperative is able to make an optimization decision independent of its membership. The pool serves to partially decouple the cooperative and its membership. When we solve the problem using either naive conjectures or Cournot conjectures the cooperative and the JOF have identical first order conditions, and it is optimal for all farmers belong to the cooperative. Both the cooperative and the IOF purchase milk from the pool for the same price, but cooperative members receive a patronage dividend as a bonus on top of the price. Assuming that there are rents available to the processors, it will pay these to the membership. As a result, the members of the cooperative will always receive a higher effective price than non members. We also find that the competitive yardstick result breaks down. The members are assumed to be atomistic, and act opportunistically. With these assumptions, economic rents cannot be secured in equilibrium by the farmers. They will compete with each other to ship milk, reducing the pool price they are willing to accept, until they are again earning zero economic profits. The lower pool price means that the IOF is able to secure its input requirements for less, and effectively recapture much of the rents that were initially taken by the cooperative.  98  Chapter 6: The Milk Market with a Supply Restriction, a Pooled Milk Supply, and a Minimum Price. The main restrictions that define dairy supply management in British Columbia are a milk pool, an externally determined milk price and an aggregate market size restriction. In chapter three we constructed the general problem and solved it without restrictions, recovering the competitive yardstick result. Chapter four was used to explore how the agents would behave under an aggregate market restriction. In general, the competitive yardstick is restored, qualified by the fact that the volume restriction prevents the true competitive outcome. Chapter four looked at an input price pool, in isolation of any other policies. The price pool allows the cooperative to purse an objective independently of its membership. At the same time, any rents the cooperative manages to capture will be transferred to the competing IOF through the opportunistic actions of the its members. In this chapter we put all the regulations together, the volume restriction, the pool price, and the price floor. As in chapter four, the price pool is only interesting when we look at how it affects the interaction between a cooperative and its membership, and between a cooperative and a competing IOF, which are the only interactions we consider.  6.1) Problem Definition Including the price pool, volume restriction, and price floor, the problem becomes: 6.1 6.2 6.3  6.4  p  =  m  =  p(y)  =  y  =  f(v)-,- f(v)  v  =  v + VP  Iw(v)  1  6.6  Pool Price  such that w(v)  v  —  c(vr  yr  max.  Supply Price.  vsuchthatw(V)<ñi  max {w(v)v’  6.5  Output market demand.  Fp(y)f(V)—  [  )}  Independent farmer’s objective.  w(v)v c  2  max ‘ p(y)f(v) 1  —  w(v)  }  1  j  v + w(v)v  99  —  c(v’  }  Cooperative member’s objective.  IOF objective function.  This construction incorporates the input price pooi, the aggregate volume restriction, and the price floor. The pooi price appears in equation 6.3. The price farmers receive for their milk is distinct from the supply price of equation 6.2. It is the amount processors pay, and farmers receive. The Aggregate volume restriction shows itself in equation 6.8. The total milk delivered to the market must be no greater than V. The price floor also shows itself in equation 6.3. The pool price is not allowed to fall below the price floor given by ñï. The combination of these restrictions means that the pool price identity we constructed in the last chapter no longer applies. Equation 6.9 shows us that the sum of the pool price and the patronage dividend must be no less than the supply price, m(v).  The aggregate volume restriction again demands a lagrangian solution: 6.8  The independent producer  L(vj’,))= w(v)vf 6.9  —  c(vf)÷  [(v  Ip(y)f(v)— w(v)v  v’ + w(v)v  .  =  —  )  c(v: +  —  v)  —  v]  [  The for profit processor  L(v, 6.11  vç)_ vr]  The cooperative member farmer  L(v,?) 6.10  —  )  =  p(y)f (v,)_ w(v)v  +  —  vj_ VP]  The cooperative processor  L(v,?)-  p(y)f(vJ- w(v)V  +  -  v)_ v]  -  We have four objective functions. The milk pool allows the cooperative to make an optimization decision independently of its members. The constraints are identical to those in chapter four. Agents are restricted  100  to produce or use no more than the gap between the production of all the other agents and the aggregate restriction.  6.2) Solution with Naive Conjectures As in chapter five, we fonnulate conjectural assumptions that are consistent with the competitive model.  0 aw(v)_  and  öv  fO,  —=  öv’  .E-=o,  °“()=0,  [1,  i  i  =  j, j,  k = c, p, k = c,p,  1 = c, p 1 = c,p  All actors believe their own actions do not affect input or output prices. The cooperative processor believes its own input decision does not affect the input use decision of the IOF. Similarly the IOF processor believes that its actions do not affect the input choice of the cooperative. Finaily, individual farmers believe that their actions do not affect the production of other farmers in the milk industry. The unique feature of this model is the pool price. With a milk pool, the cooperative is free to choose its input level independently of the decisions of its members. The conjectural relationship between the cooperative and its members is:  E=0 and  3v  —-=0 av  The cooperative believes its own input decision does not affect the production of its members, and the individual members believe that their decisions do not affect the input used by the cooperative. The first order conditions with our naive assumptions are:  101  6.12  The independent producer w(v)=  6.13  a) (9vr  The cooperative member farmer p(y)f(v)_  [P(Y)f(Vc)_  w(v)v  w(v)v  j  -  6.14  1  v  +  w(v) =  i3v,  +  The for profit processor af(v) 0 =  w(v)+  öv  6.15  The cooperative processor  ‘a’)  af(v)  1  1 =  w(v)  +  The first order conditions for the farmer and the IOF processor are little changed. The farmer sets marginal revenue—the price in the input supply market—equal to marginal cost, adjusted by a shadow value reflecting the production constraint. The IOF processor’s optimal input level occurs where marginal value product is equal to the price of the input, scaled by a shadow value. The optimal production point for the cooperative is very similar to that for the IOF. If we rearrange the relation in the table, we get the following:  af(v)  6.16  p(y)  =m(v)-i-).2v7  Under the naive assumptions we are working with, the sum of the production of the cooperative members is constant. This constant is then multiplied by some shadow value to allow the equilibrium to be located. The cooperative maximizes its objective, the size of the patronage dividend, by following the same optimization steps as an IOF with identical technology would choose. The magnitude of the shadow value is unclear a priori; we camiot determine if the IOF and cooperative operate at the same point. The solution for the cooperative member is little changed from that presented in chapter five, becoming:  102  p(y)f(v)— m(v)v  —  p(y)f(v)— m(v)v  “  =  +  ÷  m(v) av:  6.17  Patronage Dividend +  Dilution Effect  + Pnce  Marginal =  Cost  Shadow +  Value  The patronage dividend is the member’s share of the cooperative’s surplus. The dilution effect reflects how a change in member  i’s production dilutes the overall patronage dividend. The price is the amount  the producer receives from the pool. The farmer chooses a production level such that these terms equal the sum of the farm’s marginal cost and the shadow value. With a large number of farmers, the dilution effect approaches zero; members treat the patronage dividend as constant. The shadow value incorporates the effect of the constraint. In this chapter we have overlaid the model of chapter four on the model of chapter five, and added a floor price. We are solving the problem of a pooled market with an aggregate volume restriction guaranteeing  that there are economic rents available to the agents. Joining the first order conditions to locate a market solution, we first see that the cooperative behaves just like an investor owned firm. The market equilibrium is as for two oligopsonist competing in a market with a restricted overall supply of input. In figure 6.1 we are again using the technique introduced in chapter four; the IOF’s cost curves begin at V and its input use is increasing to the left. The legislated market price is given by ñï. All of the surplus below this price goes to the farmer. The difference between ñï and the supply price,  ), is the rents  available to all farmers. This amount is shown as in the figure. The shaded area marked by the label ‘Co-op’s rents’ is the difference between the average value product at this input level and the legislated price, multiplied by the total amount of milk used. This is redistributed to the members. The rents that a  member captures are equal to the difference between the supply price, and the sum of the price floor and the patronage dividend, ñï +  ), shown as  The rents generated by the IOF are not redistributed to the shippers. The absolute size of these rents is  bounded from below by the legislated price, which the IOF must pay. However, since the IOF buys milk  103  v,  V  milk  Figure 6.1: Interaction between the decoupled cooperative and an IOF.  from the pooi, and not directly from its shippers, it need not compete with cooperative members for the milk input. The independent fanner has a fairly simple problem to solve, equate price with marginal cost. Not being a cooperative member, and believing that this individual is an atomistic agent, there is no own effect on the price. However, the cooperative member’s marginal revenue is not equal to the price alone. We assume that the cooperative is capturing positive rents. We need not worry about the marginal revenue and marginal cost and need consider only the differential profits. Figure 6.2 shows the opportunity cost to the farmer of not being a member of the cooperative. The shaded rectangle is the rents that are generated by the cooperative. These are distributed among all the member’s, as shown by the rectangle. If a fanner is not a member, and produces an amount of milk equal to v•, then the area labeled RF is the opportunity cost of not being a member. If all farmers are strict profit maximizers, then one would expect every fanner to be a member of the cooperative. The size of the opportunity cost is governed primarily by the position of the legislated price. This determines the size of the rents available to the cooperative and the IOF. If the milk price is high, there are little rents available to the processor. There is not much for the cooperative to distribute, and a small  104  $  Co-op’s MVP /  1111111111 liii  2  I  -  S  /  iii  II  II lillIllIll  5  1111  I IIIIIII  [ll:f  v,  V  milk  Figure 6.2: Interaction between the decoupled cooperative and an IOF.  opportunity cost to not being a member. If the price is low, there is a large opportunity cost to not being a member. The decoupling of the farmer from the cooperative means that there is no competitive reason for the investor owned firm to pay a price equal to the effective price the members are receiving. The farmer who chooses not to join the cooperative will be distinctly worse off than the member.  6.3) Solution with Open Conjectures Relaxing all the conjectural assumptions, the solution is similar to the Cournot case in the last chapter. However, the Cournot conjectures are unreasonable. If the cooperative purchases more input, or the farmer produces more, other processors or farmers must adjust to keep overall production fixed. This imposes adding up restrictions on some of the conjectures, which can be summarized as:  0 äw(v).  , 0 t3p(y)  —1,and----=O  öv  k=p,c.  The milk price is bounded from below by a price floor, and since we are assuming that the price floor is sufficient to guarantee the required output level, the variability of the pool price is zero. The output demand function has a downward slope. If one processor increases its input use, the other must reduce its  105  input use by an equal amount. Finally, the sum of all production changes for all farmers must be zero. The complete derivatives for this problem are: 6.18  The independent producer  öw(v)  av 6.19  ac(vl’)  1thc÷p1vP+w(v)  by!  3vf1  vf  The cooperative member farmer  af(v) öv 1  [ap(y) [af(vjov j  ay [  ôv ov:  Vc  x  +  v  ()  ôv  p(y)f(vj— m(v)v  —  of(v )av öv av:  —  i’c  ‘  (V)a1C  avj  ãvc  vöv ôc(vc) V’+W(V)  =  The for profit processor  ap(y) [&f(v) + af(v) öv ô 6.21  [  ôv [ov:  p(y)f(vj_ m(v)v (  öw(v) öv c+ 3v + a [ 6.20  öw(v)  [  ãv  ôv öv Jf(vp)  öf(v) ôw(v) Fôv +11 +m(v)+? +p(y) av  ]  =  The cooperative processor  Iop(y) Fa.f(v) + j öy [ av ôv  “  x  f(v)÷p(y) 9f(v ) c  1ôm(v)[öv  1 =  —  [p(Y)f(l’c)_ m(v)v 1avd1 ôv j  J  +1] +  m(v)}  [__  2v  +  The first equation describes the optimal point for the independent farmer. From our assumption that the aggregate market restriction is binding, we first that  ôv/övf  +  is equal to zero. The total  supply of milk on the market cannot be increased. Equation 6.18 reduces to:  6.22  w(v)=  ôc(v’) ‘  övf  Once again we recover the same first order condition we saw with only an aggregate supply restriction. The supply restriction generates economic rents, which translate into a shadow value. The price pool is  106  1  responsible for the pool price relationship replacing the supply price. However, the floor price is assumed to dominate, so that the pool price does not actually have any direct impact on the independent farmer. Equation 6.19 describes the cooperative member’s solution. If we assume that the aggregate market restriction is binding, and that all farmers belong to the cooperative, then the sum  . öv /öv  must  equal zero. With these observation, equation 6.19 reduces to:  op(y) 6.23  I1  ôf(v)öv öf(v)öv f(v öv av: av av: p(y)f(v  )- m(v)v  C  +  ‘c_w(v)PY... v’ öv  )+p(y) avC  =  w(v)  öc(v) +  This equation is easier to interpret than 6.19. We can use the fact that  Ov, /Ovr + 9v /av[  =  0 to  rearrange the top term in the equation.  v  op(y) 3y  af(v) af(v) f(v , 1 öv av  C  )  ôf(V)  +  — w(v) !!-  Market Revenue Effect 6.24 p(y)f(vj-  m(v)v + w(v)  Patronage Dividend  = öc(v)  Pool +  .  Pnce  =  MC  +  Shadow +  Value  The market revenue effect represents the effect that the cooperative member has on the size of the patronage dividend through a change in farm production. It is equal to the first order condition of the cooperative, scaled by the share of the membership’s production this member produces, and the amount by which this member believes that his or her production will cause the cooperative’s input decision to be changed. However, since all the farmers are assumed to be part of the cooperative, it is somewhat unrealistic to suppose that any member believes their production will change the actions of the  107  cooperative. The cooperative is receiving the same amount of milk before and after the rearrangement of production. If a particular output level was optimal before, it should still be optimal. Therefore av/av:  0, and equation 6.19 reduces to:  =  ôc(v)  p(y)f(v)—m(v)v  6.25  ÷w(v)=  The members’ optimal decision is in fact to treat the patronage dividend as fixed. The aggregate market restriction, combined with the fact that all production in the industry comes from cooperative members, means that the dilution effect is no longer a factor. The behavior of the for-profit producer has changed a little relative to the naive case. The input use response of the cooperative must exactly balance the change made by the IOF, so that 6.20 becomes:  a()  öf(v) —  ()of(vP)  af(v) f(v)÷  6.26  =  .3v  8v  ôy  Changes in the processor’s input use decision may affect the output price. The price change is equal to the net change in output of both processors, multiplied by the slope of the demand curve. The processor’s revenue changes by the multiple of the price change and total production at this point. Due to the regulation, there is no net change in aggregate input purchased, and therefore no change in the input price. If the IOF and cooperative have the same marginal productivity, then we reproduce the first order conditions for the naive case. In the previous chapter we found that when the conjectures are wide open, the first order conditions for the IOF and the cooperative are no longer the same. In this chapter we have required öv /avf +  ôv,, /avf  =  0 and  i3v  =  0. 6.21 now takes the following form:  .  6.27  {  ‘(“) [f()  —  j  f(v  3f(v  )+ 108  )}  1  =  m(v)  +  Except for the unusual shadow value, equation 6.26 and equation 6.27 are the same. The combination of the price pool and the aggregate volume restriction completely separates the cooperative from its membership. This result relies on our assumptions about the behavior of the members, and the fact that they will all choose to be members of the cooperative. However, considering that most farmers in the B.C. dairy industry do ship first to a cooperative, this is quite reasonable. A cooperative is governed by a board of directors selected from among the members. This board hires the management group, and also sets down guidelines about organizational activities. The heterogeneous nature of the membership virtually ensures that management will not internalize the farmers’ problem. However, so far as it is possible, the guidelines imposed by the board of directors may do so. In the extreme, when almost all farmers are members of the cooperative, the board may make some decisions that are geared towards maximizing overall farmer welfare. This analysis has been built on the assumption that given the freedom to do so, a cooperative will pursue patronage dividend maximization. This approach affords us the luxury of treating the cooperative’s objective as independent of the members. Separation of the objective functions is critical to the explicit solution we have generated. This may be unreasonable. However, supply management affords the managers more autonomy than a cooperative in an unregulated market could give. In so far as management’s performance is evaluated by the payments it generates, we expect management to seek to maximize the payment to the members. Guidelines and directives from the board of directors simply limit the extent to which management is autonomous. Our principle qualitative result, that the competitive yardstick breaks down, depends on the cooperative pursuing an objective different from maximizing aggregate welfare. Maximizing aggregate welfare is accomplished by minimizing the patronage dividend, driving the market to the competitive solution. If the cooperative chooses a production level such that the competitive solution is not attained, our result still holds. Members like cash payments from their cooperative, and management likes to protect its position by satisfying the membership. These reasons alone make it unlikely that the cooperative will act to reproduce the competitive market solution.  109  6.4) Summary We have now completed our theoretical construction of a model of the B.C. dairy industry. The main features of this industry are an aggregate volume restriction, a price pool and a price floor. On their own, the components of the supply management system partially decouple the cooperative from its membership. The aggregate volume restriction accomplishes this by forcing rents to exist in the industry, and thereby preventing the complete competitive yardstick result we would expect from the traditional theory. The input price pool breaks the competitive yardstick by allowing the cooperative to pursue an objective independently of its membership. The result of this is that it may be able to pay a patronage dividend that contains economic rents. However, we have assumed that the farmers are opportunistic profit maximizes. They will compete the rents they are capturing away in this model by offering to accept a lower milk price from the pool. The final result is that any rents which the cooperative is able to capture will be transferred to the IOF competitor. When we add a price floor to the model we fix a distribution of the rents between the producers and the processors. When combined with the aggregate volume restriction it acts as a lower bound to the share of the rents which an oligopsonist might capture. With a cooperative in the market, the price floor has little effect beyond possibly restricting the small remaining rents that the competing IOF can extract. When we put all the restrictions together in one model, we find that the cooperative and the member are totally decoupled. It is still optimal for all the farmers to belong to the cooperative. However, there are no longer any interaction effects between the cooperative and its membership, even when beliefs are not restricted. The cooperative and IOF behave like an oligopsony, and members receive half of the industry rents as their patronage dividend. The patronage dividend, which is surplus to the rents which accrue to farmers from the price floor and output control, encourages all farmers to join the cooperative.  110  Chapter 7: A Linear Example. In this chapter we explicitly extract the theoretical results shown above. We build the market solution for three different market structures, the unregulated competitive market, a market with two oligopsonistic firms, and a market with a cooperative and an IOF in competition. Each of these is developed under five different policy environments, an unregulated market in isolation, a market in an importing country, an input supply restricted market, a market with an input floor price, and a market with a price pool for the suppliers. The underlying demand and supply curves are specified as linear and marginal processing cost are fixed, independent of production.  7.1) The Unregulated Market in Isolation. The problem of the unregulated market in isolation is specified as:  7.1  p(X)  7.2  c(x) = i  7.3  m(X)  =  =  10  —  Output Demand Curve  X  Marginal Processing Cost Input Supply Curve  X/2  Xis the total supply of the input brought to the market. We are assuming that the processing technology is perfectly efficient, but involves a cost of one dollar per unit of input processed. The competitive outcome is located by the intersection between the output demand curve and the output supply curve—the sum of the marginal processing cost and the input supply curve. The equation we solve is: (10—  7.4.  generating a solution where  X  =  X)  =  X / 2+1  6.  Figure 7-1 shows the outcome when we assume that the competitive result will occur. The price in the output market, p, is equal to 4. This price is found as the intersection between the output demand curve, DD, and the derived output supply curveS’S’. The price paid to the suppliers of the input is located at the  111  10  s, S 5-  D  Figure 7.1: Unregulated competitive market. intersection between the quantity of input demanded by the processors,  Q*,  and the input supply curve SS,  3 dollars. With two oligopolists, the problem constmction becomes:  7.5  =  1 p(X)x  7.6  =  2 p(X)x  —  —  )x 1 c(x  —  )x 2 c(x  First Oligopsonist  1 ,n(X)x  —  Second Oligopsonist  2 m(X)x  With two firms processing the total input supplied, the input level X is equal to the sum of the input  1 +x . Replacing the functional representations with their explicit 2 choices of the two oligopsonists, x form, and substituting for the total supply, the profit function for the first oligopsonist can be written as:  The symmetry of the problem forces the solution for the second oligopsonist to parallel the solution for the first oligopsonist. This allows us to avoid evaluating the problem twice. The first order condition that follows from this profit function is:  112  s, S  D Qo  5  10  Figure 7.2: Unregulated competitive market. 7.8  Tht 1 /öx  =  9— 3x 1  —  2 (3/2)x  This equation defines the rate of change of the oligopsonist’s profit with respect to the level of input it chooses to purchase and the amount that its competitor chooses. For simplicity, we will avoid dealing with the two stage leader-follower game. Setting this equal to zero and rearranging to isolate the oligopsonist’s own input amount generates the set of best responses to the rivals production decision. This reaction function is:  7.9  =3—x 1 x / 2 2  The synunetry of the profit functions of the two oligopsonists means that these reaction functions are  identical. We use this fact to solve for the joint equilibrium point. For this case, each oligopsonist will choose to use 2 units of input, so that the total input purchased in the market is equal to 4. The market outcome when there are two oligopsonists is shown in figure 7.2. The oligopsonists will jointly short the market, raising the consumer price top° and causing the price paid to producers to fall to m°. The shaded area represents the profits generated by the oligopsonists in this market. The  113  oligopsonists have captured a portion of the consumer and pmducer surplus, and a deadweight loss results. To complete this analysis, we replace one of the oligopsonists with a cooperative. The cooperative distributes any profits it generates as a dividend to its members. The patronage dividend and profit functions for the cooperative and the IOF are:  7.10 7.11  div  =  , c(x)x m(X)x 1 p(X)x  Investor Owned Firm  =  [p(x)x  Cooperative  —  —  —  c(x)xJ,/x  The traditional cooperative acts as an agent for its members, marketing the member’s production. The return the members see for their production is equal to the net return per unit generated by the cooperative. Once again, the total input that is used is equal to the sum of the production of the cooperative and the IOF,  x,, ÷ xe.. We can immediately simplify the cooperative’s dividend function by  eliminating terms common to the numerator and denominator. Doing this we reduce the dividend to:  7.12.  div  =  p(X)  —  c(x)  The cooperative dividend is equal to the profit per unit of input used, which is the difference between the price received for the output and the net cost of the production process. A solution can be generated by assuming that all producers of the input are strict profit maximizers. If this is true, then the cooperative dividend must equal the supply price. Adding this condition to the definition of the cooperative dividend we get:  7.13.  p(X)  —  c(x) = m(X)  When we substitute the functional definitions and the adding up condition into 7.13, we get:  7.14  [io_(x  ÷x)J—i=(x ÷x)/2  114  D  s, S  10  Figure 7.3: Unregulated competitive market in an open economy. If we simplify this relationship, the outcome we get is:  7.15  x÷x=6  This shows us the ‘competitive yardstick’ result that is often presented in the literature [Cotterill, 1987]. The cooperative will choose an input level such that the total production in the industry is equal to the competitive outcome. If we evaluate the objective of the IOF we get a reaction function that is identical to equation 7.9 above. Solving 7.9 and 7.15 simultaneously we find that the cooperative will capture the entire market. This extreme result is due to the linear nature of the curves used, but highlights the outcome predicted by Cotterill [1987] and others.  7.2) Open Economy in an Importing Country We introduce an open economy model by adding a world price for the final product that is below the competitive outcome in this economy in isolation. Output will only be produced where the total cost of processing it and procuring the required input is less than this world price. Figure 7.3 shows the market result when the final product is available for a world price ofp. The consumers demand QC of the output. However, at the world price, domestic processors demand only QC  115  for their use. This input demand results in an input price of m’. Total imports equal the difference between Q’ and  Qd  With an oligopsony, the world price prevents the oligopsonists from capturing rents at the expense of the consumer. However, we are assuming that only the output is traded, so that rents can still be extracted from the producers. Raw milk is a highly perishable product, which contains more weight and volume than many of the products it is processed into, making it unlikely that raw milk will be imported. The objective function for the oligopsonists, after the identifying substitutions have been made are:  =  2  =  pxi  —  1 x  —  —  1 [(x  —  +2  First Oligopsonist  )/2},  Second Oligopsonist  [(xi ÷ 2 )/2}2  Taking the first order conditions for 7.16, and rearranging to get the reaction function, we find:  7.18  1 x  =  —1)— x /2 2  In our example the world price has been set at 3 dollars. Using this price, and the symmetry of the problem, we can solve for the oligopsonist’s optimal production level. In this case, it is equal to 4/3.  In figure 7.4 we see the outcome when there are two oligopsonists purchasing the production of the domestic producers. The oligopsonists are still able to secure some of the surplus that goes to the domestic producers in the competitive case. However, they are unable to get any of the consumer’s surplus. The rent seeking behavior of the oligopsonists reduces the price that is paid to the producers, m’ from 2 to 4/3. At the same time, imports, the difference between QC and  Q”, are expanded from 3 to  4 1/3 units. Consumer surplus is unchanged, producer surplus is decreased, and processor profits are increased. Replacing one of the oligopsonists with a cooperative generates the following patronage dividend and profit functions:  116  D  s, S  5  10  QC  Figure 7.4: Oligopsonist in an open economy for the final output. 7.19  7.20  r  =  div  =  —  [px  x  —  —  [(xe  + ,,  Investor Owned Firm  )/2}  Cooperative  x,  Again we solve the problem for the cooperative by requiring the cooperative dividend to equal the fanner’s supply price. After substituting the supply function for the dividend, and rearranging, we find that the cooperative solution is identified by the relation  7.21  x, +x  =  4  The output point chosen by the cooperative, as a result of its interaction with its members, is such that the industry output is in aggregate equal to the competitive solution. The competitive yardstick has surfaced again. If we solve the IOF’s reaction function simultaneously with the cooperative’s solution identity, we again get the solution that the cooperative would capture the entire market.  7.3) Closed Economy with an Aggregate Volume Restriction Next we investigate the effect of an aggregate volume restriction, in isolation of any other policies. The production quota shorts the market, and restricts consumer purchases to those that are willing to pay the  117  s, S  m  D  Figure 7.5: Quantity restricted competitive market. higher price. At the same time, it reduces the supply price that producers must see to produce the required level of input. The increasing difference between the consumer price and the supply price generates rents which will be captured by some or all of the agents in the market. Figure 7.5 shows the effects of an aggregate quantity restriction. The price that consumers are required to pay has been increased top, while the producer’s supply price has fallen to m. The shaded rectangle represents the excess rents or profits that will be available to some group of agents in this economy. It is unclear who will get them, given the specification so far. The presence of arbitrage opportunities on the consumer side will quickly be eliminated, so that either the processors or the producers are capturing these rents. The oligopsonist’s problem is specified as:  7.22  =  [io  —  1 (x  2 +x  )}  —  1 x  —  1 [(x  + 2 )/2}  First Oligopsonist  2 1 Q—x subjecttox 7.23 2  =  [io  —  1 (x  +  2 )}2 x  —  —  1 + 2 )/2}2 [(x  1 2 Q—x subjecttox  118  Second Oligopsonist  The oligopsonist maximizes its return subject to a restricted aggregate market size. To optimize this we construct a lagrangian function:  7.24  , 1 L(x  1)  =  [io  —  )}  1 +x (x 2  —  1 x  —  1 +x [(x 2 )/2} ÷  +x 2  )  —  ] 1 x  The first order condition that follows from 7.22, assuming that the constraint is binding, is:  7.25  1 öL/öx  =  9— 3x 1  —  2 (3/2)x  —  We set this equal to zero, and rearrange to get the reaction function for the oligopsonist.  7.26  /2—A, 2 =3—x / 3 1 1 x  The symmetric technologies of the two firms allows us to solve for the intersection of the reaction functions. The production point for firm one is identified by the relation  7.27  1 x  We also know that the sum of the production of the two oligopsonists is equal to is three. This would allow us to place limits on the value of  Q, which in this example  ? and ? 2• The principle outcome of the  oligopsonistic model is that the oligopsonists may capture all of the rents available. However, we cannot say how the rents are distributed between the oligopsonists. The actual distribution of market share will determine what the shadow value for each firm. The shadow value indicates that this situation may be unstable. There are profits available at the margin. In the real world these may be absorbed by severe competition between the firms, with the result that the producers and the consumers may end up capturing some of the rents. Replacing one of the oligopsonists with a cooperative, the patronage dividend and profit functions become  119  7.28  div =  {[io  —  (x +  )} }/ —  Cooperative  ,  subject tox Q—x 7.29  =  [io  —  (x + x, )jx,  —  subject to x,,  x,,  —  [(xe +  )/2Jx  For Profit Processor  Q x —  The cooperative pays its dividend to its members. However, the dividend will be greater than the supply price. We need to introduce a shadow value to absorb the difference between the cooperative dividend and the supply price. If we assume that the producers are strict profit maximizers, then the IOF will have to match the cooperative’s dividend price to receive milk. Introducing such a producer’s shadow value into the equation, we get:  7.30  9—(x +x)= (x +x)/2+.,  We know that the total input use must equal  Q if the constraint is binding. The cooperative will pay a  dividend equal to the output price, less the processing cost. For the input market, the presence of a cooperative once again restores the competitive outcome. Further, the resulting output price will exhaust all the profits attainable by the IOF, so that the cooperative may take the entire market.  7.4) Closed Economy with a Floor Price A floor price is introduced as a way of guaranteeing a certain level of producer welfare. However, a floor price is not sustainable without some means of supporting it. There are at least three ways of supporting a floor price: a consumer subsidy, a government offer to purchase, and an aggregate supply restriction. A consumer subsidy is modeled as a subsidy provided by the government to allow the market to clear. An offer to purchase can be modeled in two ways, as a government payment to purchase the surplus from the market, at either the price transacted in the market or a predetermined government price. The aggregate  120  QS  10  Figure 7.6: Competitive market with a floor price. restriction will be set at the intersection between the output supply price derived from the price floor and the output demand function, and may be supported by a consumer subsidy to clear the market. Figure 7.6 shows the impact of a floor price on the competitive market. The floor price has been established atpf, where producers want to supply Q. However, consumers only want to purchase  Q” at  the after processing price of pS• Some supporting program must be introduced to maintain the support price. A consumer subsidy in the amount of QS multiplied by the difference between the after processing price p 5 and the demand price at Q 5  d  will allow the market to clear. Alternatively, the government  could purchase the surplus amount, the difference between Q and  Qd,  at a price  . 5 ofp  One further  alternative the government could use is to impose an aggregate supply restriction, setting the aggregate production quota at Q. Under this option, producers would loose some of the surplus possible if production was unrestricted. This is the triangle bounded above by the floor price, on the left by the domestic consumption Qd, and below by the input supply curve. Evaluating the problem when there are two oligopsonistic processors or a for profit firm and a cooperative when the price floor is supported through government expenditures is very complex. Each agent has two decisions to make, how much to supply to the market, and how much to supply to the supporting agency.  121  Since the oligopsonist has the most market power, we suspect that they will be the greatest beneficiaries of the government expenditure. When a cooperative is introduced, the IOF will still have to match its price, exhausting its rents. However, the government is an extra player in the game, against which the cooperative may be able to act strategically.  7.5) Closed Economy with an Input Price Pool An input price pool is easy to evaluate without a cooperative presence. We get the same outcome as if there was no pool. When there is no connection between the processor and the producer, the situation is identical to a price pool. A price pool forces one price to prevail. Generally one price is assumed to prevail anyhow, so nothing about the result will change. When there is a small number of firms, one of which is a cooperative, a price pool generates a unique result. The new objective functions are:  7.31  7.32  div  =  p(x)x  =  [p(x)x  —  —  c(xjx c(xjx  —  —  M(x)x  Investor Owned Firm  M(x)x],/x  Cooperative  We assume all farmers are members, as predicted by the theoretical development above.  M(X) is the  price that must be paid for milk that is bought from the pool. We assume farmers are strict profit maximizers. Each farmer chooses an output level such that the effective price received is equal to the marginal cost. The producers take the effective price as given. In chapter five we showed that if farmers are strictly profit maximizers, then a price pool transfers any rents the cooperative is able to earn to the competing IOF. When all farmers are members of the cooperative, the effective price is the sum of the pool price and the patronage dividend. The patronage dividend is paid to all the farmers, so that the net cost of producing the required output is equal to the  122  supply price less the patronage dividend. The relationship between the pooi price, the cooperative patronage dividend, and the supply price can be written as:  M(X) =  7.33  m(x) div —  This pool price identity defines the condition that must hold for a cooperative operating in a market where there is a price pooi for the members’ production, and all the members are atomistic and strict profit maximizers.  After substituting the cooperative dividend into this identity and rearranging, the pool price  M(X) = {Xm(X)  7.34  —  M(X) is:  p(x)x + C’(x ),ç J/(x x) —  The pool price is equal to a weighted sum of the supply price, the output demand price, and the cooperative’s processing cost. This function defines the price that the IOF and the cooperative must pay for the input that is purchased from the pooi. Substituting 3.34 into the IOF’s profit function, and simplifying, we find:  7.35  3t  =  p(X)X m(X)X —  —  c(x )x , C(x )x 1 —  The profit of the IOF is equal to the total revenues from the market, less the total cost of all the input produced, and less the processing costs of both the cooperative and the IOF. The pooling of all the milk allows the cooperative to pursue an objective independently of the membership. If the objective chosen results in a greater than zero dividend, the member will accept a lower price from the pool. The drop in the pool price allows the IOF to capture all of the rents that are available. We can use the explicit functions that we have been using above to solve for the solution in this problem. The pool price function becomes:  7.36  M(x  (3x c +x v C  ,  )_  Xx +x )_18x c  2x  123  In all the other cases looked at, the input choice of the IOF did not have any effect on the price that the producers received for their production. The cooperative chose an input level such that the competitive outcome was maintained. The cooperative’s response to the IOF’s decision always restored the zero profit result for the IOF. However, the pooling of the input generates a pool price response to the IOF’s input decision. Figure 7.7 shows the pool price as a function of the total output level. The cooperative input use has been fixed at 2, 3, and 4 units. Therefore the curves represent the price that the IOF faces for the milk delivered to the pool. The farmer is prepared to produce milk for this price. The linear demand and supply curves, combined with the fixed unitary cost and perfect efficiency of the production process generates the peculiar price lines that become asymptotic to the cooperative’s output level, going to negative zero. As expected, the curves all intersect the supply price at a production level of 6, where in a competitive market all the rents would be exhausted. These curves show that for this problem, the optimal input level the IOF will choose will be infinitesimally close to zero. A more realistic cost structure would produce a more believable result. If we substitute this into the cooperative patronage dividend function we find that the relationship simplifies to:  7.37  div  3xr =  / —  2x  +x  If the cooperative is assumed to maximize the patronage dividend, then we can derive a reaction function for the cooperative.  7.38  x  =  3_x/2  When we substitute the pool price function into the JOF profit function we find  7.39  2 i-x) / 9(x +x)_ 3(x 2  124  Figure 7.7: Pool price as a function of total milk brought to pool  The reaction function that follows from this relation is:  7.40  x+x=3  The intersection of these two reaction functions occurs where the cooperative is producing three unils of output for the consumer market, and the IOF is producing nothing. The linear functions force the intersection point to the boundary. The IOF allows the cooperative to have most of the production,, and the cooperative’s patronage dividend acts like a subsidy, allowing the IOF to purchase its input from the  pool at a much reduced price.  7.6) Summary On their own, each of the policy instruments of supply management has a distorting effect on the market. Closing the border increases the domestic price, reducing consumer welfare and increasing producer welfare. Processors with market power extract rents from both the consumer and producer in a closed  125  market, but can only act strategically against the producer if the market is open. The presence of a cooperative restores the competitive outcome in both of these cases. If we add a volume restriction to a closed market we fix available rents. The distribution of the rents is unclear, but likely accrue to the agents with the greatest amount of bargaining power. When there is a cooperative, the farmers become one of the agents with bargaining power, and all the rents are captured by the farmers. Establishing a floor price adds a new level of complexity. It is not possible to maintain a floor price without some form of government intervention to ensure that it is not optimal for the market agents to circumvent the established price. Assuming such regulations are in place, producer welfare will be increased. Consumer surplus may be increased or decreased depending on the form of the government intervention. If agents have market power, then these agents will interact strategically with the new player, the government, to capture a share of the intervention benefits. If it can strategically manipulate the government controls, a cooperative may not perfectly restore the competitive outcome. A price pool decouples the membership from the cooperative. Member production need not be processed by the cooperative, so the member’s own production decision does not need to change the amount processed by the cooperative. The cooperative can pursue an objective independent of the producer members. If the cooperative processes such that it is paying a positive dividend, it becomes optimal for farmers to belong to the cooperative. Members will accept a lower price from the milk pool. They treat the patronage dividend as a production subsidy. The lower pool price transfers the rents that the cooperative captures to the IOF competing with it. In the extreme case presented above, the IOF will choose to produce a very small amount, allowing the cooperative to act monopolistically in the output market. The subsidy on the pool price transfers all of the industry’s rents to the IOF. Supply management combines all of these policies into one program. The most critical impact on the industry is a function of the most binding constraint. Import controls may increase farmer welfare by allowing them to receive a price above the open economy price. The supply restriction may further increase farmer welfare by increasing the input price above the competitive equilibrium. The floor price guarantees that some portion of the rents created by the supply restriction accrue to the farmers. A  126  cooperative in an environment like this ensures that all rents go to the farmer. If we pool the price for milk, we may reduce farmer welfare. The pooling of the milk price allows the IOF to get away without paying a price that matches the dividend the cooperative pays, so that it can capture some share of the industry rents which would go to the farmer without this pool. However, the pool also allows the cooperative to pursue its own objective, with the result that total industry rents may be increased. In aggregate, the combination of all these policies enhance producer welfare, and reduce consumer welfare. The competitive yardstick outcome that standard cooperative theory predicts is unlikely to occur.  127  Chapter 8: The Financial Statements of the FVMPCA We have explored the definition of a cooperative, and the details of the dairy industry in British Columbia. Using these facts we have constructed and evaluated a stylized model of a cooperative in this industry, and investigated how the various elements of this environment impact on the cooperative. The cooperative behaves more like an JOF than we might expect at first glance. We now turn to the empirical structure of the dairy industry in B.C.. We begin by exploring the financial statements of the FVMPCA. After this we turn to a simple model of the quota price to see if it lends support to our model. The financial statements of a cooperative are somewhat different than that of an investor owned finn. The primary purpose of the traditional financial statement is to give information to the owners of the firm, who are the residual claimants. The residual claim is conducted through the return generated by the shares that the equity holder has. This return comes as dividends paid out or an increase in the value of the share. For a cooperative the equity has a return that is either zero or fixed at a low level. The profit generated by the cooperative is returned to the members via the patronage dividend. This dividend can appear in the form of an increased price for the conunodity the member ships to the cooperative, a reduced price for any goods the member buys from the cooperative, a final payment to the member of any surplus the cooperative generates, distributed according to the members patronage, etc. Note that a final payment is distinct from the patronage dividend. The final payment is the amount the farmers would see as cash in a particular period, while the patronage dividend is the member’s share of the cooperative’s earnings, irrespective of whether it is paid out or not. What a cooperative does not do is pay a dividend on the shares held, and the member cannot sell these shares to realize any capital gains in the asset base they represent. These differences make most of the traditional ratios used to analyze a financial statement unusable, and makes comparison between cooperatives and investor owned firms very difficult. The extensive set of ratios that reflect the position of the common shareholder are meaningless when a cooperative is  128  considered. This also means that any effort to compare cooperatives to IOFs on the basis of the shareholder’s position is of little value. The ratios that analyze the positions of the creditors are of a little more value, as are some of the general performance indicators. Several of these will be presented below, along with an explanation of how they are calculated and their meaning. Following the more common ratios, several ratios that are unique to the cooperative will be shown.  8.1) Return on TotalAssets The return on total assets measures how well the assets have been utilized. For the traditional firm this is calculated as:  Income before extraordinary items Tax rate)] + [Interest expense x (i —  =  Average total assets  Return on total assets  One takes the net income and then adds back the interest expense, adjusted  by the tax rate. The  adjustment is made to correct for the fact that taxes are not paid on the interest expense. The average assets can be calculated as the average of the beginning and end of year asset values. This is not strictly correct, but is adequate for the analysis. What this ratio measures is the return the shareholders would realize if the entire firm was equity financed, and was still in the same tax bracket. In the case of the cooperative the calculation we make is similar.  Income before extraordinary items + Interest expense Average total assets  =  Return on total assets  A cooperative is essentially an agent for its members, and so in Canada it does not pay any taxes on its earnings. This somewhat changes the interpretation of the ratio. The return on total assets for a cooperative measures how much would be available, on an asset basis, for distribution to the owners of the firm. However, in a cooperative the returns are paid to the owners according to how much they patronize  129  the cooperative. As a result, this number is, strictly speaking, of little value to the cooperative members. However, it can serve as a basis for comparison to the IOF firms that operate in the same industry. The return to assets was calculated for the investor owned firm using the data available from the Statistics Canada database CANSIM. This data series extends from 1972 to 1987. The total interest expense was calculated as the sum of the bond interest, mortgage interest, and other interest expenses. The appropriate tax rate was calculated by using the assumption that the difference between the net income before taxes and the net income after taxes is equal to the taxes paid. These numbers were used to calculate the equivalent to equity financed amount that had to be added to the net income after taxes to calculate the return on total assets. For the cooperative the amount available for distribution, less the market value of the milk that the members have shipped to the cooperative, was taken to be equivalent to the net income after taxes. This amount is the remainder after all production, interest, and tax expenses are deducted. However, the tax expenses for the cooperative are minimal, so they are ignored. The interest on long term debt and the interest due on loan certificates was then added back to the amount available for distribution, and this was then divided by the total assets. This would be the return on assets if the cooperative did not have to pay  Return to Assets Comparison 0.40  •  0.35  0.30  •  FVMPCA + Taxes  A  CANSIM  0.25 0.20  > 0 4-  0.15 0.10 0.05 0.00 1%  -0.05 •  -0.10  it  0 U it  ,-  N  cq U  U it  U  U  to  C  I-  C  0  0  U  U  0  U  (.)  ()  —  0000000<0000000000  Figure 8.1: Return on assets comparison  130  FVMPCA  FVMPCA  +  taxes  CANSIM  Average return  Variance of return  Correlat’n with time  Slope of regression  Goodness of fit  14.01%  0.00861  0.19  0.0028  0.04  8.94%  0.00478  0.09  0.0013  0.01  8.97%  0.00021  0.63  0.0019  0.40  Table 8.1: Return on assets comparison between the FVMPCA and the average Canadian dairy processor out any interest. This would be the effective return to a totally equity financed processor if that processor could escape taxes. A more appropriate comparison might be with the taxes deducted from the cooperative’s income. The tax rate that was calculated above for the average Canadian firm was used as the appropriate tax rate that would apply to the FVMPCA, should it be more traditionally funded. These results are presented as the second line in the figure. Table 8.1 presents a sunury of the results. The average return for the cooperative assets, given the tax benefit the cooperative, is slightly over 14% for the twenty years considered. However, if we deduct an amount equivalent to the taxes that would be paid for an average Canadian dairy processor, we get an average return that differs from that calculated for the IOFs in Canada by only 0.03%. The correlation with time and the slope of the time trend regression are included to see what is happening over the twenty year interval. The regression is between the data series mentioned and time. Time was a series of day numbers, and the reported number was multiplied by 365 to get a proxy for an annual figure. The main purpose of the regression is to get a sign on the slope parameter and a rough estimate of the amount of the variation explained by time. All the results show a positive trend, but it is weak for the individual firm. In terms of assets, the FVMPCA has been using them about as effectively as any other firm in the industry, both in terms of the average return generated and the general trend over time. Based on the Return on Total Assets, we cannot say that the FVMPCA behaved any differently than any other firm in the industry.  131  8.2) Current Ratio The current ratio is a commonly used measure of the liquidity of the firm. Current assets  Current ratio  =  Current liabilities  Short term creditors are concerned with this ratio. It measures the number of times that the firm would be able to pay the short term liabilities that are outstanding, if they all had to be paid now. Current assets are those assets which can be liquidated rapidly. Current liabilities are liabilities that are coming due within the year. Interpretation of the current ratio must be done cautiously. The current assets of the firm include inventory. Increases and decreases of the current ratio can be indicative of poor financial health, or indications of shifts in the inventory management. Holding large inventories can drive the current ratio up, even if the items in the inventory cannot be sold for the value recorded on the books. Disposing of these assets can cause a drop in the current ratio while at the same time actuaily improving the short term liquidity of the firm. For the average Canadian diary processing firm, the current ratio was calculated using CANSIM data.  Current Ratio Comparison 3.00  • •  .ao 2.60  o  o  o  0  0 Q  0 0  0  0 0  0 0  0  0  0 0  0 0  0 0  0 0  0 0  0 0  0 0  CANSIM  0 0  0 0  0 0  000000  Figure 8.2: Current ratios for the FVMPCA and the average Canadian diary processor.  132  0 0  FVMPCA  Average Current Ratio  Variance of return  Correlat’n with time  Slope of regression  Goodness of fit  FVMPCA  1.817  0.0503  -0.42  -0.0148  0.17  CANSIM  1.517  0.0178  0.80  0.0224  0.64  Table 8.2: Summary statistics for the current ratio analysis. Since CANSIM reports the current liability and the current asset totals of the Canadian industry as a whole, it is very straightforward to calculate these numbers. For the FVMPCA, the calculation of the current ratio took a little more work. The biggest issue was aggregation of the data over the twenty year interval. During this period a number of categories on the balance sheet were changed. Some of these categories shifted out of or into the current assets category. A consistent method of calculation was approximated by tracing where the values in a particular category were distributed to or aggregated from when a change occurred. This change was then moved forward to maintain a consistent method of analysis. This works provided that the elements in a particular category have been consistent over time. Figure 8.2 shows a peak in the current ratio of the FVMPCA occurring in 1978, followed by a large drop in 1979. The spike occurred a year after a rearrangement of the items that appear in the current assets portion of the balance sheet, indicating that it is likely not a result of a change in the factors included. Similarly, rearrangements of the balance sheet categories during 1976 and 1977 did not translate into significant movements in the current ratio. However, in 1978 a new comptroller was hired. This could indicate that there might have been some changes in the accounting practices and the management of highly liquid assets during this time. The summary in table 8.2 of the current ratios for the FVMPCA and the national average shows a slight difference. The national value is within two standard deviations of the current ratio for the FVMPCA. The downward trend indicated by the correlation coefficient and the slope of the regression is a result of the higher current ratio for the FVMPCA prior to 1979. The current ratio indicates that for a period in its history the FVMPCA was maintaining a degree of liquidity that was significantly above that of the  133  average finn in the diary industry in Canada. Following 1978 the current ratio for the FVMPCA closely parallels that of the national average. The step-like change in the current ratio probably indicates either an accounting policy change on the part of the FVMPCA, or an error in how the data was made consistent over the long time period. Overall, the FVMPCA still appears to be behaving like an IOF.  8.3) Debt to Equity Ratio The debt to equity ratio, as traditionally calculated, is a rather dubious quantity when one is dealing with a cooperative. A cooperative is financed principally through member capital. This capital is acquired by retaining from the members a portion of their share of the cooperatives earnings for some period. In the case of the FVMPCA this calculation is even more dubious since the FVMPCA allocated the retained surplus as a ‘loan certificate.’ This loan certificate was essentially a bond, and as such these funds were recorded as a liability, and the earnings they generated were considered an interest expense. For the traditional firm, the debt to equity ratio is calculated as:  Total liabilities Shareholde& equity  =  Debt to equity ratio -  -  This ratio measures the degree to which the financing of the firm is provided by equity holders, and to which degree by financiers. The cooperative raises capital by holding back for a period of time a portion of the profit that it has to distribute to the members. However, the members do have a portion of their funds tied up in the cooperative. This member capital can be grouped into four general categories. In the first category are the net short term liabilities to the members by the cooperative. Unlike most investor owned firms, a cooperative transacts most of its business with its owners. This means that a large part of the current liabilities, in the case of a marketing cooperative, are owed to the members. This is a pool of funds with a high turnover, but gives the cooperative a source of funds to use against other short term liabilities.  134  Distribution of Member Capital in the FVMPCA  OD/.  I  .  .  .  o  —  N  (  o•  o•  o0  000066 0 0 0 0 0 0 o o o o o o  o  o  o  .  . 0  0  N  N 0  o  0  o0  o0  0 — 0 000  0  00 0  0  0  o0  o0  o0  0000000000  Figure 83: Composition of member funds in the FVMPCA.  Another category is the allocated long term equity. These funds are the retained earnings that have been held in the cooperative. They have been allocated to the individual members, and are usually scheduled to be paid out at a particular date in the future. The third category is the share capital. Most cooperatives require some up-front contribution of funds that entitles the individual to the benefits of membership. These shares are usually held for the duration of the individual’s membership, and are redeemed at their purchase price when the individual stops transacting with the cooperative. The final category is unallocated equity. This category consists of retained earnings and reserve funds which are not allocated to members for payment. In an investor owned firm these funds affect the price of the shares, and thereby are effectively still part of the owner’s equity. However, in a cooperative this becomes a pool of funds which is not allocated to anyone. For the purpose of the management of the organization however, this pool of funds does provide a source of capital.  135  The figure shows how the members’ funds in the cooperative have been distributed between the four different categories. The disappearance of the share equity is an outcome of the merger with the Shuswap-Okanagan Dairy Industries Cooperative Association (SODICA) in 1982. At this time the share policy was changed so that members would only be required to hold a small number of shares. Before 1982, the retained portion of the cooperative’s surplus was allocated as shares and bonds. Following the merger, all of the retained amount was allocated as the ‘loan certificate’ bond. These different pools of member capital give us several different ways which we could form a proxy for the debt-to-equity ratio. The most obvious of these is to look at the total member capital contribution. This would be:  Total liabilities =  Total member captial in the cooperative  Debt to equity ratio -  -  This ratio would measure the total portion of the asset base which is allocated to members compared to that which is allocated to other creditors. In the case of a marketing cooperative, a large part of the current liabilities will be occupied by amounts owed to members. These amounts are member equity, but are entitled to be paid out in a short period of time. As a measure of the extent to which the asset base is financed by external creditors, including the short term liabilities to the members may not be appropriate. The short term liabilities are of no use as security  for borrowings. If we remove these short term liabilities, we get the following relationship:  Total liabilities Total member captial in the cooperative -  short term liabilities to members  136  .  =  Debt to equity ratio -  -  This more closely approximates the traditional measure of the financial structure of the cooperative. For the cooperative, this ratio measures what share of the total assets is financed by member capital contributed for this purpose, as distinct from short term liabilities. Over the long term, this measure shows how ownership of the asset base has been changing. The figure shows the time path of the Debt-to-Equity ratio for the average dairy processing firm in Canada, as determined using the CANSIM database, and the path of the debt to long term member capital for the FVMPCA. Inspection of these figures indicates that the share of debt financing for the cooperative has grown over the period shown, while the ratio has been slightly downward trending for the industry as a whole. The table shows a statistical summary of the data presented in the figure. We can see that the debt to equity ratio for the cooperative is trending upwards. The correlation with time shows a strong positive trend for the FVMPCA, while there is a weak negative trend with the data drawn from CANSIM. If we  Debt-to-Equity Ratios 3.00  2.50  •  Long Term Member Capital  A  CANSIM —  —  —  — — —  2.00  1.50  1.00  0.50  0.00  00000000000  0000000000  Figure 8.4: Comparison of the debt-to-equity ratio for the FVMPCA and the average Canadian diary processor.  137  Average ratio  Variance of ratio  Correlat’n with time  Slope of regression  Goodness of fit  Member Capital less currentliab.  1.632  0.345  0.73  0.0675  0.53  CANSIM  1.320  0.028  -0.17  -0.0059  0.03  ‘  Table 8.3: Summary statistics for the debt to equity ratio analysis.  look at the slope of the regression of the data, then we see this repeated. The regression shows that over time, the trend line has an upward slope for the FVMPCA debt to equity ratios. At the same time, the slope of the trend line for the data gathered from CANSIM is downward sloping. The statistical analysis confirms the observation that the debt to equity ratio was rising for the FVMPCA. This observation can be seen as an example of the equity crisis that many cooperatives eventually face. As the cooperative ages and becomes more stable, its members are less willing to finance its operations. They no longer share the same motivations and beliefs that existed when the original members formed the cooperative. As a result, the cooperative must rely to a growing extent on other sources of financing. The control of an organization lies with the people who contribute the capital. The equity holders do have the final decision making control, but other financial contributors place their own restrictions on what an organization can do. Excessive reliance on debt financing can place extensive restrictions on a cooperative, reducing its ability to respond to changes in the market environment. Under such conditions, a cooperutive needs some way to increase the level of equity. Refinancing methods may range from appeals to membership for extra contributions to share offerings, all of which have been seen when other cooperatives have encountered this problem. In terms of financing, the FVMPCA is not showing the same trend as the average finn in the Canadian dairy industry. However, this result is explained best by the structural problems we identified in chapter two, and not the competitive yardstick. We cannot hereby establish a production difference between the FVMPCA and a typical IOF.  138  Surplus Above Market Return 25%  20%  15% I  cn 10%  iiiilIiiI  U 0.  0%  Figure 8.5: Percent surplus above market return.  8.4) Surplus above market return This is the first of a  couple of ratios that are unique to a cooperative. If the cooperative is operating in a  market where there is a price for the product it is dealing in, then we can compare the return it generates against that generated by the market price. We will define this ratio as:  Total return generated by cooperative market value of cooperative product Market value of cooperative product -  =  Surplus above market return  This ratio measures the benefit to the members generated relative to the market. In the  case of a  marketing cooperative like the FVMPCA, which deals in only one major product from its members, that being milk, this is a fairly straightforward number to calculate. The supply management system further simplifies this calculation. The market price that the farmer can receive is simply that administered by the milk board.  139  Average ratio  Variance of ratio  Correlat’n with time  Slope of regression  Goodness of fit  Surplus above market return  6.24%  0.0028  0.21  0.0018  0.04  Surplus as capital  34.95%  0.0798  0.34  0.0153  0.11  ItUrfl t4  Table 8.4: Summary statistics for the surplus above market return.  Figure 8.5 shows the percent surplus above the market return. There is no clearly discernible trend, and over the interval considered, the surplus has generally been positive. If we look at the statistical summary in the table, we see the same result. The correlation coefficient and the regression slope show that there is a positive trend, but the low value for the correlation coefficient and the goodness of fit indicates that this relationship is quite weak. The main conclusion we can draw from this is that the FVMPCA has for the most part been able to generate a return for the members on the equity they have invested in the cooperative. If we look at the  return as a return to the members’ long term equity in the cooperative, we have another measure of the activities of the cooperative. This ratio can be expressed as:  Total return generated by cooperative market value of cooperative product Total member captial in the cooperative -  -  —  —  Surplus above market return as return to capital  short term liabilities to members  The table shows the results of this calculation as well. The average value for the return to the capital, as measured in the way above, is over thirty percent. This is the return to all the member capital in the cooperative that is not due to be paid in the next year. The total return to the farmer is higher, since the farmer is being paid a return on the loan certificates. It is not possible to compare these returns to the industry average. This ratio is unique to the cooperative. However, both of the performance measures are positive, and the return to capital appears to be quite high  140  considering the relative stability of the dairy industry. Overall this suggests that we do not see a competitive yardstick outcome in the B.C. dairy industry.  8.5) Percent of Surplus Retained. This is another ratio that is unique to cooperatives. Since a cooperative generates its operating capital by retaining a portion of the surplus it generates on operations, an illuminating piece of information is what portion of the surplus is retained. The retained amount is distinct from the conventional retained earnings. For a cooperative, the retained amount is a fraction of the patronage dividend, allocated to a particular member. It is retained for a number of periods as ‘operating capital’ for the cooperative. The conventional retained earnings is a blind pool of profits that have not been paid out as dividends. We can measure this in two ways. In the first place, we can measure this as a portion of the total surplus that the cooperative generates. This would be calculated as:  Retained portion of net income Net income  =  Share of total surplus retained  As an alternative, we could look at the amount retained as a portion of the difference between the potential market revenues and the surplus generated by the cooperative. For a marketing cooperative, this would be akin to the fraction of the final payment retained., Since the potential market revenue is sometimes greater than the surplus, this value may be negative. However, the cooperative will still most likely retain a positive amount of the surplus. In such a situation this ratio will become very difficult to interpret. For this reason, we will focus on the simpler form shown above. The figure shows the amount of the surplus retained for the FVMPCA. It is clear from the figure that the size of the retention has been declining over time. This is consistent with the observation that the debt to equity ratio has been increasing. The deferred payment is the means by which the cooperative secures its equity. If the inflow into the equity pool declines, then over time the size of the equity pool will decline, leading to a greater share of the assets financed by debt and a larger debt-to-equity ratio.  141  Portion of Surplus Retained 2.5%  2.0%  1  1.5%  I  1.0%  11111111  0.5%  0.0% U  U  U  U  U  a  U  U  U  U  U  U  U  U .  U 0  U 0  U 0  U 0  ; U 0  U 0  U 0  U 0  U 0  Figure 8.7: Percent of the surplus retained as a share of the total surplus earned.  The size of the equity pool in a cooperative is always a matter of concern. The traditional cooperative does not have any fixed equity. The financial stability of the organization is always subject to the demands of its members. Since most people are assumed to be risk averse, they will always be reluctant to conunit capital to the cooperative. From a portfolio perspective, the individual can minimize risk by diversifying out of the dairy industry, rather than investing in vertical integration within the industry. However, to be successful in the market, the cooperative needs access to capital. As the organization grows, the members become more atomistic, and act more like independent agents. At the same time, the need for large pools of funds to finance investments increases. These two forces pull in opposite directions, with the result that the cooperative may stagnate, or be forced to look to non-traditional sources for its financing.  142  8.6) Summary Analyzing the financial statements of a cooperative is a challenging task. Most of the traditional measures of a firm’s performance relate to the position of the shareholders. Since a cooperative does not have shareholders, but rather returns its earnings to its members on the basis of their patronage, many financial ratios do not provide any meaningful information. We introduce a couple of ratios which are unique to the cooperative, and make some inferences from these. The surplus above market return shows how the cooperative is performing relative to the alternatives available to the member. If we evaluate this measure in relation to the amount of member capital, we can see how ‘healthy’ a return the cooperative is generating. The percent of the surplus retained is a measure of the degree to which the members are contributing capital to the cooperative. It tracks how member equity contribution has changed over time. Combining these two new ratios with the traditional ratio analysis, we can construct a picture of the performance of the cooperative. The surplus above market return, along with the return on total assets and the current ratio indicate that the FVMPCA is performing very competitively in the milk market in B.C. In the case of the traditional ratios, the FVMPCA’s performance is closely in line with the performance of other firms in the industry in Canada. The surplus above market return echoes this fact. This measure is on average positive, and has been trending upwards, indicating that the FVMPCA has been successful at generating a return on the equity its members have invested. The ratios that measure the capital structure of the organization tell another interesting story. The debt to equity ratio, even though strictly speaking it cannot be applied to a cooperative, shows a trend to increase the amount of debt financing in the cooperative. This occurs at a time when for the average finn in the industry, there was somewhat of a decrease in the portion of debt. The percent of the surplus retained echoes this. Over the period surveyed, the amount of the surplus that was retained has fallen significantly. These facts indicate that the FVMPCA is beginning to face the same equity constraint that has forced other cooperatives to look for non-traditional sources of financing.  143  Our theoretical model predicts that under the regulatory scheme in British Columbia, a cooperative in the dairy industry would not lead to the competitive yardstick solution. We expect that a cooperative will behave analogously to the IOF it is competing with. An inspection of the financial statements of the FVMPCA lends some support to this model. We see that the FVMPCA shows a return on total assets and a current ratio that is remarkably similar to the industry average. Along with this, the surplus above market return, when expressed as a return on member capital, is quite large. It appears that over the last twenty years the behavior of the FVMPCA has been similar to that of the IOFs it competes with, and it may have been able to capture some rents on behalf of its membership. The one area where we see a difference is in the debt to equity ratio. This ratio has been climbing for the FVMPCA, while it has been stable for the rest of the industry. This can be explained at least in part by a decline in the percent of surplus retained. As the literature predicts, members of the FVMPCA seem to have been reluctant to maintain the same level of capital contribution over time.  144  Chapter 9: Econometric Analysis The financial records of the FVMPCA provide some firm specific support for our theoretical model. We can also look at the price of fluid milk quota for the presence of the implications of our model. Supply management requires all farmers hold a quota equal to the amount of milk they ship. This quota is traded between fanners. We expect its price to be related to the potential return it offers. In this chapter we explore the relationship between the quota price and the expected success of dairy cooperatives in BC.  9.1) Theoretical Background. The theoretical model of a cooperative in an imperfectly competitive supply managed sector points to a relationship between the performance of the cooperative and the price farmers are willing to pay for the quota asset. In British Columbia, for most of the last twenty years, provincially issued fluid milk quota was the only milk quota traded in the province. Federal market sharing quota was traded on par with the fluid quota, depending on the ratio of holdings by the farmer. Milk quota in British Columbia is, for all intensive purposes, a license for a perpetual cash flow [Barichello, 1984]. If one holds quota, then so long as one is able to maintain the production that the quota requires, the cash flow is guaranteed. The size of the payments is determined by an administrative body, in part using a ‘cost of production’ formula. The risk that this cash flow will change is mostly a function of policy risk. Active political involvement by the dairy industry has kept this risk to a minimum. The standard capital asset valuation models used in finance propose that the value of a capital asset will be equal to the present value of the cash flow that this asset can generate, discounted in a way that incorporates the riskiness of this cash flow. Mathematically we can write this as:  9.1.  a=E (1+i)’  The present value of the asset is equal to the sum of the expected value of the discounted cash flows.  145  Calculating the expected cash flow is difficult. A formula determines the size of the cash payments. However the rate at which the flow must be discounted is more difficult to determine, as is any riskiness associated with the cash flow. The expected discount rate is a function of the expected inflation rate and the rate of return required for an asset with similar riskiness. The expected value of the payments depends on the components of the formula and any factors that the farmers feel affect the riskiness or the absolute size of the payments. Our theoretical model finds that the cash flow will be the sum of the pool price and the patronage dividend of the cooperative. Net of costs, this is the sum of the independent farmer’s shadow value, and the patronage dividend, which equals the member’s shadow value  ),  ?. Any factors that affect the  farmer’s expectations for the patronage dividend affect how the price of quota moves. Asset valuation models are plagued by the problem of evaluating expectations. The current value of variables like the interest rate and a particular cash flow are valid only in so far as they contain information about the future. The interest rate is particularly problematic in this regard because historical data contains inconsistencies that can only be supported through unknown information. The theory of capital markets requires that arbitrage opportunities cannot be maintained. This requires that negative interest rates cannot exist. However, if we look at the historical evidence, we find negative real interest rates. This result occurs because agents incorrectly forecast the inflation rate, not because they expected a negative real interest rate. Therefore, the real interest rate is not a good predictor of expected interest rates. At best, it measures expected inflation, assuming that the underlying asset does not change its riskiness, and that the relevant opportunity cost does not change.  9.2) Data. It is impossible to accurately model the expectations of any economic actor. However, we can postulate that certain variables are more important than others in predicting the cash flow of a particular asset. In this case, we are trying to find factors that will predict the cash flow generated by holding milk quota.  146  250 —  —  —  Fluid Quota Price Blended Milk Price Fresh Fluid Price  150 10o  I%.  50  S  -50 C’) C’) C’) C’) C’) C’) C’) C’) C’) C’) C’) C’) C’) C’) C’) e) ) e 000000000000000000 ,000000000)0)00) r-r- r- r- r- r- r- r- r0)0)0)000000000000000000000 C)  )  )  )  I-  1  00000  1  -  I-  1  1  1  I  1  I-  1  V-  Quarter  Figure 9.1: Quota price with fresh fluid price and blend milk price.  The model we have built indicates that the surplus generated by the cooperative should be an important factor in determining how much farmers will be willing to pay for the quota. We therefore need to identify variables that might predict the return generated by being a cooperative member, and the return generated by producing milk. The data presented in figures 9.1, 9.2 and 9.3 shows a twenty year history for the quota price and a number of other variables that might be important. Several of these series, such as the quota price and the composite leading indicators, show a large nominal increase over this period. All data shown has had a linear time trend removed. Each series is also adjusted to have an overall range of about two hundred, centered around a value of 100. These adjustments make the series more comparable. In figure 9.1, the time trend for the price of fluid quota follows a pattern that is similar to the pattern shown by both the blend milk price, which farmers receive, and the fresh milk price as paid by consumers, lagged by two to three years. Looking closely, the extra volatility of the fresh milk price is somewhat  147  250 200 —  —  —  Fluid Quota Price Composite Indicators Prime Loan Rate  150 100  .—-,  / __% I  0  ,•  g  50  I ‘I  -50) C  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  C’)  000000000000000000000000 0)0)0)0)0)0)0)0)0)0)0)000 FFFFF- F- FFFF000000000000000000000)000 1  I  ,-  I-  I-  I-  I  I  ,-  1  Quarter  Figure 9.2: Quota price and economic indicators. more reflective of the quota price. This is particularly evident in the downward movement of both curves towards the right of the figure. Here the fresh milk price and the quota price have almost identical slopes, while the blend price is not as steep. The consumer price is grossly determined by the administrative price represented by the blend price line. All processors pay this blended price to the farmers, making large deviations a practical impossibility. However, fine deviations will occur as a result of the competition between the various participants in the marketing channel. A cooperative processor’s patronage dividend is expected to deviate with variations in the marketing margin. This should change the producers expectations about the return to cooperative membership. The lag effect may result from information delays, and ‘wait and see’ behavior by industry participants. Qualitatively, the figure indicates that the consumer milk price appears to be having some impact on the price farmers are willing to pay for quota. We would not expect this relationship if farmer return was independent of the consumer price level.  148  In figure 9.2 we see a Statisitics Canada composite leading indicator, composed of ten leading indicators tracked by this agency. The indicator have more variability than the quota price, but we can see that it is on a roughly similar path. The peaks and troughs on both sides of the figure are generally consistent with the quota price trend, although the amplitude of the variation is quite different. The peculiar feature of this pattern is that the quota price leads the indicator by between several quarters and several years. The largest discrepancy in the relationship occurs in the center of the figure. Here we can look at the interest rate. The early eighties corresponded to a period of high inflation. The administrative formula is lagged through a moving average, so inflation would eat up some of the profits generated on the farm. This high inflation period seems to correspond to a lower quota price. Overall, the quota price follows a pattern similar to the economic indicators. However, the leading behavior of the quota price hints that farmers who buy quota are incorporating more information than just economic performance variables into their investment decision. Figure 9.3 shows the quota price along with the surplus earned by the FVMPCA per unit of milk received, and the TSE 300 composite index. The cooperative’s surplus follows a pattern that is similar to the quota price, with the quota price lagging behind anywhere from two to six years. We expect the present performance to be some indicator of its future performance, but as for any organization it is not a complete indicator. The predictive value of the present surplus of the cooperative is further complicated by the changes that this organization has undergone. During the period modeled, the FVMPCA acquired one large dairy processor and merged with another dairy cooperative. At the end of the series, the organization was in the final stages of a merger which would lead to a radical restructuring and a huge increase in size. These factors would greatly complicate the process through which members make their expectations, making the surplus earned an unreliable indicator. The TSE 300 composite shows a remarkably close fit with the quota price, if we ignore the significantly higher volatility. Like the other indicators shown in figure 9.2, the TSE 300 seems to lag behind the  149  250  Quota Price  I’ I’ I’  TSE300  200 —  —  —  Surplus per Unit  I’ I’  150  100  I  —  ‘S  •  ‘_. ••  S  5/  S  51  15  —  I  50  S,bs  s S  S  I  I  S  “  S  0 1  Cl  C’)  —  —  _  —  0—Cl  U)  U)  1%  U)  U)  —  —  —  _  — — —  —  i  _  —  —  —  _  N  C’)  — — — — —S —  -50  Year  Figure 9.3: Quota price against surplus earned and TSE 300 composite.  quota price by about a year. However, unlike the other indicators, the TSE 300 does not present an inconsistent relationship during the high inflation period of the early eighties. As shown above, quota prices are often modeled like other asset valuation models. This model comes from capital market theory, which has as its most obvious application the stock market. The agents buying quota likely uses much of the same infonuation as the capital market participant, such as interest and inflation rates. If the nominal cash flow generated by quota were guaranteed, then it would be similar to a bond, which we expect to be quite counter cyclical. It appears to have a smaller variance, and be slightly out of phase with the TSE 300. The security of the cash flow can explain the lower variance, while the stability of the return to producers may account for the leading behavior. The stability might lead farmers to believe their maximum earning potential occurs before the economy peaks, generating a somewhat leading behavior.  150  The relatively close match between the TSE 300 and the quota price and between the composite leading indicator and the quota price supports our model. Many of the factors that increase stock prices, such as increases in consumer spending, also increase the return generated by a marketing cooperative. With supply management, such increases in earning potential are transfonned into a willingness to pay for quota. If the quota is considered as guaranteeing a fixed income stream over time, then one would best consider the quota asset as a bond. The price of a bond is inversely related to the expected inflation rate. By assuming that the riskiness of corporate bonds is fixed, we can assume that the prime corporate interest rate incorporates the same real rate of interest over time. This number then reflects the expected inflation rate that investors believe will occur in the future. The corresponding effect on the willingness to pay for access to this income stream on the part of producers should be negative regardless of the source of the funds. Holding fluid milk quota gives one access to the income generated by the cooperative. We want to separate variables that will indicate future success of the cooperative from those that identify changes in the underlying fixed income stream generated by the quota. Two variables that could indicate this are the market price for the good the cooperative sells, and the rate of return generated by the Toronto Composite 300 (TSE 300) index. The price for the cooperative’s output acts as a relatively short term indicator of cooperative performance. If the cooperative is engaged in a price war, and the price for milk is low, then its profitability will also suffer. A rational member would therefore expect the final payment generated by the cooperative to be reduced, all other things remaining unchanged. If this is true, we would expect to see a positive relationship between the price for milk in the consumer market and the price quota is being traded for. Conversely, if the quota’s income stream is believed to be fixed, then its price should be unrelated to the consumer milk price. The variable we will use is the consumer price index for milk in British Columbia.  151  The TSE 300 is an index of the price of stocks on the Toronto stock exchange. Basic finance theory proposes that the price investors in the capital market are prepared to pay for a share reflects accurately all the presently available information about the future cash flows this stock will generate [Brealey et a!, 1992]. Using a diversified composite index eliminates any finn specific effects from the return calculated, allowing one to see a risk correct expected present value for the income stream generated by the market. The period by period changes in the value of the TSE 300 will reflect changes in the expectations of the investors. The stock market can be seen as indicating two different key pieces of infonnation about the future value of the income stream generated by the fluid milk quota. On the one hand, the TSE 300 return can be seen as reflecting the opportunity cost of keeping ones assets involved in the dairy industry. If this is true, when the TSE 300 falls, the price of fluid quota should rise. One would expect a negative relationship between changes in the value of the stock market and the price of milk quota. On the other hand, the stock market is also an indicator of the strength of the economy in the future. Investors bid up the price of stocks when they discover that the expected future income stream of a stock has increased. If an index of the market increases in value, then investors expect the income stream of most stocks in the market to increase. If this is true, then a cooperative member is also likely to expect that the return generated by the cooperative to increase. A simple causal mechanism may be increased consumer spending which will include increased spending in the food sector. One would expect a positive relationship between changes in the value of the TSE 300 index and the price of fluid quota. Another important factor to consider is technological change. Theoretically, technological innovations are not adopted unless they reduce the cost of production relative to current technology. As technology has improved we expect the present value of the income stream to increase, all other things constant. We expect the relationship between an index of technology and the price for quota to be positive. For this analysis, a time trend was used to try and capture the effect of technical change.  152  Table 9.1.1. Summary statistics for quarterly series.  Milk  Quota Mean Standard Error Median Standard Deviation Standard Dev. Percent Sample Variance Range Minimum Maximum Count  197.34 6.72 205.60 55.44 28.1% 3074.00 226.02 75.89 301.90 68  122.80 0.78 124.65 6.45 5.25% 41.59 26.64 108.28 134.92 68  Rate 11.67 0.35 11 2.89 24.8% 8.36 14.59 7.08 21.67 68  TSE rate 0.20 0.98 0.67 8.10 4050% 65.57 43.87 -24.02 19.85 68  Table 9.1.2: Summary statistics for annual series.  Milk  Quota Mean Standard Error Median Standard Deviation Sample Variance Range Minimum Maximum Count  199.49 13.08 206.80 52.31 2736.57 194.02 85.65 279.67 16  123.25 1.56 124.62 6.24 38.98 21.98 111.08 133.06 16  Rate 11.79 0.70 10.79 2.79 7.78 10.79 8.50 19.29 16  TSE rate -1.57 4.91 1.68 19.65 386.23 70.82 -42.53 28.29 16  The quota values were gathered from three different sources. The data for the 1981 to 1991 period was taken from BC Ministry of Agriculture, Fisheries, and Food publications. The 1972 to 1976 period was covered by data taken from a study by the Fraser Institute [Grubel, 1977]. The interval from 1977 to 1980 was filled in through prices taken from the records of Paten and Smith Auctioneers in the Fraser Valley. The TSE 300 data series, the CPI for milk in Vancouver., and the Prime Rate were taken from Statistics Canada sources. The Quota Price, the TSE 300 close, and the CPI for milk were adjusted using the Industrial Product Price Index to generate a ‘real’ value. Tables 9.1.1 and 9.1.2 show the summary statistics for the quarterly and annual data series being used. The Quota series is the price paid for quota trades. The Milk series is the CPI series for fresh fluid milk from Statistic Canada. The Rate series is the prime bank rate on corporate loans. The TSE rate series is  153  Table 9.2.1. Quarterly Regression Results  Regression Statistics Standard Error 0.8576 Observations 0.7356 Durbin-Watson 0.7029  Multiple R R Square Adjusted R Square  Constant Milk Rate TSE Rate Time  Standard Error 81.87 0.6087 1.379 0.4701 0.1936  Coefficients -46.24 1.508 -1.771 1.042 2.287  29.40 68 0.3533 t Stat -0.5648 2.477 -1.285 2.219 11.81  Table 9.2.2. Annual Regression Results  Regression Statistics Standard Error 0.9274 Observations 0.8601 Durbin-Watson 0.7183  Multiple R R Square Adjusted R Square  Constant Milk Rate TSE rate Time  Standard Error 148.0 1.074 2.378 0.3383 1.405  Coefficients 18.34 1.063 -1.792 0.7721 8.535  22.85 16 1.249 t Stat 0.1239 0.9895 -0.7536 2.282 6.076  the percentage change in the value of the ThE 300 composite between the last period and this period. All the summary statistics are for the data actually used in the regressions. Notice that the milk price variable has a standard deviation equal to 5.25% of the mean, and that the range is only a little over 10% of the mean. The other variables have much more dispersion.  9.3) Results. The hypotheses were tested using both annual and quarterly data. The quarterly data had 9 missing observations over the 68 observations used. These were patched by linear interpolation. The two series were analyzed to see if there would be any difference in results beyond sample size effects. A priori there was no reason to expect a difference in the results. The coefficients generated by the regressions are presented in tables 9.2.1 and 9.2.2.  154  The coefficients on the TSE and the Milk data series have a sign that is consistent with the hypothesis that the quota price is dependent on the profitability of the cooperative. For the annual data, the four variables explain more than 70% of the variation in both cases. For the quarterly data, the parameters on Milk, TSE rate, and Time are significant. If we believe that the quota price is independent of the expected performance of the cooperative, then we would expect the sign of the TSE rate parameter to be negative, reflecting the opportunity cost of remaining in farming. However, the sign is positive, which is consistent with the dependence on expected cooperative return. The fact that the Milk parameter is significant provides a further challenge to the traditional model. If the fanner’s expected cash flow from producing milk is exogenous, then there is no reason to expect it to be related to what the consumer pays for milk. The significance of this variable supports our model that cooperative performance affects quota value. When we move from quarterly data to annual data we go from 68 observations to 16. With annual data we find that the parameter on the milk variable is no longer significant. The return the fanner expects to receive from the cooperative will depend on the cooperative’s market share, the consumer price, and other factors such as the weather which will affect consumer demand. During a hot summer we expect the milk price to be higher, as consumers are demanding more products like ice cream and fluid milk. This would likely indicate a larger final payment in this period, which would be reflected by a higher willingness to pay for the quota. The annual averages would eliminate or flatten many of these seasonal variations, with the result that we may not be able to see the effect. The relatively high R-squared values, along with small t-stats indicates that there is probably a problem with multicollinearity. To explore the interactions between the different variables, we have constructed auxiliary regressions between the different variables. Table 9.3.1 and table 9.3.2 present the results for the annual and quarterly data respectively. There is a high degree of collinearity between the milk price and the constant term. This result agrees with the small variation that this data shows. The multicollinearity between the constant and the milk price is probably largely responsible for the low t-stat for these variables. The signs are all consistent with the expectations, but the confidence interval around  155  the parameter estimates is quite large. This fact would aggravate the difficulty in finding a relationship between the quota price and the milk price. Table 9.3.1. Quarterly Data Auxiliary Regressions.  Milk  R Squared 0.16271  Milk  (0.1078)2 Rate  TSE  Time  Corist  Rate  TSE  Time  Intercept  -0.7270  0.1330  -0.0251  132.1271  (-2.71 06)  (1.3987)  (-0.6345)  (42.0874)  0.1886  -0.1417  -0.0464  0.0329  27.9368  (0.1349)  (-2.7106)  (-1.1000)  (1 .9262)  (4.2664)  0.1096  0.2230  -0.3996  0.1065  -26.1986  (0.0523)  (1.3987)  (-1.1000)  (2.1409)  (-1.2174)  0.1199  -0.2485  1.6670  0.6276  45.4482  (0.0631)  (-0.6345)  (1.9262)  (2.1409)  (0.8650)  1.0000  0.0073  0.0079  -0.0009  0.0003  (0.9844)  (42.0874)  (4.2664)  (-1.2174)  (0.8650)  Table 9.3.2. Annual Data Auxiliary Regressions.  R Square Milk  Rate  TSE  Time  Const  1  Milk  Rate  TSE  Time  Intercept  0.2257  -0.8669  0.0516  -0.2440  135.6243  (0.0322)  (-1.4732)  (0.5755)  (-0.6578)  (19.0317)  0.2091  -0.1767  -0.0184  0.0745  32.9054  (0.0114)  (-1.4732)  (-0.4508)  (0.4402)  (2.1579)  0.2129  0.5201  -0.9071  1.81 48  -70.4018  (0.0162)  (0.5755)  (-0.4508)  (1.6836)  (-0.5648)  0.2220  -0.1426  0.2134  0.1053  23.7237  (0.0275)  (-0.6578)  (0.4402)  (1.6836)  (0.8004)  1.0000  0.0071  0.0085  -0.0004  0.0021  (0.91 67)  (1 9.0317)  (2.1579)  (-0.5648)  (0.8004)  Unadjusted R-Squared  Adjus R-Squared 2 T-Statistic 3  156  Table 9.4.1. Autocorrelation corrected quarterly Regression Results  R Square Adjusted R Square Standard Error  Constant Milk Rate TSE Rate Time  Regression Statistics Observations 0.9341 Durbin-Watson 0.9299 14.68 Rho -  Standard Error 104.3 0.7648 1.712 0.2046 0.6590  Coefficients 37.17 0.9034 -2.4740 0.1557 2.023  68 1.4627 0.9017 t Stat 0.3562 1.182 -1.445 0.7610 3.069  Table 9.4.2. Autocorrelation corrected annual Regression Results  R Square Adjusted R Square Standard Error  Constant Milk Rate TSE rate Time  Regression Statistics Observations 0.8852 Durbin-Watson 0.8435 Rho 20.70 Standard Error 138.7 1.015 2.708 0.2647 2.170  Coefficients 16.41 1.116 -2.649 0.4396 8.698  68 1.2791 0.5547 t Stat 0.1183 1.100 -0.9784 1.660 4.008  Our regressions generated Durbin-Watson values of 1.247 and 0.3533 for the annual and quarterly data respectively. The critical values for the Durbin-Watson statistics are 0.73 and 1.94 for the annual data indicates that we cannot say if there is a problem with autocorrelation. However, the residual plot appears to be quite cyclical, with a normal statistic of -1.941 for the runs test. The small value for the Durbin Watson value with the quarterly data clearly indicates the presence of autocorrelation. When we correct for autocorrelation with the Cochrane-Orcutt procedure, we generate the results in table 9.4. The only parameter estimate which remains significant is the parameter on the time trend. However, all signs except that on the constant term are unchanged. In the case of the annual data, the parameter on the TSE variable retains the largest t-stat after the time trend. It appears that there is a relationship  157  between the TSE 300 index and the quota value, although the TSE 300 is probably not the best proxy for the factors that affect the price of quota. Our empirical hypothesis is that factors which are used by agents who purchase quota includes information investors use when purchasing equity shares. If the farmers willingness to pay for quota depends on the expected return of a cooperative, then the quota price should follow a path similar to the TSE. A proper analysis requires the identification of all the factors that determine economic activity, and possibly regression by simultaneous equation to eliminate any endogeneity problems. This analysis is beyond the scope of the current work. These results indicate that the price farmers are willing to pay for quota rights is a function of more than just the formula price and the interest rate. The results are also consistent with our proposal that the return generated by cooperatives in the BC dairy industry impacts on the price farmers are willing to pay. The principle source of the change in the quota price is the time trend, which reflects technological change. The net cash flow the farmer is receiving is increasing because costs are falling faster than the price received. However, the next best explanatory variables are those that would affect the expectations members have about the return their cooperative generates.  94) Summary In the preceding chapters we developed a theoretical model of the interaction of a cooperative, its members, and an IOF competitor. One principle conclusion of this model was that the opportunity cost to the farmer of the supply management quota will be related to the expected success of the cooperative. In this chapter we investigated how some variables that might be expected to predict the success of the cooperative relate to the quota price. We find that the price of quota is positively related to the consumer milk price and the TSE 300 price, a powerful economic indicator. These results provide empirical support for our model of the relationship between farmers and their marketing cooperative.  158  Conclusion Cooperatives are a pervasive feature in production agriculture throughout the world. This is true in almost all jurisdictions, seemingly irrespective of the different regulations which may exist. This fact may explain why the interplay between the structure of a cooperative and government regulations has not been investigated to any great extent. With this work we have tried to look at this problem for a highly specific case, the dairy industry in British Columbia. We have developed a theoretical model of this market structure, and have looked to the FVMPCA and the quota market for supporting evidence. In the first chapter we identified the key characteristics of a cooperative and looked at some of the implications that follow from these. A cooperative is a form of industrial organization where those who benefit most from the services of the organization also contribute the capital and control its activities. Cooperatives almost by definition do not pay a return on the member’s equity. They return their earnings to their members on the basis of their patronage. Control of the organization is conducted democratically, and membership is usually, though not always, open. There is an inherent attractiveness to the cooperative structure, arising primarily from the egalitarian flavour that naturally follows from the democratic style of the organization. This attractiveness however comes at a price. At least in theory, if not also in reality, cooperatives are plagued by problems arising from conflicting incentives and objectives on the part of its membership. One of the most glaring of these is the unwillingness of the membership to contribute equity to the organization. It is ironic that this problem becomes more significant as the cooperative grows in size, and would otherwise be considered successful. Chapter two developed the specifics of the supply management system. Supply management arose as a response to a prolonged and painful adjustment in the dairy industry. It is the latest and perhaps most pervasive response to the perception, true or not, that dairy farmers suffer from irreconcilable asymmetries in their bargaining power vis-a-vis the processors who buy their production. The main features of this system are an aggregate volume restriction, a restriction or prohibition on imports, a controlled price, and  159  a price pool for all the milk produced. The first three features have been studied at length, and the result that the institutional shorting of the market generates rents which may accrue to the producers is well known. The further result that the atomistic producers will capitalize these rents away has also been widely commented on. The price pooi however is little studied. It is this price pool which we found generates a change in the traditional behavior expected for a cooperative. In chapters three through six we developed a theoretical model of a cooperative and explored how it behaves under the regulations of supply management. When the regulations are not present, we restore the standard competitive yardstick result. We find however, that if there are entry barriers in the processing sector, a cooperative will not be able to completely restore the competitive outcome. The competing IOFs are able to optimize against the producer price, which is now determined by the cooperative’s average value product rather than the producer’s marginal cost. Adding an aggregate volume restriction changes this result little. The principle effect of a supply restriction is to guarantee that there are profits available. For two competing oligopolists, an aggregate restriction on the amount of input available seems to generate a Bertrand outcome from a model with Cournot conjectures. However, the presence of a cooperative essentially restores the competitive yardstick. The competing IOF is still able to use the cooperative’s AVP as the pricing curve, but it can only do so provided that total production remains within the aggregate volume restriction. Adding a price floor to the volume restriction has one principle effect, it restricts how far the down the farmer’s price can fall. A price pool does have a significant impact. Under a price pool, the cooperative is no longer forced to process all of the milk its members produce. The cooperative and the IOF both purchase their input requirement from the common milk pool. There is now little reason for the cooperative to behave differently from the IOF. The patronage dividend now consists of the profits the cooperative is able to generate above the input price, allocated according to the production of the members. Those who choose to belong receive a ‘bonus’ which non-members do not receive, with the result that it is optimal for all farmers to belong to the cooperative. However, if the pool price is determined as an ‘average offer price’  160  from the processors, the net result of the pooling structure will be that all rents captured by the cooperative will be transferred to the IOF. When we put all the pieces together, we find an even stronger result. We now see that regardless of the conjectures, the cooperative will behave exactly like an IOF with identical technology, and the cooperative member will not internalize any of the cooperative’s objective. They will take the patronage dividend as a fixed price, and make their production decision accordingly. The price pool has decoupled the cooperative from its membership, and the volume restriction has removed the impact of the farmer’s production decision on overall industry prices. In chapter seven we built a simple model of the cooperative, based on linear supply, demand, and cost curves. As we expect, the linear functions translate into extreme outcomes. In most of the situations modeled, we find that the competitive yardstick is completely restored. However, the price pool does not reproduce this outcome. Here we find that the IOF will choose to produce nothing, while the cooperative produces at the monopoly point. This result is a consequence of the complete rent transfer to the IOF, because the maximum rents available occur at the monopoly outcome. The financial statements of the FVMPCA are explored in chapter eight. The FVMPCA is the largest receiver of milk in British Columbia, so that it might well represent the cooperative we are modeling. We see in the financial statements that the FVMPCA is generating a return for its members that is almost perfectly in line with the performance of other dairy processors in the Canadian dairy industry. This is what we expect if the cooperative is performing in the same way as an IOF would. However, we also see that the FVMPCA is being plagued by the equity crunch that often befalls cooperatives. In the last chapter we explore how the price of quota might vary with the expected success of the cooperative. The valuation of the quota ‘asset’ is traditionally modeled like any other capital asset. The asset should capture in its value the present value of the expected cash flows it entitles the holder to. Since most of the dairy producers in B.C. are members of the FVMPCA, their expected cash flow from milk production will be tied to the expected success of the FVMPCA. It is difficult to measure ‘expected  161  success.’ However, we propose that the ThE 300 serves as a measure of expected economic strength, and therefore likely indicates something about the expected success of the cooperative. We find that the TSE 300 has significant predictive power for the price of quota, and the sign of the parameter agrees with our proposed model. Our model results and empirical investigations indicate that the supply management system can effectively separate a dairy marketing cooperative from the adverse incentives of its membership. The cooperative is free to follow an objective in the same way as a competing IOF would. At the same time the market for quota between farmers in effect becomes a capital market, where farmers capitalize the expected patronage dividend into the quota price. This result relies on the interaction between all the elements of the supply management system. Each of the components on its own does not produce this result. In the present time of uncertainty, and of pressure to restructure supply management, we need to consider what form we want the industry to take in the future, and how the different policy instruments interact with each other and with the main agents that are active in the dairy industry.  162  References Agricultural Products MarketingAct. R.S., c. A-7, s.1. Queen’s Printer for Canada, Ottawa, 1985. B.C. Legislature. Reg.101/90: Natural Products Marketing Act (BC): British Columbia Milk Marketing Board Regulation. Victoria. 1990. Bailey, Jack. The British Co-operative Movement. Hutchinson & Co. Ltd. 1955. Barichello, Richard R. Analyzing an Agricultural Marketing Quota. University of British Columbia, 1984. Barichello, Richard. The Economics of Canadian Dairy Industry Regulation. University of British Columbia Technical Report No. E/I 2. March 1981. Barichello, Richard. Quota Allocation and Transfer Schemes in Canada. Working Paper, Department of Agricultural Economics, University of British Columbia. 1987. BCJ, British Columbia Judgements. No. 1748. 1993 BCMMB. British Columbia Milk Marketing Board. Annual Report. Vancouver, British Columbia. 1990-1992. BCMMB. British Columbia Milk Marketing Board. General Order 133. Vancouver, British Columbia. 4 October, 1991. BCMMB. British Columbia Milk Marketing Board. General Order 31. Vancouver, British Columbia. 4 October, 1991. Bogardus, Emory S.. Principles of Cooperation. The Cooperative League of the U.S.A.. Chicago, Illinois, U.S.A. 1952. Bonin, John P., Derek C. Jones and Louis Putterman. “Theoretical and Empirical Studies of Producer Cooperatives: Will Ever the Twain Meet?” Journal of Economic Literature. September 1993. ppl29O-132O. Brealey, Richard, Stewart Myers, Gordon Sick, and Ronald Giammarino. Principles of Corporate Finance. McGraw-Hill Ryerson Limited. 1992. Cotterill, Ronald W.. “Agricultural Cooperatives: A Unified Theory of Pricing, Finance, and Investment.” Cooperative Theory: New Approaches. United States Department of Agriculture, Agricultural Cooperative Service, ACS Report #18. pp 258 July 1987. 171 Forbes, J.D., R.D. Hughes, and T.K. Warley. Economic Intervention and Regulation in Canadian Agriculture. Economic Council of Canada and The Institutw for Research on Public Policy. Ottawa, Canada. 1982. FVMPCA. Fraser Valley Milk Producers Cooperative Association. Annual Report. Vancouver, British Columbia. 1971-1992. FVMPCA. Fraser Valley Milk Producers Cooperative Association.  163  1983.  Gandolfo, Giancarlo. Economic Dynamics: Methods and Models. North-Holland Publishing. Netherlands. 1980. Grant, Wyn. The Dairy Industry: An International Comparison. Dartmouth Publishing Company. England. 1991 Grubel, Herbert G. and Richard W. Schwindt. The Real Cost of the B.C. Milk Board: A Case Study in Canadian Agricultural Policy. The Fraser Institute. Vancouver. 1977. Hardie, Ian W. “Shadow Prices as Member Returns for a Marketing Cooperative.” Journal of Farm Economics. (51) 1969. . 833 818 pp Haydu, John Joseph. Barriers and Opportunities Facing Cooperative in Improving the Economic Coordination of the Farm Supply Industry. PhD Dissertation, Michigan State University. 1988. Hazlett, Judith. Alpha: The History of the Central Alberta Dairy Pool. Central Alberta Dairy Pool, 1992. Helmberger, Peter, and Sidney Hoos. “Cooperative Enterprise and Organization Theory.” Journal of Farm Economics. 44:2. May 1962. . 290 275 pp Horowitz, Ira. “On the Effects of Cournot Rivalry between Entrepreneurial and Cooperative Firms.” Journal of Comparative Economics. 15, 115-121 (1991) ILO, International Labour Office. Co-operative Management and Administration: Second (revised) Edition. International Labour Office. Geneva. 1988. Ireland, Norman J.. “The Economic Analysis of Labour-Managed Firms.” Bulletin of Economic Research. 39:4. 249-272 (1987).. Jackson, Robert 3. and Doreen Jackson. Politics in Canada: Culture, Institutions, Behavior, and Public Policy. Prentice-Hall Canada Inc. Scarborough, Ontario. 1990. Law, PJ. and Stewart, G. “Stackleberg Duopoly with an Illyria and Profit Maximizing Firm.” Recherches Economigues de Louvain. Vol 49. 1983. pp 212 207 Lerman, Zvi and Claudia Parliament. Financing of Growth in Agricultural Cooperatives. University of Minnesota Staff Paper 9 1-33. July 1991. St Paul, Minnesota. Levay, Clare. “Some Problems of Agricultural Marketing Co-operatives’ Price/Output Determination in Imperfect Competition.” Canadian Journal of Agricultural Economics. 31. March 1983. pp 105-110. Manchester, Alden C. The Public Role in the Dairy Economy: Why and How Governments Intervene in the Milk Business. Westview Press. Boulder, Colorado. 1983. McKinley, Kathryn. Alternative Arms Length Pricing Mechanisms for Dairy Products in Canada. Commodity Coordination Directorate, Policy Branch, Agriculture Canada. Working Paper, September 1990. Meade, J.E.. “The Theory of Labour Managed Firms and of Profit Sharing.” Economic Journal. 1982. pp402-428. Natural Products Marketing (BC) Act British Columbia Milk Marketing Board Regulation. B.C. Reg. 101/90. Queen’s Printer for British Columbia, Victoria, 1990.  164  Parliament, Claudia, Zvi Lermau and Dana Huseby. Risk and Equity in Agricultural Cooperatives. University of Minnesota Staff Paper P91-37. September 1991. St. Paul, Minnesota. Rhodes, V. James. “Large Agricultural Cooperative, on the Road to Where?” Cooperative Theory: New Approaches. United States Department of Agriculture, Agricultural Cooperative Service, ACS Report #18. July 1987. . 170 155 pp Sertel, Murat R. “Workers’ Enterprises in Imperfect Competition.” Journal of Comparative Economics. 15 pp698-710. 1991. Sexton, Richard and Julie Iskow. Factors Critical to the Success or Failure of Emerging Agricultural Cooperatives. Giannini Foundation Information Series No. 88-3. University of California. 1988. Shaffer, James D.. “Thinking About Farmers’ Cooperatives, Contracts, and Economic Coordination.” Cooperative Theory: New Approaches. United States Department of Agriculture, Agricultural Cooperative Service, ACS Report #18. July 1987. pp6l-86. Shelford, Cyril M. An Overview of the British Columbia Dairy Industry. November 30, 1988. Shelford, Cyril M. Report to the Hon. John Savage, Minister of Agriculture. April 30, 1988. SSCA, Select Standing Committee on Agriculture. Marketing Boards in British Columbia. vol III. British Columbia Legislative Assembly. October 1978. SSCA, Select Standing Committee on Agriculture. Supply Management and Quota Values in Primary Agriculture. British Columbia Legislative Assembly. December 1978. SSCA, Select Standing Committee on Agriculture. Milk Production. Legislative Assembly, Province of British Columbia. 1979. SSCA, Select Standing Committee on Agriculture. The British Columbia Milk Board. Legislative Assembly, British Columbia. 1978. SSCA, Select Standing Committee on Agriculture. The Dairy Industry in British Columbia. Legislative Assembly, British Columbia. 1979. Staatz, John M.. “Farmers’ Incentives to take Collective Action Via Cooperative: A Transactions Cost Approach.” Cooperative Theory: New Approaches. United States Department of Agriculture, Agricultural Cooperative Service, ACS Report #18. July 1987. . 107 87 pp “The Structural Characteristics of Farmer Cooperatives and their Behavioral Consequences.” Cooperative Theory: New Approaches. United States Department of Agriculture, Agricultural Cooperative Service, ACS Report #18. July 1987. . 60 33 pp Veeman, Michele M. “Social Costs of Supply-Restricting Marketing Boards.” Canadian Journal of Agricultural Economics. vol. 30. March 1982. pp2l-35. Ward, Benjamin. “The Firm in Illyria: Market Syndicalism.” American Economic Review. Sept 1958 48(4). pp 89 566  165  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items