Back to Working PapersAbstract
This paper examines the evolution of the dairy genetics industry from the Herd Bull Era to the Artificial Insemination Era using situation, structure, and performance variables. During the herd bull era, the good in question, dairy semen, had high information cost characteristics. Structural issues of product liability were important. One structural alternative, bull pooling, created new situational characteristics; those associated with marginal cost equals zero and common pool resources. Accordingly, pricing, quality choice, and "who does what" issues are examined.
With the development of the Artificial Insemination Era, the high information costs of the Herd Bull Era were reduced. However, the good took on situational characteristics of increasing returns to scale good. Accordingly, structural parameters such as price differentiation, predatory pricing, demand adulteration, new variety introduction and who chooses quality and their impacts on performance are examined.
Also, included in this paper is a proposed, testable hypothesis to solve a long running debate, whether or not a market or administrative structure performs significantly different in providing optimal genetic variety.
I. Introduction
This paper concerns the evolution of the dairy genetics industry (DGI). It examines the DGI from an institutional economics context using a Situation-Structure-Performance (SSP) framework (Schmid 1987). The paper traces the DGI's evolution from the herd bull era through the bull pooling era to the current artificial insemination (AI) era.
In Section II, the SSP framework is applied to the herd bull era. Interdependencies associated with high information cost (HIC), marginal cost equals zero (MC=0), and common pool goods will be discussed. The institutional economic issues associated with increasing returns will be a primary focus of Section III's AI era. A testable hypothesis will be presented in Section IV attempting to settle a long running debate: whether a "free enterprise" or government regulated DGI will produce optimal genetic variety. Section V concludes this paper with a summary and some considerations for the future.
II. The Herd Bull Era
The good in question is a unit of dairy semen. In the herd bull era, this good was produced by a dairy or group of dairies maintaining one or more herd bulls. The bull(s) would then transfer genetic material to the respective dams.
II-a. The Herd Bull Era -- Situation
Under ideal conditions, it takes a minimum of 2.75 years to start determining the genetic improvement ability of a given bull. The 2.75 years refer to the nine-month gestation period and the minimum two years required to grow a heifer into a productive animal.
Any given offspring sired by a particular bull is subject to wide genetic variability. If judgement about bull merits is based upon a few daughters, the conclusions could be completely incorrect. Accordingly, it takes many offspring and perhaps generations to determine the relative merits of a given bull. During the herd bull era, by the time enough information was available, if collected at all, the bulls would have been already removed from service. Furthermore, there is a myriad of other factors affecting cow performance. A good manager with good facilities can make an animal appear to have wonderful genetics when, in fact, the animal is mediocre. Most bull evaluations were then based upon the animal's pedigree and rules of thumb, such as color, temperament, and/or structure.
Producers looking to procure a herd bull potentially faced a situation of high information cost relative to the genetic potential of the animal. Depending upon the circumstance, this high information cost could impact the buyer or the seller.
For an example of the buyer facing high information cost, assume that a bull in question had all of the correct rule of thumb characteristics. Nevertheless, it and all the sires and dams of its pedigree had poor milk production traits. If the seller of the bull was aware of this, he or she could elect to hide this from the potential buyer. Because the buyer saw an animal with the correct rule of thumb characteristics, the seller could have charged the buyer a premium when its value should have been discounted because of the milk production issue.
Now consider a hypothetical scenario where the buyer was an experienced trader someone who bought and resold bulls and cows for a living. Because of his or her experience in trading, the trader may have developed a unique set of rules of thumb to determine bull quality. If the trader's rules of thumb were better than the typical seller's, the seller would face high information cost. For instance, a trader may see a bull, which the typical seller's rules of thumb indicated as a poor animal, was actually a bull with great potential. If the trader elected to hide this fact from the buyer, the seller could have accepted a price from the trader that was less than its true value.
Another source of high information cost associated with herd bulls is safety. Maintaining a bull is dangerous. The mood of a bull is highly variable, with the variability generally increasing with age. Estimating the irritability of a herd bull is difficult and very costly when estimated wrong.
One type of bull ownership was bull pooling. With bull pooling, a group of dairy producers owned a bull or the rights to be serviced by a particular bull. Under this scenario, the good would exhibit marginal cost equals zero characteristics(MC=0). A dairy could be added to the bull pool without increasing the costs (up to the bull's capacity). This MC=0 good is of the avoidable nature. If they did not like the bull or the structure of the bull pool, a producer could voluntarily elect to not participate in it.
II-b. The Herd Bull Era -- Structure and Resulting Performance
The first type of institutional structure for the herd bull era is market. Under the market structure the issue of product liability is important. The structure had to decide who was responsible for the performance of the bull. In the first hypothetical type of this structure alternative, the seller was deemed responsible for product liability. Under this scenario, the seller had to determine and assure the quality of the bull. The types of penalties for poor bull performance were numerous under this scenario. The seller could have refunded all if not part of the price of the bull. No matter what the penalty was, if the bull's quality turned out to be poor, the seller's reputation was hindered. This was arguably the worst penalty for the bull seller as his or her bull values would be discounted in the future.
To prevent this, the seller collected information about the bull's ancestry. General thumb rules were applied. If the breeder decided to "play-it-safe," he or she only sold bulls with a "good pedigree" and the correct rule of thumb characteristics. In many instances, this was redundant as some classifications of pedigree were based on rule of thumb characteristics. A possible result of this was a decrease in genetic variety and problems associated with inbreeding. This was especially probable in areas where a particular breeder had received a status identity for supplying good bulls, having good structured animals, or superior production.
Another hypothetical type of market structure alternative was to have the buyer responsible for product liability. With this scenario, the buyer would be forced to gather the information required to select a good bull and prove its faults if poor performance was achieved. With this scenario, the buyer would apply rules of thumb, but would more likely do so across bull suppliers to assure proper quality. Accordingly, the chance for inbreeding and lowering genetic variety was potentially less than in the seller liability scenario.
However, collecting this information could have been cost prohibitive or technically impossible. If this information cost was deemed too high, a producer could elect to use their own bulls bred from within their own herds. This would help reduce the information costs associated with determining a bull's quality. At one point in time, it was once thought that a registered breeder could not be a "good" stockman and use bulls purchased from other operations (Herman, 1981). A possible reason for this was that the breeder using bred-within-bulls had better information concerning the genetic merits for a bull. Unfortunately, if careful selection processes were not observed, genetic variety loss and inbreeding, causing a potential decrease in genetic progress, could result.
A third structure was the administrative structure. With this type of structure, a government edict dictated what bulls were used. This is especially apparent with the case of the Dutch Belted breed. This breed was bred to supply Dutch royalty and nobility with their dairy products. In the Netherlands, only the royalty and nobility were allowed to have ownership of the breed (Becker, 1973). Accordingly, selection was conducted with government input. Selection was based on appearance, having black fore and hind quarters with a continuous white band in the middle of the body. This selection structure also caused unique milk characteristics. The milk has a soft curd, a high protein percentage relative to its fat content, and is reputed to be easier to digest when compared to other breeds (Hoffman, 1996). It is unclear whether this was planned or a by-product of the selection process.
Once again, my main concern here is whether or not genetic advancement and the minimum genetic variety needed was promoted. After the Dutch royalty and nobility lost their prosperity, the breed offered little to the commercial dairyman. With this particular administrative structure, the characteristics are like the integration structure associated with specific assets and uncertainty goods. Once the nobility lost its prestige, the value of the specific asset, the Dutch Belted breed, was greatly reduced. The breed eroded in the Netherlands. By 1976, only 2% of the identified Dutch Belted cows were well marked. This has improved since pure Dutch Belted semen was reintroduced in Holland from the United States (Hoffman, 1996).
The erosion of the Dutch Belted has modern day comparisons. With the United States consumer preferences now favoring lower fat dairy products, the resulting administrative change of moving to a component-based pricing system which favors fluid milk yield over fat yield has left some of the "colored" breeds undesirable to modern commercial dairy managers.
A fourth structure to consider with the herd bull era was the bull pooling structure. This structure developed to decrease the maintenance costs associated with the bull. It also decreased the aggregate information cost related to the risk associated with bull safety. For instance, instead of five farmers being exposed to danger at anyone time under individual ownership, the risk exposure to any single producer is greatly reduced if they pool together.
As mentioned earlier, the good assumes MC=0 and common pool characteristics under this structure. Once the quality issues of the bull type were decided through administrative, market, or status/grant structures, how the costs of the bull were to be shared and how work common to the bull was to be conducted had to be decided. For instance, who was responsible for the costs of housing the bull? Should a larger herd, which would have required more sire service, pay more for the bull? Who was responsible for the care of the bull in the event it became ill or injured? If the bull's box stall or manger needed repaired, who among the pool was supposed to do it? For work common to the pooled resource, would shirking occur? The answers to these questions determine the performance of the bull pooling structure. It would be interesting to conduct a historical study of this system. The results could prove pertinent to present day applications with MC=0 and common resource goods.
III. The Artificial Insemination Era
Once again, the good in question is a unit of dairy semen. The process of delivering the semen to a prospective dam is via artificial means. AI is a very ancient technology. AI was first practiced by Arabic horseman in the fourteenth century (Herman, 1981). However, it wasn't until the twentieth century that semen freezing technology developed. This allowed for many females and different generations to be bred by an individual bull. Accordingly, it permitted the DGI to develop a record keeping system on the genetic enhancement and transference ability of bull sires, enabling the information costs associated with bull selection to be greatly reduced.
While not completely solving the problems associated with high information costs, these costs were greatly reduced. Accordingly, high information costs will not be a focus of this section.
III-a. The Artificial Insemination Era -- Situation
When the AI industry was in its infancy, there were numerous AI service groups in the United States and many bulls to choose from. In fact, in 1950 there were 97 groups in the United States. By 1976 there were only twenty such organizations. Today, there are very few firms in the domestic industry. The reduced numbers were the result of mergers to take advantage of efficiencies and the exiting of generally smaller firms. As the industry became characterized by fewer, larger firms and using fewer bulls more often, it decreased its cost structure. Accordingly, the frozen semen good is characterized by an increasing returns to scale situation.
III - b. The Artificial Insemination Era -- Structure
A structure that the AI industry needed to decide upon was the issue of price differentiation and predatory pricing, which contribute to industry instability. In the United States, the infant AI industry formed the National Artificial Breeders Association (NABA). This organization was to create and enforce ethical standards for the industry to follow. Originally, AI services were not supposed to cross territorial borders and steal customers or service technicians from each another.
However, the rewards of the advantages of economies of scale were too great. In 1959, J. Rockefeller Prentice of Artificial Breeder Services declared the rule was illegal in the United States and crossed territorial borders, offering better wages to its competitors' technicians and better prices to their customers (Herman, 1981). This structural change away from coordination to a more competitive market system resulted in a very unstable industry that endured many transitions as firms merged or were forced out of business through price competition.
This does not necessarily mean that the firms who were forced to merge or exit were inefficient at producing AI products and services. For example, it is possible that some of these firms had too confining external capital allotments to finance growth. Without expansion capital, the firm would be unable to take advantage of the decreasing cost structure of the industry. An AI firm caught in this scenario would be forced to search for a merger partner or exit the market as it became increasingly uncompetitive as other firms grew.
For the AI industry to fully capture increasing returns, it was imperative that the producers have homogeneous preferences concerning desired cow traits. It can be shown by the amount of intra and interbreed variety prior to the AI industry's start that producer preferences were not homogeneous. A system was needed enabling dairy producer demand to adulterate so that animals with similar characteristics were desired. Accordingly, it was important to have a structure in place to decide upon the rules of demand adulteration.
In the United States, there are three primary factors contributing to demand adulteration. First, the DHIA Coop provides both production and genetic information. Second, breed associations provide information to producers on the structural and performance characteristics a breed should have. Third, the DGI supplies sire statistical information to semen buyers. Using this information, a dairy producer makes mating decisions based upon the improvements the information says is needed. The information provided by these organizations can be geared, either directly or inferentially, to promote certain animals, types or breeds over others.
The DHIA Coop's and breed organizations' information might be contradictory. For instance, the DHIA Coop information may promote milk and milk component yield efficiency. This matches well with the goals of both the Holstein and Jersey breed organizations. Accordingly, the demand for Holstein and Jersey breeds may increase over such dairy breeds as the Guernsey, whose breed organization has primarily emphasized "show ring traits," unique milk characteristics, milk fat and temperament.
This is analogous to the previously mentioned Dutch Belted scenario. For today's commercial dairy producers, the Guernsey breed is being demanded less and less. As more and more start using Holsteins and Jerseys and join those breed organizations, the AI industry is able to concentrate on those two breeds. This results in a loss of interbreed variety as the unpopular breeds have fewer and fewer and increasingly expensive sire service alternatives.
One problem that the DGI is having with its demand adulteration is that it does so too completely. For years now, the dairy producers have been instructed to not use herd bulls and only use sires with good statistical proof. Unfortunately, this has resulted in producers being wary of participating in young sire programs(Demfle and Grundl). Young sires are those bulls that haven't had enough matings and offspring to have complete statistical records.
This puts the DGI in a catch-22. If they tell producers that it is critical to their genetic prosperity to use only proven bulls, they are unable to get their young sires proven. If they have difficulty getting their young sires proven through producer participation in young sire programs, it takes longer to prove a young sire and fewer sires are given proven status. Once again, variety in choice is potentially lost.
A third structure concerning goods exhibiting increasing returns is the introduction of new products or variety structure. However, what if preferences change or poor performance arises from previous decisions in sire approval? It is well known that to enhance some genetic traits others become suppressed. If a threshold of minimum animals needed to insure genetic variety is breeched, altering the genetic characteristics of a breed could prove difficult. Due to the increasing returns nature of the DGI, this could become a major concern for the DGI, especially with the less popular breeds.
For instance, if the Guernsey industry decided to try to enhance production traits and move away from their traditional show ring type to a more commercially suitable type, it may prove difficult. From an overall perspective, the DGI is path dependent on Holsteins and, to a lesser extent, Jerseys. Therefore, Guernsey sire alternatives are limited. Within the breed itself, the industry may be path dependent on the current show ring type. Therefore, Guernsey sire variety alternatives may also be too limited. The combination of these two facts increases the likelihood that changing body structural traits would prove difficult if possible.
A structure type closely linked to demand adulteration and the introduction of new varieties are policies deciding who chooses quality. This also affects both inter- and intra breed genetic variety. In the previous paragraph, it was mentioned that the Guernsey breed organization emphasizes body structure type. If the majority of Guernsey semen buyers demand genetic traits accordingly, the AI organization, on the grounds of efficiency, will only accept those Guernsey bulls who exhibit them.
However, there may be those individuals who desire a genetic improvement in milk yield efficiency. Because Guernsey bull acceptance onto a sire list is based primarily on type, finding a bull with the genetic potential to improve production may be difficult. If a bull is found that exhibits those characteristics, the cost of the semen will be high to that producer as only that one bull's semen will be available to him or her. The cost to the demander of the more typical Guernsey sire's semen will also increase as a non typical bull will have to replace a typical bull on the sire list, thereby decreasing their economies of scale. Who chooses quality results in a price/variety trade off.
In the DGI, there are two primary types of structure determining the "who chooses quality issue" -- market and administrative. The United States is considered to be a free enterprise (market) system. While there are some administrative components such as the collection and tabulation of the genetic data affecting the industry, I will tentatively accept this definition.
Proponents of market structures for the DGI feel that the "free enterprise" system is the best way to assure that the genetic variety demanded by the marketplace is being achieved at low cost (Herman, 1981). However, others favor a more administrative system where the government plays a major role in dictating the genetic direction of the firm. An example of this occurs in Great Britain where the government actively participates in the industry in policy formulation, and AI cooperatives provide the implementation of the services allowed under that policy. Because of the increasing returns nature of the industry favoring homogeneous characteristics, supporters of this system believe that the DGI needs to have governmental influence to assure optimal genetic variety, which gives its AI industry more flexibility in correcting past genetic improvement mistakes.
Of course, both sides have valid points. Due to the increasing returns nature of the industry, assuring genetic variety means that the semen will be more costly. However, without governmental influence, the DGI is free to take advantage of increasing returns and deliver to producers a less costly good -- albeit a good with little or no variety. In Section IV, I will present a study proposal designed to determine whether the two systems offer significantly different economic results.
Section IV. Testable Hypothesis -- Comparison of DGI Structures
The DGI could prove to be a fascinating study of the interdependencies associated with an increasing returns industry. What is extremely fascinating to this author is that variety effects are compounded due to the very nature of genetics itself. This makes the issue of deciding who chooses quality paramount. Does the market structure, ala the United States structure, or a more administrative structure, ala the Great Britain structure, offers significantly different economic results from their choice decision systems?
In order to evaluate the two systems, it is essential to have a basis of comparison. For this proposed study, I will be using optimal genetic variety. Optimal genetic variety is the amount of genetic diversity needed to produce efficiently relative to the respective market and production economic characteristics of a society.
For instance, the dairy industry in the United States has become highly specialized in milk and milk component production. Beef from the sale of dairy steers is considered a byproduct of little importance. As a result, milk characteristics are stressed in the United States' industry. In Europe, although it may still be considered a byproduct, dairy beef production is considered much more important. In fact, a majority of beef consumption is from dairy steers. Accordingly, the DGI of Europe is concerned with both milk and meat characteristics.
Another factor compounding the optimal genetic variety issue is the nature of genetics. In order to enhance one trait in animals, other traits tend to be suppressed. For instance -- to emphasize milk production -- meat, fat production, protein production, reproduction, and body structure traits are generally suppressed.
Accordingly, to compare the market and administrative systems, I will use a model where the total annualized receipts of dairy producers are determined. These receipts would include: milk, milk fat, milk protein, breeding stock, cull cow, bull calf, and dairy beef production. These receipts would then be compared to annualized costs associated with veterinary expense or premature culling due to genetically induced production or preference problems. Preference problems would include value judgements such as color, color pattern, temperament, etc. . . . Genetically induced production problems would include poor production, reproduction, and body structural characteristics. If a particular institutional structure is providing optimal genetic variety, the difference between the annualized receipts and costs will be larger than a system that is not promoting optimal genetic variety. This model would be valid regardless of that society's reward system for a dairy producer's production.
Unfortunately, this model is historical in perspective. As it stands today, it cannot account for the costs associated with changing the future genetic direction caused by changes in preferences. However, by comparing administrative and market DGI types by this method, we will be able to determine if there is a significant performance difference regarding optimal genetic variety. Furthermore, if there is a difference, we will be able to determine which system is better at delivering the optimal genetic variety.
V. Conclusion and Future Considerations
This paper has examined the DGI's evolution from using herd bulls to the current era of using artificial insemination
from an institutional economics approach. The primary interdependency of the herd bull era was those associated with a
high information cost good. The evolution of the DGI into the artificial insemination era helped alleviate some high
information cost concerns. However, some arguably equally troubling interdependency resulted; those associated with an
increasing returns good. With embryo transfer, embryo sexing and cloning technology -- the increasing returns issues of
the DGI will be compounded as fewer and fewer sires and dams are used to produce offspring.
Bibliography
Becker, R. 1973. Dairy Cattle Breeds Origin and Development. University of Florida Press: Gainsville.
Dempfle, L. and E. Grundl. !988. "Identification of Superior Animals and Their Use in Improvement Programmes." Advances in Animal Breeding. Pudoc Wageningen: Netherlands.
Herman, H. 1981. Improving Cattle by the Millions. University of Missouri Press: Columbia and London.
Hoffman, W. 1996. "Dutch Belted Cattle: History of a Rare and Noble Breed." Breeds of Livestock. Department of Animal Science -- Oklahoma State University.
Schmid, A. 1987. Property, Power, and Public Choice. An Inquiry into Law and Economics. Praeger Publishers: New York.
SSP Framework for the Dairy Genetics Industry
| Good |
Situation |
Structure | Performance |
| Dairy Semen | High Information Cost (Genetic Potential, Inbreeding, Safety) | Market
1. Product Liability on the Seller 2. Product Liability on the Buyer Status/Grant Administrative |
Market
1. Seller absorbs information costs 2. Buyer absorbs information costs Belief structure places information cost on the buyer Information for selection held by administering body. Unclear who pays costs associated with information costs. |
| Dairy Semen (Bull Pooling) | MC=0; Avoidable
Common Resource |
Group Ownership
1) Cost sharing 2) Who is responsible for work for common interest? |
1)Who pays fixed and who
pays variable costs?
2) Shirking? |
| Frozen Dairy Semen | Increasing Returns | 1) Price differentiation and
predatory pricing
2) Rules of Demand Adulteration -- DHIA, Breed Organizations, and DGI 3) Introduction of new variety 4) Who chooses quality? -- Market versus Administrative |
1) Unpopular breeds pay
more per genetic
improvement. AI
Cooperatives underbid
each other to gain demand
an lower cost structure.
Unstable industry.
2) Dairy operations switch to Holstein and Jersey 3) Path dependence. How do you reverse past mistakes? 4) Price/variety trade-off -- optimal genetic variety assured? |