Paul Herbig's Working Papers

Return to Home Page

1997

INTERFIRM RELATIONS: COMPETE OR COOPERATE
AN EXPLORATORY STUDY

By Paul Herbig

Abstract of INTERFIRM RELATIONS: COMPETE OR COOPERATE
AN EXPLORATORY STUDY

This paper extends Deutsch's (1949) Theory of Cooperation and Competition from its original level--that of individuals within an organization --to a macrolevel viewpoint-- within an industry and the effects upon interfirm relations. A game theoretic experiment is conducted whereupon teams are either rewarded based upon total industry performance or compared to others within their industry. The hypothesis that group rewards would instill cooperation while individual rewards would create competition was supported. Knowing or not knowing one's rivals does not make a difference upon a team's eventual behavior. It is speculated that Cartels can survive long term only under cooperative conditIons. Implications and Areas for future research are given.

INTRODUCTION

Deutsch's (1949) classic theory of cooperation and competition states that the perceived interdependence of goals significantly affects the dynamics and outcomes of social interactions. In positive interdependence (cooperation) persons perceive that their goals are positively related; one's movements towards one's goals facilitates others' goals. In negative interdependence (competition) persons believe their goals are negatively correlated; one's goal movement interferes with and makes it less likely that others will reach their goals (Tjosvold 1986). Deutsch theorized that perceived goal interdependence affects both the patterns and consequences of interaction. In cooperative pursuits, persons encourage each other, expect help and assistance from each other, and trust is promoted and Productivity is enhanced. Contrarily, under competition persons are tempted to mislead and interfere with others so as to reach one's own goals. They are reluctant to assist others or to ask others for assistance. These interactions arising from negative interdependence result in frustration, hostility, and low productivity.
Researchers are generally in agreement with Deutsch's theories on interaction effects of competition and cooperation. But Deutsch's theory concerns the behavior of individuals within an organization. Are his concepts and their consequences transferable to interfirm activities at the industry or global level?

Cooperation versus Competition
It is a modern fad of management to compare and contrast the competition based American economic system to that of the Japanese economic system with its highly popularized cooperative arrangements between firms. The secret of the success of the Japanese economy (so goes the logic behind the fad) is their reliance on cooperation between firms within an industry. The moral of the story is if American companies are to survive in this global economy they must make more efforts to cooperate with their competition and not engage in sometimes deadly competitive feuds. The cry is in the air, "Cooperate. Cooperate."
Cooperation does brings out the best in us. Research studies indicate clearly that Cooperation is a more productive state than competitive behavior(confirming Deutsch's theory). Yet for all their fabled cooperativeness, the Japanese are still fierce competitiors-- the intense competition by exam ("hell week") to be admitted into an elite university has no comparison in America. But competition is externalized against the outgroup not exerted upon one's own group, family, company, or nation. (Rauschenbach 1988).
The Japanese social structure minimizes competitive behavior so as to maintain the social status quo, one example of which is the firm policy of paying all employees with the same start date the same pay. In business, lifetime employment and the seniority system is the basic organizational principle. The Japanese worker does compete with others within the organization throughout his working life but in a covert fashion. But he still considers himself part of the organization, part of the group, and will subordinate some of his personal goals to that of the group; thus workers often consider total profits before individual profits, creating a spirit of cooperation. (Yoshida 1985) (Alston 1985)
In America, the economic system and its culture encourages competition. The American spirit of rugged individualism espouses a competitive approach to life which has been hailed by both media and folk hero example alike(Who hasn't dreamed of emulating Horatio Alger or Steve Jobs-The Entrepreneur who against all odds succeeds and becomes rich and famous, "Go Out and Do it on your own"). It is also certainly true that nothing succeeds like competition in weeding out the incompetent and unworkable. Competition is the best system for Darwinism: survival of the fittest for what the market (customers) want or need. But Competition has adverse effects: low productivity, oftentimes resulting in a less than optimum behavior, and high social costs. American Business strategies are based upon the military perspective on competition; The business environment is seen to be made up of threats and opportunities, with competitive strategies determining success or failure. But more to it, the American style of competition has not been as successful as it was in previous eras. Perhaps a paradigm change of interfirm relations is in order to meet the new elements of modern times--the international marketplace and global competition. Cooperative strategies where business can share resources, jointly bear the costs and risks of entering new markets and resist foreign competition may now be more appropriate and allow survival of the company(Ulrich 1985).
Interfirm cooperative activities are becoming increasingly popular as instruments of strategic action (note the semiconductor consortium in Texas). However, understanding of cooperative action in a competitive economy needs further improvement if the hoped for benefits of interfirm collaboration are to be realized (Murphy 1988). The political system and the courts have attempted to deal with the interfirm cooperative issues through antitrust laws. The courts have recognized some interfirm cooperation need not necessarily harm competition but actually may enhance it. The courts/ legislature have had difficulties in separating the legitimate benevolent effects of joint ventures from the illegal malevolent effects.
But no matter how much the hue and cry and the overwhelming evidence of the superiority of cooperation, when the American economy is based upon individual firm performance and corporate rivalry rather than cooperation among firms , then competition will be the dominant economic form of behavior. When the reward structure, culture, and especially the disincentives provide further impetus for competitive behavior, competition is what we will get. CoOperation may be preached but the reward structure dictates what we will get (competition in America's case).
This paper tests the concept that Deutsch's Theory of Cooperation and Competition is applicable to interfirm relations, that it is the reward structure that determines behavior; that we can cry cooperate until our face is blue but that behavior will not change until the incentive structure (and just as importantly the disincentive structure) is changed. This exploratory study attempts to study the behavior of teams under different reward conditions and determine whether or not competitive or cooperative behavior will result as a direct result of the type of rewards provided to the firms.

HYPOTHESES

Competition or Competitive Behavior is defined as that behavior which results in a competitive equilibrium when activity stabilizes. That is when all firms or individuals in a competitive situation are actively seeking the best outcome for themselves, an equilibrium will be reached which can be predicted apriori via game theory. Cooperation or Cooperative Behavior is defined to occur when the optimum for the group as a whole is obtained when activity stabilizes, regardless of whether the individual firms can make themselves better off; When firms, entities, or teams will sacrifice potential individual gains for the good of the entire group. Stabilization is said to occur when no change in any firm's decisions is observed for three consecutive periods.
Those Companies (Teams) who are rewarded based upon total industry profits (summation of the profits of the three teams within an industry) should stabilize at the (high,high,high) point which is not a competitive equilibrium position but the cooperative equilibrium point. Because they are rewarded based upon the group results and not their own individual entity's performance comparative to all other groups, more cooperative behavior should be seen with efforts to reach the group maximization point.

Hypothesis 1) Companies (Teams) rewarded based upon group reward systems will stabilize at the cooperative equilibrium point, that point of maximum group profitability.

On the other hand, those Firms whose reward system is based upon their individual firm results and whose rewards are compared to all other firms within their industry will actively compete. Even though there exists a group maximum, the act of competition and the rewards to look out for their own (individual firm or entity) best results will drive the resulting industry (and the firms within it) into a competitive equilibrium which, while perhaps good for an individual firm will not be the optimal point for the industry as a whole. This process may take some time but the eventual equilibria should be at a competitive equilibria when activities stabilize.

Hypothesis 2) Firms rewarded based upon individual firm results should stabilize at one of the four competitive equilibrium points.

Knowing or not knowing one's competitors should not effect behavior. In the real world, some of the most fiercely competitive behavior is shown by firms intimately knowledgeable with their competitors personnel, behavior, and strategy. Knowledge of one's competition should therefore have little impact on one's ultimate competitive strategy and whether one knows or not knows, should be overshadowed by the reward system and the incentives upon behavior therein. After sufficient time, the knowledge of one's competitors should not influence one's behavior; it is strictly the reward structure alone that influences the competitive behavior and performance of one's firm.

Hypothesis 3) Knowing one's competitors or not knowing who they are will have no effect on the final competitive or cooperative behavior of the firm.

Cartel formation and collusion have long been studied as per their formation, conduct, and eventual disintegration in the competitive environment. Most economists have indicated (and real world examples such as OPEC seem to confirm) that competitive forces and pressures to seek maximum individual gain rather than group optima will eventually break up a cartel or collusion. However, in a cooperative environment, when the rewards are given according to group performance, the motivation to defect to better one's own performance should be negated. The reward structure is skewed towards one's group and not one's individual efforts. If the group reward structure can be strictly adhered to (through governmental fiat or high peer pressure) the tremendous temptation to defect and better one's individual performance will be withheld. OPEC did not (and still does not) have a sufficiently prohibitive reward or punishment system to inflict the necessary group reward process necessary to set and maintain its desired price and to deter the individual members from defecting and gaining individual benefit at the cost of the group. Although cartel formation is speculated to occur under both competitive and cooperative environments, it is supposed that only at a cooperative equilibrium could a cartel survive long term. Group reward systems enhance collusion and necessarily support it as a part of the co-operative efforts. Cartels so formed under cooperative systems will tend to be stable when the reward and punishment system is sufficient to deter individual profit maximzation behavior.

Hypothesis 4) A stable and long term cartel will be created only in a group reward setting resulting in cooperative behavior.

The flip side of collusion is that it should be unstable in purely competitive environments, when individual rewards are more important than maximizing group performance. Therefore we should see some initial cartel formation efforts but eventually the temptation to defect is too much and the cartel will break down. The pressures of the competitive environment is not inducive to cartel formation.

Hypothesis 5) Collusion in pure competitive situations is an unstable situation and will inevitably result in a competitive equilibrium when activity stabilizes.

METHODOLOGY

A game theoretic experiment was conducted. One of the strengths of game theory is the precision and completeness with which it describes and analyzes multiperson decision problems. The rules of the game are completely specified, the strategy set is known to all players, an equilibrium predicts not only what is expected to happen but also the anticipated consequences of any deviation from the optimal strategies of the players. The control inherent in a game makes its use in a laboratory experiment worthwhile.
Eighteen student teams comprising approximately fifty undergraduate marketing students from a prestigious Midwestern University were utilized. The use of student subjects as a convenience sample was a major consideration in the research design. Students were arbitrarily assigned to teams of three students each and treatments were randomly assigned to teams. Teams were used to emulate the multiple influences found in the industrial decision making process. Eighteen game turns were conducted.
The experiment concerned airline industry pricing on a single route. The airline industry is a particularly good choice because pricing is highly dynamic in nature (daily or even hourly rate changes can easily be made in response to competitor actions), competitive response can be quick, competition is well defined and have bordered product lines (routes) which are different and distinguishable from other products (preferred for ease of measurement purposes), competitors are well known with usually no more than 3 or 4 competitors on any route (a manageable number to simulate), a history on any one competitor is well established (a known track record), and the outcomes(profits) are typically not perfectly known (uncertainty exists on the profit profiles for each competitor).
The teams were randomly divided into industry groups of three who then competed against one another according to reward structure and knowledge of competition. The six industry groups of three were further divided by a random process into the proper treatment type. Half of the teams were told they would be rewarded based upon their own individual firm's cumulative profits compared to all other teams at the end of the game. This was the individual treatment. The other half of the teams were told they would be rewarded (graded) based upon the sum of their particular industry group. This was the group treatment.
The second treatment was knowledge of the competition. The teams were again randomly divided into halves. Half were given the group number and the name of the members of those teams within their own industry group; this was the knowledge treatment. The other half knew their competition only as competitor one and competitor two and could communicate only through comments (marketing signals) submitted on the same game sheet as their choice for that period; this was the no-knowledge treatment. Therefore four combinations of behavior were provided. The payoff matrix was a 2 by 2 by 2 owing to having three competitors and two possible pricing choices for each competitor or 8 possible options in all.
A Competitive Equilibrium exists when there is no benefit for any competitor to change his position. Let us examine the competitive situation if all members were to play high (high,high,high). They would all earn a profit of $400. The group profit level of $1200 is at a maximum. If each company is rewarded based upon their individual performance relative to other firms in its industry, the key question is can any firm do better? Any one of the firms can increase its profits by changing from high to low pricing and thus earn $500--a gain of $100. If the other two players remain at high the situation now is (low,high, high) and the two high players each earn $150. This represents a loss of $250 from their prior level. Can any one firm do any better from this position? Player 1 is at the max so he can't; if either high player moves to low he will reduce his gain from $150 to $100. So at this point no movement will occur as none of the players can gain by changing their positon. This is a competitive equilibrium. As can be readily seen, any combination of one low and two highs result in a competitive equilibrium: (low,high,high); (high, low, high); (high, high, low).
Let us examine the consequences when two players simultaneously defect from (high,high,high): a situation that often occurs when several firms have identical notions of gaining by cheating(defecting) on other firms within its industry. This presents a (low,low,high) scenario. Players one and two each earn $100 while player three earns $0. Can any player do better? Player one or two can gain by going high and earn $150. Player three has highest motivation to switch to low and earn $50 ilnstead of $0. This situation is highly unstable. If players one or two switch to high (high,low,high) occurs; as seen before this is a competitive equilibrium and further activity ceases. If player three changes, as is usually the case, all three players are now playing low (low,low,low). Can any player change and increase his profit? No. Any move by any player results in a loss. So (low,low,low) is also a competitive equilbrium.
Therefore four competitive equilibriums exist at: (Low,High,High), (High,Low,High), (High,High, Low) and (Low, Low, Low). Note that all teams playing high is not an equilibrium as the incentive exists for a single team to defect (play low). Nonetheless, this point (all playing high) is the best scenario for the industry as a whole. The industry total is greatest if all choose to play high (1200 versus 800 for next highest combination). This point of group maximization is called the Cooperative equilibrium. If Cooperation were seen and the industry members did indeed choose for all to play high and receive the group optimum then the sheer fact that it is not a competitive equilibrium should be construed as evidence that the cooperation was enhanced by the group reward structure.
At the beginning of play, the experimenter described the conduct of play, the use of the payoff matrix, and answered any questions the playing teams had. Sufficient incentives for the playing team with the highest cumulative profitability at the end of the game were given to provide incentives to compete (it was surmised that being exempt from the final was a powerful enough incentive). The reward structure was expanded to provide part of the class grade as a reward for participation in this study. There was no noticeable learning curve.
The suspected course of play is for those teams who do not know their competition is to feel out their competitions' intentions by signals as well as initially playing high. If a firm were to get burned, being the only industry member to play high, then it is suspected that the firm will play low indefinitely. Thusly one of the four equilibriums should be reached quickly under the individual reward system. However, under the cooperative condition either through signaling or direct collusive effort, the industry members should agree and follow through on playing the high price and earning the maximum profits.
At the half way point, a shuffling of treatments occurred. Those teams who had group reward and knowledge of the competition were told that they were now being graded on individual results and had new competition whom they did not know. Likewise reversal of all treatments for all groups occurred. Although, the industry groups of three did not in reality change, the subjects were told that they had to present the notion of starting from scratch. This midstream change was done to provide both balance for the experiment (every team would have equal time under all treatment conditions) and also to provide motivation for the subjects and equal likelihood of being the team with highest profitability.
To eliminate any end-game actions, the last turn was not declared until after the surveys were turned in. After the full eighteen turns, the game administrator debriefed the playing teams, surveyed them about the exercise via a written questionnaire, and interviewed selected samples on whether they had surmised the intent of the experiment and elicited suggestions on improving future experiments. Results were tabulated, awards given to the winning teams, and a presentation of the findings were given to the participants.
Although the different treatment levels were quite visible, few participants guessed the purpose of the experiment. This was not a concern as even if the experiment's objective were presented on day one, the motivational factors and the team's self interests would have driven the groups to identical results. Since an individual team could not control their own destiny, their resulting behavior and performance was mandated not by any drive to please or confuse the experimenter but purely by self interest.
If finite, the game must also be sufficiently long and its expiration uncertain to prohibit end game strategies from occurring. An end game strategy tends to be played if the game is finite, the duration of the game known and a small number of periods remaining to be played in the game. With eighteen turns, it is felt sufficient time was given to allow teams and industries to stabilize. Thusly, we feel confident that no end game strategies were played.

Results

To analyze the results one could use standard statistical tests (T-tests, ANOVA), that is if the industry was under the group reward treatment then one could proceed to calculate the mean and test to determine if it were significantly different from $400 (firm) or $1200 (industry). But these techniques would only provide deceiving information. The short timeframes (9 periods) mean cartels could exist even under competitive pressures and dissolve early (period 4 or 5), end up at a competitive equilibrium stabilized yet statistically impact the results if data from all the periods were used. If the game was of long enough duration, say 20 to 25 turns with the same treatment, we would expect to see the same final behavior but with the initial cartel effect considerably dampened by the later competitive efforts. OPEC if viewed only from 1974-1981 was a major success and profitable cartel. However, if viewed over the longer range 1930-1990 the highly competitive aspects of the oil industry becomes apparent and the 8 years of OPEC would be absorbed by the long term competitive trends. Therefore standard statistical tests using data over the entire course of the game are inappropriate. A more accurate statistic is a review of the outcomes after activity had stabilized. Table 4 shows the stability analysis. A ChiSquare value of 5.586 indicates significance of .019 and a clear difference between reward treatments. One observes eleven out of the twelve industries settled at the equilibria predicted.
Hypothesis 1 states cooperative behavior should result from being rewarded upon group performance. An analysis of Table 6 indicates group behavior resulted in cooperative behavior in 5 out of the 6 cases. A chi square of 5.586 and a probability of less than .02 results. One industry out of 6 treatments became involved in a price war and failed to reach the cooperative equilibrium. The small sample size would have greatly amplified the effect of one team upon the standard statistics. Since we did define cooperative behavior as reaching a cooperative equilibrium and 5 out of 6 industries did so, we cautiously declare hypothesis 1 supported conditionally until replication with larger sample sizes can be accomplished.

TABLE 1: STABILITY ANALYSIS

Co-Op equilibrium Competitive Equilibria
GroupReward 5 1
Individuals 0 6

Know 3 3
Not know 2 4

We test hypothesis 2 in the same fashion using information from Table 1 to determine final stability points. Six out of 6 industries rewarded individually stabilized at a competitive equilibrium. This was obviously highly significant with probability of less than .001 . We can confidently support the second hypothesis.
To test hypothesis 3, the proposition that knowledge of one's competitors or non-knowledge should have no effect upon one's final behavior and equilibrium state, Table 4 once again is used. Reviewing Table 1, almost even balance is seen from the know versus no know equilibrium: a chisquare indicates no significance; hence, knowledge of one's competitors had no impact on the final outcome. T-tests and ANOVA revealed significant differences between knowledge and no-knowledge, Two Way ANOVA results indicate knowledge treatment was a significant factor in predicting both team and industry profit and the interaction between the treatments was significant (indicating that group knowledge treatment was far superior). We can account for this disreptancy as cartel formation effects. A cartel can be predicted to fail but it might take a considerable timeframe before it begins to unravel. The end result may be a competitive equilibrium but a statistical review of the timeframe would indicate above average profits. The OPEC example also would serve us well here. We must in summary indicate that hypothesis 3 is confirmed pending replication of the experiment over a longer term timeframe to smooth out any initial cartel formation effects.
To test hypothesis 4 and 5 we only grade final behavior. Table 1 provides the evidence to judge these two hypothesis. Cartel behavior was exhibited by both groups in the knowledge condition. Its formation and maintenance was confirmed in 5 out of 6 instances for the group reward treatment. All three group knowledge treatments resulted in a stable cartel. Although nine turns do not make forever, suspicion is that if the game was extended considerably with the reward structure, the results would be the same. We believe This provides sufficient evidence to support hypothesis 4.
Hypothesis 5 indicates that cartel formation by industries rewarded by individual performance will not be successful and will eventually result in a competitive equilibrium. Six out of 6 treatments resulted in competitive equilibriums. Of the three industries given the individual knowledge treatment, two quickly (by turn 2) converged on a competitive equilibrium. The third industry maintained its overt cartel for the first five periods before falling out the sixth period. Interviews with subjects involved in the third industry indicated that at first price setting was agreed to by all teams within the industry. However, this was done with the intent of "setting up" the opposition for the kill to take advantage of the agreement at a future time. The most surprising element is that all three teams within that third industry entered the cartel arrangement with the intent of defecting at some later time. IT was only a matter of time until one or more members broke ranks and defected, resulting in a competitive equilibrium. It was who would be the first to defect that mattered.
Of the five hypothesis, two are supported outright (numbers 4 and 5, the two cartel propositions) and the other three are conditionally supported. Statistical analysis and behavior analysis provided conflicting and ambiguous results for the first three hypothesis. This is due to small sample size and the short duration of the game played. Replication on a larger sample size and over a longer duration would provide more confidence in the results.

Future Research

We Recognize the limitations of this study due to its small sample size and short duration but accept them within an exploratory study framework. The objectives of this initial study were: 1) to act as a pretest to determine areas for improvement and debugging if the experiment were to be enacted later on a much larger scale; 2) to view reactions and behavior of the subjects; and 3) to see if the initial results meant with prior expectations.
Initial results are encouraging. We believe the experiment and its results have shown potential value for policy makers. To make the study more effective we need larger sample size (30 industries) , longer time frame without treatment changes (20 to 25 turns), and allow total collusion to occur. The results if they replicate the preliminary findings of this paper would substantially validate the study and provide more confidence towards its findings.

MANAGEMENT IMPLICATIONS

The results of this exploratory study tends to confirm the hypothesized extension of Deutsch's theory of cooperation and competition from individuals to the interfirm level. The consequences of cooperation and competition at the industry, interfirm level appear to be analogous to that which occur at the individual, intraorganizational level with the concept of productivity replaced by combined industry profits. Results indicate conclusively that providing group rewards leads to Cooperative behavior among firms and a resulting optimum level of industry profits.
We have conditionally proven that the reward structure does influence the resulting competitive behavior of the industry and the firm. For firms which are rewarded based upon its own performance and compared to all other firms in the industry, competitive behavior will be seen. For firms which are rewarded based upon the performance of its group, a cooperative behavior resulted. Cartel formation and maintenance was seen only under cooperative behavior. Attempts to establish cartels in the face of competitive pressures led to inevitable defection and a resulting competitive equilibrium.
The message to American policy makers is when the American economic system encourages competition, competitive behavior will result. If you really wish to emulate the Japanese cooperative methods ( a step the author would caution against as the cure might be worse than the disease) then you must change the incentive and disincentive system to allow cooperative activities to blossom. It is at cross purposes to encourage with one hand and to take away with the other. Changes to encourage cooperation include less stringent antitrust regulations, industry cartels, R&D consortiums, and a longer term perspective from the financial community. Competition is encouraged from birth in the United States and attempts to lessen it would take severe cultural changes with potentially diastrously side effects. Before policy makers act, they should ask several questions: 1) Is the Japanese success really due to cooperative elements of their firms or is it culturally engrained down even to the individual worker?; 2)Do the Japanese firms cooperate as much as the myth has it or is cooperation a governmental (MITI) tool used by Japan to develop and capture new markets and technologies?; and 3)What would be the effects of minimizing the competitive drive of the American people; how can the system be altered to provide global competition without losing the drive, creativity, and entrepreneurial aspects that competition provides?

References

Alston, Jon P. (1986), The American Samurai, Walter De Gruyter, Berlin

Deutsch, M. (1949), " A Theory of Cooperation and competition", Human Relations, Vol 2, 129-152

Gleeson, Russ (1987), "CoOperation Breeds Content,"Rydge's (Australia) 60/4,68-71

Littlejohn, Stephen E. (1986) "Competition and CoOperation: New Trends in Corporate Public Issue Identification and Resolution,"California Management Review,29/1, 109-123

Leung, Kwok and Hun-Joon Park (1986), " Effects of Interactional Goal on Choice of Allocation Rule: A Cross National Study", Organizational Behavior and Human Decision Processes, Vol 37, 111-120

Morris, Michael H., Gordon Paul, and Don Rahtz (1987),"Organizational Rewards and Coalitions in the Industrial Buying Center" International Journal of Research in Marketing (Netherlands), 4/2,131-146

Murphy,William J.(1988),"Interfirm Cooperation in a Competitive Enviornment",American Business Law Journal,26/1,29-56

Rauschenbach, Thomas M. (1988),"Competitiveness and CoOperation in a Global Industry,"International Journal of Technology Management,3/3,345-349

Tjosvold, Dean (1988),"Cooperative and Competitive Dynamics within and Between Organizational Units,"Human Relations,41/8,425-436

______________(1986)," The Dynamics of Interdependence in Organizations", Human Relations, Vol 39, Issue 6 (JUne), 517-540

Ulrich, Dave (1983),"Governing Transactions: A Framework for Cooperative Strategy,"Human Resource Management, 22/1/2/ (Spring/Summer), 23-39

Yoshida,Kosaku (1985),"Sources of Japanese Productivity: Competition and Cooperation,"Review of Business,7/3 (Winter),18-20

ormation can only occur under .A modern fadsof American Management is to compare the competitive-based American economic system to that of the Japanese system with its highly publicized and popularized Co-Operative elements. Almost unanimously, the studies conclude that the secret of the success of the Japanese economy is their reliance on Co-Operation between firms within an industry and propose that if American companies are to survive in this global economy ,they must make more efforts to co-operate with their competition and not engage in sometimes deadly competitive feuds. The cry is in the air, "Cooperate. Cooperate." But no matter how loud the hue and cry, when the basic underpinnings of the American economy is based upon individual firm performance and rivalry rather than cooperation among firms , then competition will be the dominant economic form of behavior.

This paper is a game theoretic experiment of this dilemma. The major premise is it is the reward structure that determines which of the dichotomy is chosen. If the performance of the entire group is the basis of rewards,the resulting tendencies will be towards cooperation and a stabilization at the point where group profits are maximized; this despite the fact (and consequently carrying all the more weight) that the group maximization point is not an equilibrium. Contrarily, If the individual firm is rewarded based upon its individual performance compared to all other firms, then competition will result to the detriment of the industry in total. Even if the competition is known and collusion allowed, the reward structure will tend to induce competition and prohibit long term cooperation between firms within an industry.
The conclusion of this exercise is not to make any judgments about the relative merits of cooperation versus competition but to indicate to policy makers that the reward structure must be in place (and penalty disincentives such as anti-trust or cartel formation removed) to incite the cooperative behavior seemingly so much desired nowadays.
This paper is a game theoretic experiment of this dichotomy-- competition versus cooperation. The major premise is that it is the reward structure that determines which of the dichotomy is chosen. If the performance of the entire industry is the basis of rewards, the resulting tendencies will be towards cooperation and a stabilization at the point where group profits are maximized; this despite the fact (and consequently carrying all the more weight) that the group maximization point is not an equilibrium. Contrarily,If the individual firm is rewarded based upon its individual performance compared to all other firms, then competition will result to the detriment of the industry in total. Even if the competition is known and collusion allowed, the reward structure will tend to induce competition and prohibit cooperation.

Table 6: 2 way anova between treatments and team scores

Table 4: 2 way anova between treatments and industry scores

Table 5: Multiple Regression upon Team and Group Scores

5a. Team score regressed upon knowledge and rewards

Table 5b: Industry scores regressed upon knowledge and rewards

Table 4: Table of Means

Know-notknow 3.770 .000 1.98 .052*

Group-Indiv 9.51 .000 * 6.316 .000*

GroupK-GroupNk 5.871 .000 3.536 .001

IndivK- Indiv NK 1.144 .256* .268 .790*

* indicates results agree with hypothesis

It is not the intent of this exercise to make judgments concerning the relative merits of cooperation versus competition but to indicate to policy makers that the reward structure must be in place (and penalty disincentives such as anti-trust or cartel formation removed) to incite the cooperative behavior so seemingly desired nowadays.