A Laboratory Science Approach to Teaching in the Agricultural Economics Curriculum
Abstract
We argue that economic laboratory experiments can facilitate active learning of theory in the agricultural economics curriculum by placing students in decision-making roles. Three types of experimental protocols are described, along with a discussion of the costs and benefits of experiments. The Double Auction Experiment provides a platform for demonstrating concepts of equilibrium, price discovery, externalities, excess supply, surplus, and speculation. The Monopoly Experiment is a platform for demonstrating concepts of advance production, inventory carryover, monopoly pricing, and search strategies. The Oligopoly Experiment is a platform for demonstrating concepts of thin markets, coordination, and imperfectly competitive equilibria.
Can agricultural economics be taught like a natural science, with experiments in the classroom or laboratory? We say yes. Without even donning lab coats or goggles, experimental economics can motivate students to operationalize theories and test hypotheses just like the laboratories of the natural sciences. Our notion of a laboratory experience in an agricultural economics course is simply to use experimental economics to deconstruct and critically examine the underlying economic theory presented in the course.1
Experimental economics is a subset of pedagogical tools known as active learning or experiential learning. Courses in agricultural economics often introduce other types of active learning exercises in computer laboratories. For example, farm management often has a farm plan that may require instructional time in a computer laboratory. Econometrics has laboratory times scheduled to teach the software and show examples of econometric modeling. Agribusiness management may have a market simulator operated through a computer laboratory. However, these activities are not experimental economics, which is defined as a constructed decision-making experience where students develop data and test hypotheses derived from economic theories.2
In this paper, we explore a number of implications arising from the basic premise that many agricultural economics curricula are structured around microeconomic theory and applied economics, even when these curricula include business-related courses such as agribusiness management, marketing, and finance (Biere 1988, 1999; Heiman et al.; Boland and Akridge). This structure in the agricultural economics curriculum is hardly surprising, considering that much of our academic preparation is grounded in microeconomic theory, and microeconomics is the classical foundation of the agricultural economics profession, so that most of us are thoroughly steeped in it.
Faculty members who teach theory-based agricultural economics courses such as agricultural trade and prices use their training directly in these courses. However, a tension may arise when these same faculty members teach other, less theory-based, courses, especially those in agribusiness. In this tension we may find ourselves asking: Are we economists or are we business professionals, or something else? The answer that most often prevails is that we are (agricultural) economists. Boland and Akridge asserted that “…agribusiness management programs remain based on a core curriculum in agricultural economics…” (p. 564). Thus, as relative latecomers to the pedagogical playing field, some agribusiness courses are compromises between the instructional methods of applied economics and those of other fields such as business. Following a path through those other fields to agribusiness, many instructors use a variety of active learning activities—such as case studies and market simulators—that do not fully exploit our training in economics. However, there is an alternative path that is better integrated with economic theory. We suggest that not only agribusiness but many other agricultural economics courses can have an experimental laboratory component, and this laboratory may be the most effective way to demonstrate the economic theory that is already in our courses.
Active Learning Versus Experimental Economics
The field of agricultural economics is replete with active learning experiences. In their survey of undergraduate agribusiness programs, Boland and Akridge argued that more experiential learning activities are needed such as case studies, group projects, or simulation exercises. Bastian et al. extended this list to include class discussions and debates, problem solving, and interactive lectures. Indeed, agricultural economists have a long history of incorporating active learning in their classes, and have also contributed materially to this pedagogical literature, including papers on general active learning (Bastian et al.; Bonwell; Koontz et al. 1995; Wilson), business and management games (Babb, Oesterle, and Erickson; Curtis; Eidman et al.), decision strategy simulators (Arellano, Hine, and Thilmany; Dahlgran; Koontz et al. 1992), commodity trading exercises (Fackler and McNew; Tierney; Trapp), and role-playing activities (Blank; Park et al.).
To our credit, the agricultural economics profession has widely embraced and promoted active learning. For example, Bastian et al. observed that their active learning exercise encouraged more questions which, in turn, prompted more teaching moments. Bonwell suggested that because lectures make up the largest share of our teaching time, instructors should incorporate active learning into lectures. Referencing pedagogical studies from different fields, he found evidence that active learning can have a significant effect on learning outcomes. Wilson used the example of a group exercise in his managerial economics course to make the case that students are better prepared for the workforce because of the problem-solving and team activities associated with active learning.
Wilson also identified a tension between economics and management instruction that is similar to that experienced in agribusiness courses. The economics component requires solving firm-level problems with specific knowledge—the outcomes are knowable and based on theory. The management component requires knowing a process, which is affected by messy human interactions, and may or may not generate optimal economic outcomes. The active learning that supports the management component likewise often leads to an “explanation and description rather than a prediction” (Wilson p. 289). Consider many of the market simulators used in agribusiness management classes where the interactions are often too complex to yield predictions of core economic phenomena. We submit that the weakness of many active learning tools in theory-based curricula is that they lack a theoretical orientation, or they are specific to a particular context, or they present such complicated scenarios that analysis based on theory becomes intractable.
Bastian et al. asserted that students gain experience in applying concepts through the active learning process, and these experiences are precursors or catalysts to understanding theoretical principles. In contrast, experimental economics may be more efficient in helping students understand theoretical principles because the experiences are the applications of the theories. In fact, Barnett and Kriesel proposed the following criteria to distinguish classroom experiments from other active learning: “(1) the activity is conducted primarily for instructional purposes to demonstrate specific economic concepts, (2) the activity involves participants' making economic decisions within a somewhat controlled environment, and (3) participants' economic decisions are compared to behavioral hypotheses derived from economic theory” (p. 322). Compared to the active and experiential learning approaches mentioned earlier, the laboratory experiments described in this paper have predicted, measurable outcomes grounded in theory—just like experiments in the natural sciences. In short, laboratory experiments offer a bridge between theory and observations (Nelson and Wilson).
Value of the Laboratory
What advantages do laboratory experiences in general provide in an academic curriculum? According to the Committee on Undergraduate Science Education, laboratory experiments in the sciences are useful for a number of reasons that, in our context, particularly include:
- experiments that help students develop an intuition for theoretical concepts presented in lectures,
- experiments that allow students to experience basic phenomena, and
- experiments that help students develop insight and skills essential to the discipline.
For example, to illustrate the reaction of two chemicals and show that the result is consistent with stoichiometric theory, a chemistry instructor will actually generate the reaction in the lecture hall or assign it as a lab experiment. Similarly, to demonstrate the theory of equilibrium price and quantity in a competitive market, an instructor can actually generate the market results in the classroom or laboratory, with students playing the roles of buyers and sellers (e.g., French and Turner). In both cases, experiments help students observe theory in action.
As an illustration of the second reason experiments are useful, when two common chemicals are mixed together and burst into flames, the result is surprising and interesting—a real attention-getter. Similarly, in the classroom, students may be surprised to find themselves generating a speculative bubble in an asset market even when they all agree that the true value of the asset is far below its current market price (e.g., Ball and Holt).
In the third case, laboratory exercises prepare students for professional life in the natural sciences by exercising their imagination and critical thinking in developing and testing hypotheses (Ladd; Garcia and Nelson; Hudson), and by conditioning them to work in controlled environments with standardized quantitative and analytical procedures. Similarly, laboratory experiments can expose students to a variety of controlled economic environments such as those discussed in this paper, including institutions for exchange (such as bilateral bargaining or Dutch auctions), market structures (monopoly or oligopoly), and strategic behavior (monopoly pricing or arbitrage). By integrating experiments with theory, the agricultural economics student may learn not only how to make better decisions, but also how to recognize and even manage the situations in which certain decisions are more or less appropriate.
Experiments for the Agricultural Economics Curriculum
Experiments are an innovative way of putting flesh on the skeleton of economic theory. Students exposed to a variety of market environments learn that markets are dynamic entities that adjust to new information, institutions, and policies. Even though experimental models are abstractions, they help students learn how real markets work.
It is well accepted that students vary in their learning styles (Oltmans; Ellerbrock and Norton; Wetzstein). One of the challenges of any curriculum is to motivate this variety of students by exposing them to a variety of learning experiences. As Kolb observed “We have often emphasized the unitary linear trend of human growth and development at the expense of acknowledging and managing the diverse developmental pathways that exist within different disciplines and professions” (p. 233). Experimental economics is simply one more learning experience for accommodating the different learning styles of students.
While there are many experiments that can be used in the agricultural economics curriculum, we concentrate on three protocols, or “platforms,” that we feel offer the widest range of application to standard lecture material and theory in agricultural economics and agribusiness. These are the double auction, monopoly, and oligopoly platforms. We have used all of these in our classes, and we describe a number of extensions used by other teachers.
The Double Auction Experiment: A Platform for Demonstrating Concepts of Equilibrium, Price Discovery, Price Controls, Externalities, Taxes, Permits, Excess Supply, Surplus, and Speculation
A classic experiment for the agricultural economics classroom is the double auction, so called because it is a market where both buyers and sellers make offers simultaneously (Davis and Holt; Holt). Over the course of the experiment, the students—playing the roles of buyers or sellers—can actually see the market converge to the equilibrium. The double oral auction is a particularly lively market institution because students make trades by shouting out bid and offer prices, and accepting exchanges. In addition, the double auction is an excellent kinesthetic complement to lectures on price theory because it consistently and reliably generates the competitive equilibrium and is relatively insensitive to variations in experimental protocol or control (Davis and Holt).
In the experiment, the instructor divides the class into buyers and sellers. Buyers have decreasing unit valuations, and sellers have increasing unit costs for the homogeneous, usually nondescript, commodity. Each seller is given one or more units that she wants to sell at a price higher than her cost. Each buyer wishes to obtain one or more units at a price lower than her unit valuation. Buyers are most conveniently characterized as wholesalers who can resell any unit they acquire at its unit valuation, which is often described to them as its “resale value.” The instructor serves as the auctioneer, opening the market and posting ascending price bids from buyers and descending price offers from sellers until a deal is made. Either single- or multiple-unit exchanges can be accommodated. The auctioneer usually closes the market when there are no more improving bids or offers.
In our experience, only a few buyers and sellers enter the market at first, and those with some trepidation. But after the first few trades, the market picks up in the volume of trading (and noise), prices converge, bidding slows, and the trading period ends soon after the last tradable unit is exchanged. At this point, students do not know that prices have converged to the competitive equilibrium, only that they have stabilized. Nor do they know when the competitive quantity has been exchanged, only that they can no longer interest any of their classmates in a trade at any affordable price. It might be surprising to ask students how they explain the outcome. Holt made a pointed comment about the influence that conventional teaching methods have had in his ensuing class discussions: “I used to ask students what theory would explain why prices converged to the observed level, but even those with prior knowledge of supply and demand would not make the connection. Students often come up with theories based on averaging all values and costs. This says a lot about the ineffectiveness of the usual authoritarian, smooth-curve presentations of supply and demand” (p. 198).
In the context of an agribusiness marketing course, the main purpose of the experiment is to demonstrate pricing and trading volume in a competitive market, but there are also elements of supply and demand, equilibrium, price discovery and determination, institutions, information, structure, conduct, and performance. We have used the exercise to explore some of the received doctrine about the requirements for perfect competition, particularly the conditions of large numbers of buyers and sellers, and perfect information. It frequently happens that the competitive equilibrium can be obtained without large numbers in classroom markets as thin as three buyers and sellers. Nor can perfect information be essential, inasmuch as each student starts out completely ignorant of the valuations or costs of the other students. Finally, the “invisible hand” can actually be observed in its role as an equilibrating mechanism. In this institution it is manifested as the information contained in the converging bids and offers which, though few result in exchanges, are nonetheless noted and used in making subsequent pricing decisions. An entirely different manifestation of the invisible hand can be revealed by modifying the conditions slightly to impose a price ceiling or floor (Netusil), or by introducing a Walrasian auction market (Mestelman), both of which produce the palpable sensation of excess supply or demand.
The double auction can be adapted to show more than just convergence to equilibrium and the relaxing of conditions for competitive markets. With simple modifications, students can explore the effects of a variety of policy interventions that might be of interest in an agricultural policy course. Netusil described several variations of the double auction, including price floors, price ceilings, quantity restrictions, shifting supply or demand curves (through taxes or subsidies), cartelization, suggested prices, and externalities. On the price ceiling variation, Netusil commented: “Buyers, who were initially elated by the price ceiling, quickly became frustrated by their inability to make a trade; the inefficiency from this restriction is immediately clear to students—fewer trades occur because of the price ceiling and all trades occur at the restricted price” (p. 268). In her cartel variation, attempts to manipulate the market by a group of sellers resulted in retaliatory actions by buyers, and the cartel ultimately dissolved after only a few rounds.
A number of classroom experimenters have extended the double auction platform to consider negative externalities relevant to resource and environmental economics (Crouter; Kilkenny). Bergstrom and Miller introduced a negative externality by penalizing every market participant for the “pollution damage” caused by each trade. They then incorporated three trading sessions that showed how negative externalities lead to inefficient outcomes, how sales taxes can improve efficiency, and how trading pollution permits can be used for regulation.
Another application of the double auction market—this time in agricultural finance—is in the context of financial asset markets (Shrader and Helgeson; Adams and Kluger). Smith, Suchanek, and Williams were among the first researchers to generate speculative bubbles and crashes under experimental conditions. The lesson to be learned from this phenomenon could be lost in lectures because students are inclined to dismiss speculation by others as ignorance or error. But when the students themselves create the bubble and experience the subsequent crash they gain a far more intuitive understanding of why financial markets can appear so capricious. Ball and Holt developed an exercise for the classroom where each asset has a one-sixth probability of becoming valueless in any forthcoming period, as determined by the roll of a six-sided die. This probability makes the asset worth less in the future, which induces discounting. Traders earn money either by holding assets (which pay a $1 dividend at the end of each period and $6 at the end of the exercise), or by buying and selling assets in a double-oral auction conducted each period. When transactions prices for trades are plotted on an overhead transparency as each exchange is made, the sequence of price discovery is displayed for all to see. The surprising feature of this experiment is that although the fundamental value of the asset remains constant at $6 in each period (as demonstrated by a lecture on backward induction given before or after the experiment), trading will frequently generate a speculative bubble that may reach $8 or more in some periods. This behavioral phenomenon may continue to be exhibited even after the asset's constant $6 value is clearly demonstrated, and after the transactions prices are prominently displayed on a transparency, and after students have participated in the exercise to the end several times! A number of explanations for this behavior, such as irrationality, greed, optimism, etc., can be explored in the subsequent “debriefing” session.
The double auction is not the only platform for demonstrating competitive behavior. Other auctions (including the Dutch, English, and sealed-bid), as well as other institutions (such as posted-offer and bilateral negotiation), often perform equally well, although some are better suited to computer labs or multiple periods. The following platform uses a multiperiod exercise to demonstrate the opposite extreme in cooperative behavior: monopoly.
The Monopoly Experiment: A Platform for Demonstrating Concepts of Advance Production, Production-to-Order, Inventory Carryover, Monopoly Pricing, and Search Strategies
The monopoly experiment is a polar opposite of the double auction. Not only does it demonstrate the most imperfectly competitive institution, it does not require in-class participation and can be run as a homework assignment over several weeks (Nelson and Beil 1994). It also serves as a benchmark equilibrium from which to compare other imperfectly competitive markets.
The basic monopoly experiment places individual students in markets as monopolist producers with simulated consumers; that is, the instructor uses a constructed demand curve, and the student sellers are told they will face “The Demand Curve,” rather than live buyers as in the double auction. The students know all about their costs of production, but they do not know their demand schedule initially. The instructor asks each student to select a quantity of product and decide on a single price at which she will offer the entire quantity for sale. Using this information, the instructor simply reads off the demand schedule how much each student sold at the price asked.
When the entire quantity offered is not sold, there are three options for dealing with the excess supply. If production units are characterized as services, then there can be no excess supply, since they are not produced in advance. In contrast, if units are represented as durable goods, then they can be produced in advance, put into inventory, and carried over to the next period. But if units must be produced in advance and are perishable, like many agricultural products, then the full cost of excess production is incurred whether the product is sold or not. This last condition is particularly challenging for the producer since overproduction can be very costly (though valuable as strategic information about demand), while underproduction provides no information on the whereabouts of the demand curve and, consequently, how much more money could have been made if a higher price had been asked.
To be successful, students must somehow determine the demand for their product, but they approach the task in different ways. Some take their market power too much for granted, and initially set their price so high that they sell nothing at all. Others try mark-up pricing based on average cost, or offer the quantity corresponding to the minimum average cost, but soon discover that asking higher prices for larger quantities generates much higher profits. Still others iterate their way to a maximum by trial-and-error processes, including lucky guessing and, occasionally, even the use of marginal principles. In the debriefings, students generally agree that theory can be helpful in getting to the maximum quickly. An important lesson that students take from this experiment is that monopolists are not so much price makers as price searchers, and are constrained by demand and technology just like firms in more competitive markets.
Variations in the monopoly experiment of interest to an agricultural marketing or prices class can include different shapes of the demand curve, shifting or rotating the demand curve, or combining students into cooperatives. The demand curve does not have to be linear and downward-sloping; there are interesting pricing implications when it is step-shaped, discontinuous, unitary elastic, or even upward sloping in places. In later sessions, the demand curve can be shifted in or rotated by invoking changes in consumer tastes and preferences; or it can be shifted out or rotated by providing investment opportunities in research or advertising. Hudson and Lusk described a web-based experiment that extends the monopoly platform into price discrimination, principal-agent models and the economics of information.
A more complex exercise, of interest to an agricultural cooperatives course, allows groups of students to form seller cooperatives wherein the members have to agree on a single price and quantity even though they each have different cost functions. However, the behavioral principles involved in a cooperative are more apparent in the oligopoly experiments that we describe next.
The Oligopoly Experiment: A Platform for Demonstrating Concepts of Thin Markets, Coordination, and Imperfectly Competitive Equilibria (Cournot, Stackelberg, and Bertrand)
In light of concerns about the consolidation of firms in the agribusiness sector, an increasingly important topic for agricultural policy and agribusiness management courses is imperfect competition. The most common mode of inquiry used by our discipline to investigate imperfect competition tends to emphasize theoretical models. However, several experiments are available to explore the behavioral environment of oligopolies without resorting solely to a graphical or mathematical exposition.
The basic structure of the oligopoly experiment is to provide a market with exogenous demand (simulated buyers) for a small number of student sellers—typically two to five—and let them make decisions about how they will supply the market. In the Cournot or quantity-setting version of this experiment, students make their decisions in terms of the number of units they choose to deliver, and the output price is simply calculated for them as a function of the total quantity placed on the market. At the beginning of each round, students submit their individual quantities, and then the instructor totals the output, generates a market-clearing price, and returns this information to each student/firm. After a few rounds, students usually begin to appreciate that a reduction in total supply would increase everyone's profits, but they are usually frustrated in their attempts to influence the actions of the other members of their market, even when these members would also benefit from such coordinated action.
In one of our agribusiness management classes, we conducted an oligopoly experiment where the students explored three different oligopoly institutions: no communication, coordination, and Stackelberg leader/follower conditions (Wilson and Nelson). We operated three markets with three students in each market, and the experiment lasted for about an hour. For the first five baseline rounds, opportunities for coordination were minimal as the students were not allowed to talk and did not know who was in their market. At the end of the fifth round, we identified the participants in each market, and permitted discussions among the participants for the next three rounds. After exploring the Cournot market institution, we switched to a Stackelberg institution in rounds 9, 10, and 11. For this market we randomly selected one student to be the Stackelberg Leader, and allowed him/her to deliver first. The Leader submitted a quantity, which we announced, and then the two Followers each decided what quantities they would deliver.
During the debriefing session, our students readily agreed that coordination without discussion is difficult to attain, and that thin markets are not necessarily collusive. This would not have been so apparent if the students had not had the opportunity to confer. Meister conducted a classroom experiment in which he prohibited his students from communicating explicitly, and found that all tacit efforts to reduce supply through signaling failed to achieve the collusive outcome. Nelson and Beil (1995) allowed talking after the 10th round and found that one market coordinated immediately, while the other never did.
Apart from variations in rules on communicating and delivery sequencing in the quantity-setting exercise, there are other variations that can be explored in the agribusiness marketing or management classroom. We conducted one experiment where the students participated as buyers, rather than sellers, since oligopsony is a market structure that is arguably more representative of current trends in agribusiness. Another version is the Bertrand oligopoly, where firms choose prices instead of quantities. Davis and Holt recommended this price-setting version for providing the “distinct advantage of having a direct analogue in those natural markets where sellers independently post and advertise a price” (p. 37).
In summary, we have attempted here to make a strong case for the role of laboratory science in agricultural economics curricula with courses spanning the continuum from introductory agricultural economics to game theory. Applications can be found for almost any course, including agricultural finance, policy, trade, prices, production, and cooperatives, as well as agribusiness marketing and management, resource and environmental economics, futures and options markets, and even law and labor issues.
Cost and Benefits of Experiments
A number of authors have focused attention on the costs and benefits of using experiments in the classroom. In their compilation of 113 noncomputerized classroom experiments, Brauer and Delemeester developed a bimatrix of costs and benefits to the student and the instructor that is useful in capturing many of the dimensions of effectiveness.
Among costs, most often mentioned is the opportunity cost of the additional time the instructor needs to develop, coordinate, administer, and integrate an experiment into the curriculum. Another potential cost is the possible adverse effect on the instructor's confidence, authority, or reputation when an experiment does not produce the expected result, or the classroom environment seems less organized or structured (King; Yandell).
Costs to the student may include reduced achievement among students with certain learning styles, personalities, aptitudes, or motivations. In a study of undergraduates randomly assigned to either a lecture-discussion section or a simulation-gaming section of introductory economics, Fraas found that students with low precourse knowledge, no previous economic training, or low SAT scores performed better on the exam after exposure to the gaming treatment than the lecture. In contrast, Dickie found that while experiments improved learning of microeconomics principles among the majority of students, they did not benefit students whose GPAs were below 2.04 or whose ACT scores were below 16.6 (both about one standard deviation below the mean).
Among benefits, the vast majority of anecdotal evidence is that experiments are extraordinarily popular and successful in the classroom, and to our knowledge no authors have reported dissatisfaction with the results of their efforts to introduce the method in their classes. Empirical tests have consistently shown that students exposed to experiments on average do at least as well as those exposed to lectures and often do better (Brauer and Delemeester). Several other student benefits are frequently mentioned, such as more fun in class, more engagement, better attendance, greater attention, and higher effort. However, inasmuch as these are inputs in the function that produces higher exam scores, including them separately on the benefit side of the ledger would seem to be double-counting. On the other hand, Becker notes that students who have saved time through more efficient learning in one subject may invest that time in other subjects, in part-time jobs or in leisure—all contributing to a general increase in student welfare.
Brauer and Delemeester remarked that a frequently ignored quadrant in their bimatrix is the benefit to the instructor. This can take many forms, including: pleasure in a more genial relationship with students (King); better student evaluations (Yandell); pedagogical efficiency by laying empirical foundations for understanding more elaborate concepts (Nelson and Beil 1994) or by substituting a single experiment for a number of lecture examples (Haupert); sense of accomplishment in providing a connection between theories and key features of markets and institutions (Smith); inspiration for generating and exploring research topics (Noussair and Walker; Becker and Kennedy); and reputation for projecting the image of a scientist (Barnett and Kriesel).
Conclusions
Experiments provide students with opportunities to interact and learn in simple market situations that appeal to a kinesthetic learning style, and offer a more integrative learning experience than lecture presentations of theory alone. Through subsequent laboratory reports and class discussions, students are challenged to explain and defend their choices and strategies, and to learn from their errors and the experiences of others. Moreover, the advantage that experiments have theoretical underpinnings may help reduce the angst of teaching microeconomic theory in agricultural economics courses. A variety of dynamic settings can be used to demonstrate or test the theory, and the flexibility of experiments allows the instructor to modify the basic platform so that different topics can be explored. Thus, experimental economics offers the agricultural economics curriculum an adaptable, kinesthetic teaching technique that challenges students to think critically in a stimulating decision-making environment. Whether experiments should take any time away from lectures is—and probably should remain—a matter of instructor preference. However, experiments may provide a way to allow students to try out different theories in ways that lectures and other active learning cannot. As Kenneth Arrow and the 4-H slogan avow, students “learn by doing.”
