Subsequent state and federal legislation has allowed for the formation of related types of organizations. MMOs were designed to mitigate the “disruption of the orderly exchange of commodities” and protect the “value of agricultural assets which support the national credit structure.” . The continued presence in today’s U.S. Code of these justifications and others first voiced in the 1930s suggests that MMOs are still considered a valuable tool for meeting these objectives. However, the structure of agricultural production has changed considerably since the 1930s, suggesting that the value of MMOs may have changed as well. When MMOs were introduced during the Great Depression, farms in the United States were relatively homogeneous compared to today. In 1934, 1.3% of farms were larger than 1000 acres and represented only 6.7% of harvested cropland. In contrast, in 2012, 8.2% of farms were larger than 1000 acres, representing 64.2% of harvested cropland. In 1934, 39.5% of farms were smaller than 50 acres, and represented 11.2% of harvested cropland. In 2012, 38.5% of farms were smaller than 50 acres, representing 1.40% of harvested cropland . In the 1930s, there were fewer large farms, and large farms represented a considerably lower percentage of the nation’s harvested cropland. Thus, promoting the orderly exchange of commodities meant managing the interests of smaller farmers with negligible or nonexistent potential to affect the market individually.
In contrast, there is considerable heterogeneity in farm size and revenues today. Of course, cost— which is an important factor in determining farm size and revenue—is only one of many dimensions of heterogeneity among producers. Producers may also be heterogeneous in terms of operator age, crop mix, land quality,black flower bucket or non-farm income sources. Nevertheless, for simplicity we limit our discussion of heterogeneity to farm sizes, which has changed considerably over time. The implication of this increase in heterogeneity is that it could cause MMOs to yield an unequal distribution of benefits across farm sizes, even proving costly for some producers. In turn, these differences could cause producers in the same industry to disagree about the need for MMOs. This circumstance may explain why growers in industries covered by MMOs have been suing the U.S. government over issues related to their mandatory nature during the last several decades . These differences of opinion suggest a need for MMOs to have a mechanism to adapt their functions and assessment rates over time… and they do. Producer referenda play an important role in determining whether organizations continue and in some cases whether major amendments that change the focus of activities conducted by the MMO should be passed. But, when disagreement occurs among producers because an MMO benefits some producers and not others, the scope for change using these referenda can be limited due to the requirements of the voting rule. Voting rules are particular to each marketing order and are developed in collaboration between industry and the USDA or equivalent state-level department.Voting rules may require a single, double or triple majority. An example of a single majority rule would be that the MMO would be formed if more than half of the producers voting favored formation. An example of a double majority rule would be the requirement that more than half of the producers voting and producers of more than half of the total output of those voting favor formation.
Finally, an example of a triple majority rule would be the requirement that more than half of the producers voting and producers of more than half of the total output of those voting favor formation, with a quorum requirement of 50%.In this paper we consider the most common formation and termination rules for federal marketing orders. Formation of a federal marketing order requires approval by at least two thirds of those producers voting by number or producers of at least two thirds of the total output of those voting. Termination requires approval by more than half of those producers voting by number who must also produce more than half of the total production of those voting. These rules do not include a quorum requirement, although the Secretary of Agriculture has taken the participation rate into account on certain occasions. In our model we assume 100% participation. Given these rules, if an industry comprised of homogenous producers desires to form or terminate a MMO, industry members have the ability to make those changes. This may not be the case in industries with heterogeneous firms where not all industry members are aligned in their profit maximizing choices. In such cases, the question remains of whether heterogeneity among firms may lead to disproportionate voting power among firms, and how this affects an industry’s ability to adapt and change its MMO to meet its needs. This question is the central topic of our paper. This work relates to veins of literature in agricultural economics and political science. Several recent studies have examined the distributional implications of generic commodity promotion programs, and there is a rich literature in coalition formation among agricultural producers. In the political science literature, the work on voting power indices stretches back to the 1940s. Although this literature has a large theoretical component, there is a spate of more recent applied work.
Much of this work is focused on recent changes to member state representation on the Council of the European Union.Like our analysis, this work focuses on the importance of firms, in the role of the voter. Within the extensive literature regarding generic commodity promotion programs, three recent studies consider firm-level heterogeneity and its effects on the benefits firms obtain from MMOs. Chung and Kaiser showed that producers with a less elastic supply response capture more benefits per dollar of generic advertising expenditure than producers with a more elastic supply response. Building on this analysis, Chung and Kaiser suggest that producers with steeper marginal cost curves and fewer fixed factors have a lower marginal benefit per unit of generic advertising expenditure than producers with flatter marginal cost curves and more fixed factors. Crespi and Marette consider firm heterogeneity in output quality. They find that when a crop covered by a marketing order is horizontally differentiated,square black flower bucket the benefits of a uniform assessment are not equitably distributed between the producers of the two different varieties. The newest papers in this vein consider the effects of market power. Crespi and Marette model an agricultural market as a monopoly with a potential entrant and consider a demand-enhancing check-off program. They find that generic promotion can be pro-competitive due to cost-sharingby firms and consumers’ increased exposure to the product. This paper shows that relaxing the assumption of price-taking firms can alter conclusions regarding the effects of MMOs, a theme we continue. Our work most closely relates to that of Zheng, Bar and Kaiser . They consider how benefits of generic advertising vary with farm size assuming an asymmetric oligopoly. In their work, they vary both the convexity of the demand curve and the type of demand shift and demonstrate that there are cases in which firms differentially benefit from generic advertising. Our work, like theirs, considers the implications of cost heterogeneity among farms and the implications of market power. However, we focus on the distribution of voting power. In addition, our work differs from theirs in other important ways. First, we consider heterogeneity in terms of costs and multiple market structure scenarios; they only consider cost heterogeneity under asymmetric oligopoly. Second, they only consider the distribution of the associated benefits of generic promotion compared to a counterfactual of no promotion and do not consider the referenda that lead to these generic promotion programs. Our work builds on theirs by considering how these differential benefits interact with the institutional framework of the MMO. In addition to this work on generic promotion, there is a literature in agricultural economics related to coalition formation in the context of agricultural cooperatives stemming from Staatz and Sexton that draws on the methodology of cooperative game theory. While there are some strong parallels to this work, the nature of MMOs does not lend itself to this framework because of its disaggregated nature.
Although we consider some strategic interactions among firms in the market, in the formation of MMOs, referenda are mandated by law and administered by the government, and therefore the type of producer interaction modeled in cooperative game theory is not necessary for an MMO to exist.4 The sharing rule of the benefits stemming from the formation of the MMO is simply that each producer keeps his own profits. In addition, we must make an important distinction between the mandatory nature of MMOs and the voluntary nature of other types of organizations such as cartels, cooperatives, and strategic alliances. These organizations may form among firms that benefit from them within a larger industry. Thus, a dominant producer choosing not to participate is not necessarily a death knell for an agricultural cooperative. However, the vote of this producer in a MMO referendum could be vital. To measure voting power, we use a metric developed by Banzhaf and also attributed to Penrose which measures the ability of each voter to cast a swing vote. Straffin characterized the Banzhaf Power measure and another well-known and oft-used voting power measure, the Shapley-Shubik Index, according to their assumptions about probabilities. Our Feasible Banzhaf Power Index follows Straffin as well as Laruelle and Valenciano , who developed a unified approach to the Banzhaf Power Index and related voting power indices by providing a more general characterization of the probabilities underlying these measures. The voting power literature is expansive, and a number of academics have proposed new or related voting power indices throughout the 20th century. In addition to theoretical work regarding voting power, there is considerable work applying these concepts to specific empirical contexts. Voting power indices have been utilized recently to examine the representation of nations on the Council of the European Union. Multiple majority rules for the Council were established by the Treaty of Nice and later amended in the Treaty of Lisbon, and political scientists have considered the implications of various voting rules implemented or under consideration by the Council at various points in time . Our work is similar to theirs in that it deals with questions of changing relative size and its effect on voting power, as well as what the optimal voting rules would be depending on the goal of policymakers. While much of the applied focus has been on political representation in the EU, there have also been applications to other settings. Leech examines the validity of voting power measure among the shareholders of large British companies by calculating voting power based on the share of votes held by the largest shareholders and comparing the result to expert evidence regarding the level of control these shareholders exert on the companies they own shares of. This work is similar to our own in that it considers a context in which the goal of voters is likely to be profit-maximization. As Leech points out, in a vote that would yield an unambiguous increase in share value for all shareholders, under the assumption of profit-maximization, we would expect all shareholders to vote for the measure, making the context fairly uninteresting. However, this information can also be incredibly useful, as we show in this paper. Finally, in addition to these applications, some authors have developed empirical voting power indices that incorporate information on voter preferences . In these empirical approaches, the authors propose rules that limit the number of winning coalitions by locating voting bodies relative to each other in a uni-dimensional policy space. Locations are defined by observed voting behavior. While incorporating information on voter preferences is appealing, determining voting bodies’ relative positioning in policy space is mathematically intensive, and the resulting ordering does not provide an intuitive link to policy positions. We avoid some of these challenges by incorporating the accepted neoclassical theory of the profit-maximizing firm into our development of our Feasible Banzhaf Power Index.5 In Table 1 we report the pre-vote market shares and Banzhaf Power Index and Feasible Banzhaf Power Index values associated with formation for the low cost firm and average high cost firm.