- Email: [email protected]
Morality in marketing: Oxymoron or good business practice?
Journal of Business Research 64 (2011) 922–927
Contents lists available at ScienceDirect
Journal of Business Research
Morality in marketing: Oxymoron or good business practice? Alison Watkins a, Ronald Paul Hill b,⁎,1 a b
Information Systems in the College of Business at University of South Florida St. Petersburg, United States The Richard J. and Barbara Naclerio Chairholder in Business and Senior Associate Dean of Intellectual Strategy at the Villanova School of Business, United States
a r t i c l e
i n f o
Article history: Received 1 April 2010 Received in revised form 1 June 2010 Accepted 1 August 2010 Available online 11 April 2011 Keywords: Marketing Morality Agent-based simulations Profitability
a b s t r a c t Marketing morality as oxymoron or good business practice is the primary focus of this research and is a topic deserving of attention by the field. Much theory on morality involves issues like salesperson behavior that fails to examine the impact on important goals. For instance, ethicists evaluate and provide recommendations and prohibitions without concern for their influence on the financial well-being of firms. This disconnect has unintended consequences. First, moral dictates often are naive by the standards of marketing managers, outof-touch with complexities of real world firms. Second, scholars discount prescriptions for change since they do not reflect the full range of management requirements for success. Is marketing morality an oxymoron or good business practice? The time is right for marketers to take a serious look. © 2011 Elsevier Inc. All rights reserved.
1. Introduction Marketing morality as oxymoron or good business practice is a rhetorical question. Recent financial and corporate governance scandals notwithstanding, some commentators tie the marketing profession to the “whipping post” to purge in public settings corporations' lack of morals. From deceptive advertising to illicit sales tactics, business reporters traditionally charge marketers with a lack of good faith in the creation and conduct of exchange relationships. Of course, none of this behavior makes sense in light of current thinking about the intersection of moral business practice and relationship marketing (Murphy et al., 2007). Firms and individuals seek trustworthy partners who make and fulfill commitments during transactions. Less is unacceptable and leads to search for other opportunities. Walker and Pitts (1998) attribute some portion of this moral dilemma to the lack of an identifiable moral ideal, something even leading philosophers struggle with over time. Marketers and managers wrestle with the practical implications, asking that they make deals in the best interests of their firms while simultaneously serving the needs of others (Desmond and Andrew, 2004). The potential conflicts are intuitive and obvious. Nonetheless, the significance of developed moral reasoning capability is clear, especially learning the skills necessary to navigate such competing demands (Monga, 2007). This requirement occurs within a variety of
⁎ Corresponding author at: Villanova University, 800 Lancaster Avenue, Villanova, PA 19085, United States. Tel.: + 1 610 519 3256. E-mail address: [email protected] (R.P. Hill). 1 He is thankful for the support provided by the Naclerio family. 0148-2963/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2011.03.009
business settings where testing of different hypotheses takes place (Trevino, 1992). One framework with potential for organizing our thinking is Kohlberg's (1984) cognitive moral development theory. His approach is useful for understanding ethical and unethical behavior in accounting, management, and marketing (Robin et al., 1996). According to this theoretical paradigm, moral reasoning advances along three stages from pre-conventional to conventional to postconventional. The first stage is characterized by extreme egoism that concentrates on maximizing self-gain and minimizing self-loss by all means (Monga, 2007). Using the principal-agent model popular in marketing and management, the only barrier to dishonesty in exchange relationships is fear of getting caught and punishment. Adherence to a set of rules or regulations that arise from reference groups such as peers in a marketing department and/or larger societal or cultural collectives defines the second stage (Falkenberg, 2004). Persons at this developmental level are more consistent in their behaviors, and some stability in their interactions occurs. The third and final stage emphasizes growing moral autonomy rising above external influences to produce an individual set of ethical principles. In an advanced form, selflessness emerges that heightens awareness of others' situations and compassion for helping them achieve goals (Kjonstad and Willmott, 1995). Armstrong (1977) presents such an environment in his discussion of subjects who must represent interests of stakeholder groups, which changes willingness to bring a dangerous product to market. This model suggests that moral reasoning in applied contexts exists along a continuum from pure selfishness to pure selflessness (Watkins and Ronald Paul, 2005). At one end are marketers with poor cognitive skills; these marketers are unable to comprehend or care about the needs of their customers because of personal flaws that
A. Watkins, R.P. Hill / Journal of Business Research 64 (2011) 922–927
predispose them to self-centeredness focusing exclusively on personal interests during exchanges (Ashkanasy et al., 2007). At the opposite end are marketers who operate as moral individuals, empathically and compassionately nurturing their relationships with customers (Walker and Pitts, 1998). They have what Werhane (2006) calls “moral imagination,” requiring awareness of and sensitivity to ethical issues surrounding exchanges. Simulations are increasingly popular methods of exposing students and executives to the complex environment in which morality plays out in businesses. For example, ISLAND TELECOM is a multiplayer, Internet-based game that allows participants to resolve moral problems resulting from globalization (Bos et al., 2006). Further, the SLIPPERY SLOPE simulation captures consequences of acting with integrity, or trying to do what a person believes is right despite negative outcomes (Teach et al., 2005). While games have learning value, they do not examine the impact of specific moral types on key success factors. As a result, other forms of simulation that allow testing are in use for a variety of purposes. Agent-based social simulations for investigating problems that derive from the prisoner's dilemma and the tragedy of the commons show promise (Gotts et al., 2003). These simulations include various sorts of egoism and altruism and lead to research efforts with advanced modeling techniques. Artificial Life reproductions of living systems represent another category of simulation. Creation of moral agents who are capable of developing trust that may guide perceptions and actions are possible in these environments (Tuomela and Hofmann, 2003). Thus, these simulations place autonomous agents with different moral foundations in ethical quandaries where they support (or detract from) the well-being of others (Sullins, 2005). This discussion presents opportunities for research that form the principal basis for the investigation. The present study examines impact of morality among marketers on profitability of their firms. Manipulating moral imagination or the lack there of cannot occur easily in real-world situations; researchers may create business-tobusiness simulations using SUGARSCAPE versions of Artificial Life computer-generated models to resolve this dilemma. Possible contributions of this project are in the next section, as well as additional grounding in the marketing literature. The specific SUGARSCAPE application is in the section that follows, with an emphasis on marketers as moral agents and the cumulative impact of their decisions. The paper closes with implications for understanding simulations and morality in marketing. 1.1. Theoretical orientation The concept of morality refers to perceptions and behaviors that flow from norms or rules that distinguish right from wrong (Brinkmann, 2002). Values, beliefs, and attitudes help determine one's cognitive moral development and one's recognition of and reactions to particular ethical dilemmas (Singhapakdi et al., 1996). Conventional frames that drive personal choices are posited by research on moral decision-making. For example, organizations that people belong to dominate these contexts, and they implicitly or explicitly direct employees to conform to a common set of principles (Gick, 2003). Nevertheless, individuals who deviate from organizational standards but are successful at meeting or exceeding larger goals and objectives of firms may survive or even thrive regardless of moral underpinnings (Schroeder, 2000). Attitudes and actions of individuals in boundary-spanning roles like salespeople often are the basis for application of morality to marketing (Hunt and Vitell, 2006). Marketplace activities among exchange partners optimally are equitable, transparent, and productive as well as stable over time (Klein, 2003). Nonetheless, relationship marketing notwithstanding, this boundary-spanning nature may cause or exacerbate many ethical dilemmas (Valentine
923
and Barnett, 2002). Goolsby. and Hunt (1992) posit that moral problems arise because of the role marketing plays in assimilating wants and needs of various constituencies, especially when demands, desires, and/or expectations are in conflict. Problems are particularly germane with sales professionals because of intense pressures on agents to meet their goals (Cherry and Fraedrich, 2002). Hill and Watkins (2007) address differences in marketer morality within a framework that is oriented toward moral reasoning theory. Their study examines impact of moral philosophies of selling agents in business-to-business settings on firm and industry sales. Philosophies are extremes of the cognitive development spectrum as well as several intermediate positions. “True egoists” are sellers at one end; these sellers act in completely self-serving ways consistent with preconventional morality. “True altruists” are sellers at the other end; these sellers seek to meet the needs of buyers without regard for themselves like individuals at the post-conventional stage of morality. In between these extremes are three additional prototypes that seek to span the gap from one endpoint to the other. The Iterative Prisoner's Dilemma informs the creation of these simulations with variations in moral maturity, payoffs, and decision-making heuristics. Results demonstrate the long-term positive impact of moral maturity on financial performance. For example, simulation firms made up of altruists earn more income than firms containing egoists. This outcome is corroborated by results from other simulations with multiple philosophies where moral maturity advances or declines over time. Under these circumstances, economic success improves from the most egoistic collective to the most altruistic grouping. Sensitivity analyses that manipulate payoffs and heuristics take place, and the results remain robust except under extraordinary circumstances that over-reward more egoistic behavior relative to more altruistic behavior, or hide selfishness from marketing partners for an extended and unacceptable length of time. Investigation of relationships between marketer moral maturity and buyer satisfaction/firm profitability in business-to-business contexts occurs in this study. With a basis in research on embedded morality, selling agents have an array of orientations including egoists, realists, loyalists, or altruists. This method also involves more complex and dynamic simulations of business-to-business exchanges that represent a significant modification beyond previous studies of agent-based computer models of artificial life. In this environment, buying agents seek selling agents to exchange with according to previous experiences and, in some circumstances, wordof-mouth. Market knowledge accumulates over time and reflects tolerance for variation in price offerings relative to actual delivered value. Morality plays an important role within buyer–seller exchanges in a number of contexts, but the area of pricing is most appropriate for the conduct of this research (Xia et al., 2004). Buyers are able to evaluate aspects of transaction quality as a result of marketers' pricing strategies, leading to subjective determinations of ethical and unethical treatments. Such evaluations result from comparisons to internal norms or external standards as judgment bases (Rallapalli et al., 2000). When value (quality-to-price ratio) is lower than expectations, sellers' reputations may suffer through the spread of negative information. High levels of price dispersion within the marketplace are necessary, allowing for enough variation to produce diverse outcomes. 2. Simulating marketing exchange relationships Usage of SUGARSCAPE versions of agent-based social simulations occurs because of its unique approach to the creation of interactive and dynamic environs (Epstein and Axtell, 1996). According to Agar (2005), such models test hypothetical situations rather than represent realworld systems, with the analysis capabilities of quantitative models. Epstein and Axtell (1996) employ SUGARSCAPE to build historical
A. Watkins, R.P. Hill / Journal of Business Research 64 (2011) 922–927
models of civilizations to understand effects on evolution of social phenomena. Subsequent uses include modeling communication and cooperation within animal societies (Buzing et al., 2003), estimating effects of taxation on evolution (Bäck et al., 2002), and studying assortments of cooperative behaviors among agents (Nishizaki et al., 2004). A similar philosophy advances agent-based computational economics (ACE) models, which involve computer science, cognitive science, and evolutionary economics (Vag, 2004). Within ACE models, agents learn from their actions and the actions of others, and modify individual behaviors (Tesfatsion, 2003). For example, Kirman and Vriend (2001) present an ACE model using two groups—buyers who learn loyalty results in lower prices and sellers who learn lower prices bring greater loyalty. In reinforcement learning, participants begin with simple rules about their environment and learn through a system of rewards and punishments which behaviors achieve the desired results. Their research provides a starting point for this study, which concentrates on a more complex model of reinforcement learning that involves multi-agent/environment interactions that synchronize and occur in discrete cycles. The orientations of selling agents capture extremes of moral maturity that are consistent with previous research (Hill and Watkins, 2007). Agents exist within single systems or industries to avoid contamination from extraneous factors or interactions, and each agent affiliates with one buyer or seller firm so that a set of players and their actions are easy to identify for purposes of relationship building. Seller firms contain 100 agents and two types of agent distributions were examined in this study; a uniform approach where agents within a firm had the same moral maturity or a mixed mode consisting of one dominant type and an even mix of the other types. Additionally buyers in the initial simulation did not directly communicate their experiences with one another but in the latter model communication occurred whenever buyers met. Agent Types on the selling side of exchanges include egoists who seek short-term financial gain for themselves with disregard for well-being of their trading partners; altruists who selflessly try to meet the desires of trading partners even if the result is negative impact on short-term profitability; realists who seek to maximize gain but do so with the guise of maintaining relationships with buyers; and loyalists who attempt to advance firm reputations among buyers even if the result negatively impacts personal success. Buyer firms also contain 100 agents in each, whose primary job is to find and consummate exchanges of good value. Sellers' ethical orientations guide price setting decisions. The base price of the simulated product is $750, and sellers can add markups up to $250. Sellers' transactional incomes are ½ base prices plus the full markup; therefore, one might expect more egoistic selling agents with a profit maximization goal to charge highest possible amounts, yet altruistic sellers will find that no markup brings greater personal satisfaction. However, as with human agents, an element of chance exists within the simulation that they might select a lower markup at any given point. Please note that the other orientations are more sensitive to buyer reactions to prices. Buyers move within the marketplace searching for sellers with whom they can have positive experiences. Although buyers find satisfaction when sellers make their offerings at base price, they do have a level of tolerance for markups. Buyers use a uniform system to rate transactions so that sellers get instant feedback on their actions. While real-world buyers may not have a uniform manner for such recognition, most individual appraisals are based on fair treatment and this is reflected in the model. Sellers receive ratings as feedback once exchanges are complete and all but the egoists use these to assist pricing decisions. Buyers, firms, and the industry transform transactional ratings into: (a) an individual buyer–seller relationship rating (B2S); (b) an average firm rating of the seller (F2S); and (c) an overall industry average rating of
Income (in millions) Amount
924
8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 2%
25%
50%
75%
100% 125% 150% 175% 200% 225%
% of Sellers Met by Buyers Egoists
Realists
Loyalists
Altruists
Fig. 1. Mean seller income (in millions) for uniform moral codes without communication.
the seller (I2S). Buyers use ratings and past data on positive/negative experiences to determine whether to complete a purchase. 2.1. Simulation results: uniform and mixed morality without communication Uniform moral maturity within seller firms and no internal or external word-of-mouth among buying agents defines the first baseline simulation model (Fig. 1). As a result, seller firms present a one-dimensional corporate culture, while buying agents are left to accumulated experiences to determine with whom they desire to transact. Buying agents are naive during initial exchanges; however, this scenario does not last long since they learn how to discriminate among selling agents. By the time experience reaches 225%, the point at which all buyers communicate with all sellers at least twice, income rises from egoists ($788 million), to realists ($3.3 billion) to loyalists ($5.5 billion) to altruists ($6.0 billion). Significant differences across seller firms occur (ANOVA: F(3,76) = 163.94, pb0.0001), and t-tests of mean differences are as well. Useful and counterintuitive findings concern intermediate levels of experience between 2% and 225%, as income and order of leadership change regularly during this interval. For instance at 25% (buying agents communicate with one-fourth of selling agents), egoists ($301 million) are in fourth place, altruists third ($482 million), loyalists second ($529 million), and realists first ($571 million). The ANOVA is significant (F(3,76) = 878.25, pb0.0001), as are t-tests. Another change happens at 125%, with egoists ($590 million) still least profitable and the remainder in an increasing order from realists ($2.4 billion), to altruists ($2.8 billion), to loyalists ($3.1 billion). Once again, ANOVA and t-tests are significant (F(3,76) = 198.25, pb0.0001). The capability of SUGARSCAPE to establish theoretical relationships between moral maturity and financial success within a business-to-business sales context occurs with the previous model. Nevertheless, this situation lacks diversity that typically exists within and across seller firms. A completely homogeneous corporate culture where every sales agent has the same moral sense is unlikely. Instead, firms may have a dominant approach that captures most of their employees along with minority perspectives that represent individuals outside the mainstream. Therefore, the model revision includes a majority of one type and a minority of the remaining three. This simulation contains 4 seller firms of 55 egoists, realists, loyalists, or altruists and 15 each of the other maturity levels. Similarities exist between uniform and the mixed morality. Incomes of seller firms at 2% penetration are identical; while the egoist dominant firm falls to fourth place and remains in this position for the rest of the simulation after the 25% rate. Surprisingly, a stable pattern of relative success occurs earlier at the 125% level, and this ordering is distinctive beyond this point in that the predominantly egoist firm ($1.5 billion) makes significantly less than the other three, but the realist ($2.5 billion) and altruist ($2.4 billion) firms earn about
2%
100%
225%
% of Sellers Met by Buyers Egoists
Realists
Loyalists
925
100%
150%
Altruists
Loyalists
Egoists
Realists
Altruists
Loyalists
Egoists
Realists
Altruists
Loyalists
Egoists
2%
Realists
Altruists
Realists
Loyalists
0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 Egoists
Altruists
Loyalists
Egoists
150%
Realists
Altruists
Loyalists
Egoists
Realists
Altruists
Loyalists
Egoists
Realists
Altruists
Loyalists
Egoists
Mean Amount Earned per Agent (in millions)
0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 Realists
Mean Amount Earned per Agent (in millions)
A. Watkins, R.P. Hill / Journal of Business Research 64 (2011) 922–927
225%
% of Sellers Met by Buyers
Altruists
Egoists
Realists
Loyalists
Altruists
Fig. 2. Mean individual seller income (in millions) at intervals for mixed moral codes without communication.
Fig. 4. Mean individual seller income (in millions) at intervals for mixed moral codes without communication.
the same amount while the loyalist firm ($2.6 billion) acquires more by a significant margin (F(3,76) = 213.51, pb0.0001). Finally, ultimate success of the loyalist dominant firm notwithstanding, income totals at 225% penetration supply some interesting findings compared to their complement. The primarily egoist firm earns about $2.4 billion, an increase of about 300% over the purely egoist firm, and the mainly realist firm ($4.3 billion) makes more than its parallel firm at 129%. Conversely, the mostly loyalist ($4.8 billion) and altruist ($4.5 billion) firms do less well in comparison at 88% and 75% respectively. Some basic similarities as well as some distinctive findings are the result of an attempt to provide a realistic representation of the role of morality in sales/financial achievement when compared to the uniform simulation. Early naivety gives way to discrimination powers that doom the primarily egoist firm to a distant fourth place in profitability. In comparison to their uniform complement, relative success that finally emerges over-rewards two least morally mature collectives relative to two more mature firms. The underlying cause is that some less morally mature selling agents are able to prosper in firms that present a mixed but morally advanced corporate culture, potentially masking self-centered intentions and allowing opportunities to transact as a consequence.
Given increasing speed experience accrues relative to previous models, powers of discrimination develop quickly. Egoists, realists, loyalists, and altruists earn the same amounts at low penetration rates within baseline data from the uniform scenario. However, by 25% egoists are at the bottom of the income hierarchy, earning half that of their closest competitor. By 75%, realists relegate to third position and a growth curve that is increasing at a decreasing rate. Finally, profitability of sellers relative to one another stabilizes at 150%, and the same pattern as the previous uniform run emerges as income amplifies from egoists ($348 million) to realists ($2.2 billion) to loyalists ($3.6 billion) to altruists ($4.1 billion) (F(3,76) = 1872.70, pb0.0001/all t-tests significant). Yet, while egoists fall into disrepute early, those who exist where information sharing takes place earn less than half (43%) as much as their no-sharing counterparts at 225% penetration. Further, both realists (78%) and loyalists (85%) are less profitable than their no-sharing complements, with only altruists (125%) gaining ground. This mixed simulation model also includes internal and external communication. While flip-flopping of profit leadership does occur across runs, the final pattern surfaces at the 100% level. Then, the mostly egoist firm ($1.1 billion) remains at the bottom, with the mainly realist firm ($1.9 billion) and altruist ($2.0 billion) and loyalist ($2.1 billion) dominant firms following. (ANOVA/F(3,76) = 1872.70, pb0.0001/t-tests significant between egoists and realists and the other two types.) Seller firms containing chiefly egoists ($2.5 billion) bring in 600% as much as their uniform with communication counterparts, while the principally realists firm produces 150% of their corresponding firm's income at 225% penetration. The loyalist and altruist dominant firms do less well in comparison at 100% and 65% respectively. One distinction is simulation results establish a stable pattern at earlier penetration, although the final result fails to fully discriminate among the stages of moral maturity. Another comparison involves findings from two mixed simulations. Sharing of information within and between seller firms has a modest dampening effect on total income by the end of the run for both egoist and realist oriented firms, with income falling slightly to 94% and 92% correspondingly. Also, the principally loyalist firm experiences a drop to 97% whereas the largely altruist firm expands to 108%. The sharing model causes realist, loyalist, and, to a large extent, altruist agents to improve average profitability as the primary orientation of firms shifts to the more mature in patterns similar to the no-sharing model. However, the altruists eventually stand out in the sharing circumstance, while loyalists tend to do so in the no-sharing situation. Addition of internal and external information sharing impacts results, marginally reducing penetration rate necessary to gain a stable marketplace and shift profit towards altruists. Nonetheless, clear demarcations according to level of moral maturity and selling agents' interest in firms' reputations with or well-being of exchange
2.2. Simulation results: uniform and mixed morality with communication
Income (in millions) Amount
Consistent moral maturity in contrast to mixed seller firms also takes place in the second set of simulations, but internal communication and external word-of-mouth are added to decision calculus of buying agents. When buying agents make contact, they provide each other with their storehouse of seller knowledge on agents that are unknown. Such data are amassed as agent-to-agent, agent-to-firm, and firm-to-firm. Buyer agents who are given access to the quality of exchange relationships of another agent encode data as if their own. 8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 2%
25%
50%
75%
100% 125% 150% 175% 200% 225%
% of Sellers Met by Buyers Egoists
Realists
Loyalists
Altruists
Fig. 3. Mean seller income (in millions) for uniform moral codes with communication.
926
A. Watkins, R.P. Hill / Journal of Business Research 64 (2011) 922–927
partners fails to materialize as they did in the uniform/sharing simulation. Thus, discrimination capabilities among buying agents seem less robust than they do under ideal conditions of onedimensional corporate cultures with widespread information exchange. In the end, buyers may allow sellers with selfish motivations to prosper, especially if they have the good fortune to reside in morally mature firms with largely selfless agents. 3. Discussion and conclusions The real value of models of ethical behavior flows from the ability to capture the complexity of moral decision making (Treviño and Brown, 2004). Thus, this simulation permits examination of the impact of agent moral character on exchange relationships within a business-to-business sales context. Selling agents transact according to assigned levels of moral maturity, which advances from the most selfish to the most selfless in the following order: egoists who maximize their gain from every transaction, realists who extract as much as possible without jeopardizing future opportunities, loyalists who place their firms' reputations above all else, and altruists who meet the needs of exchange partners as their first priority. Buyers respond to these offers from personal experience and word-of-mouth. The industry maintains four firms on each side of the exchange equation, with seller firms containing agents with uniform or mixed moral maturity who operate accordingly. The uniform scenarios provide validation for basic tenets of the artificial life model of living systems as a reproduction of ethical marketing theory. Specifically, buyers and sellers work in an environment where the possibility exists for long-term associations that have a basis in shared understanding of transactional ingredients. Final ordering of profitability meets the expectation that mature agents and their firms prosper, with significant differences in net income rising from firms with entirely egoists to realists to loyalists to altruists. Communication acts as catalyst that speeds identification of sellers by buyers (Fig. 2). While development of a suitable model mirroring this theoretical understanding is important, unexpected nuances often provide the most interesting findings. The inability of buying agents to fathom the intentions of selling agents during early stages of market interactions is clear, but the changes and differences in profit leadership among the levels of moral maturity at various penetration rates are an unexpected twist. Additionally, the reduced capacity of buyers to comprehend the same intentions when seller firms house agents with a variety of moral maturity levels is intuitive, but the speed with which stable patterns of interaction occur and the preferences that emerge suggest considerable marketplace distortion. Finally, ability of less moral sellers to prosper in mature cultures is in contrast to the belief that such corporate environments disallow this possibility (Treviño and Brown, 2004). Marketing managers and other practitioners have several potential takeaways. For instance, the common finding that buying agents lack discriminatory powers in searches for selling agents until the marketplace reaches a critical mass suggests the possibility of widespread manipulation. While certainty is impossible regarding moral makeup of prospective exchange partners until sufficient information and/or experience accumulates, buyers who are aware of this deficit can avoid the pitfalls by employing additional scrutiny and limiting the number of transactions until they are better able to judge seller intentions. This strategy must take into account selfimposed restrictions that lead to shortages in goods or services necessary for ultimate success (Fig. 3). Since discriminatory powers fail to work properly when operable, they lead to suboptimal choices by significant margins. This finding is troublesome given the smaller size and intricacy of the artificial environment and the extent and accuracy of available intelligence relative to real world contexts. Such results beg the question: How can buying agents protect themselves from overconfidence until the
situation truly merits loyalty to particular sellers? While the answer may seem trite, the best remedy involves avoiding premature decisions until a fuller measure of selling agents takes place. The criteria to ensure appropriate selections are made include a wide range of interactions and/or information that triangulates around the same conclusions. The third inference is from mixed morality models, with potential value for buyers and sellers. Preference patterns that arise from SUGARSCAPE simulations pose difficulties in comparison to uniform morality and its theoretical underpinnings. Thus managers in buyer firms, who are likely to encounter at least some heterogeneity within seller firms, must recognize limitations in agents' powers of perception and selection as decision-making guides (Fig. 4). As before, seeking assurance from a larger variety of sources and avoiding early and extensive commitments are the best strategy. Managers in seller firms, who attempt to follow moral directions of the relationship marketing philosophy, can improve chances of success through recruitment and promotion of consistently morally mature selling agents. References Agar. Michael H. Agents in living color: towards Emic Agent-Based Models. Journal of Artificial Societies and Social Simulation 2005;8(1) http://jasss.soc.surrey.ac. uk/8/1/4.html. Armstrong Scott J. Social irresponsibility in management. Journal of Business Research 1977;5:185–213 (September). Ashkanasy Neal M, Windsor Carolyn A, Trevino Linda K. Bad apples in bad barrels revisited: cognitive moral development, just world beliefs, rewards, and ethical decision making. Business Ethics Quarterly 2007;16(4):448–73. Bäck Thomas, Vermeulen Daniel, Eiben Agoston E. Effects of tax and evolution in an artificial society. In: Caulfield HJ, Chen S-H, Cheng HD, Duro R, Honavar V, Kerre EE, Lu M, Romay MG, Shih TK, Ventura D, Wang P, Yang Y, editors. Proceedings of The Sixth Joint Conference on Information Sciences, (JCIS 2002); 2002. p. 1151–6. Bos Nathan D, Sadat Shami N, Sara Naab. A globalization simulation to teach corporate social responsibility: design features and analysis of student reasoning. Simulation & Gaming 2006;37:56–72 (March). Brinkmann Johannes. Business and marketing ethics: concepts, approaches and typologies. Journal of Business Ethics 2002;41:159–77 (November/December). Buzing Pieter, Agoston Eiber, Schut Martjin C. Evolving agent societies with VUScape. Lecture Notes in Computer Science, 2801. Springer Berlin; 2003. p. 434–41. Cherry John, Fraedrich John. Perceived risk, moral philosophy and marketing ethics: mediating influences on sales managers' ethical decision- making. Journal of Business Research 2002;55:951–62 (December). Desmond John, Andrew Crane. Morality and the consequences of marketing action. Journal of Business Research 2004;57:1222–30 (November). Epstein Joshua M, Axtell Robert. Growing artificial societies: social science from the bottom up. Cambridge, MA: MIT Press; 1996. Falkenberg Andreas Wyller. When in Rome … moral maturity and ethics for international economic organizations. Journal of Business Ethics 2004;54:17–32. Gick Evelyn. Cognitive theory and moral behavior: the contribution of F. A. Hayek to business ethics. Journal of Business Ethics 2003;45:149–65 (June). Goolsby. Jerry R, Hunt Shelby D. Cognitive moral development and marketing. Journal of Marketing 1992;56:55–68 (January). Gotts NM, Polhill JG, Law ANR. Agent-based simulation in the study of social dilemmas. Artificial Intelligence Review 2003;19:3-92. Hill Ronald Paul, Watkins Alison. A simulation of moral behavior within marketing exchange relationships. Journal of the Academy of Marketing Science 2007;35(2): 417–29. Hunt Shelby D, Vitell Scott J. The general theory of marketing ethics: a revision and three questions. Journal of Macromarketing 2006;26:143–53 (December). Kirman Alan P, Vriend Nicolaas J. Evolving market structure: an ACE model of price dispersion and loyalty. Journal of Economic Dynamics & Control 2001;25:459–502. Kjonstad Bjorn, Willmott Hugh. Business ethics: restrictive or empowering? Journal of Business Ethics 1995;14(6):445–64. Klein Sherwin. The natural roots of capitalism and its virtues and values. Journal of Business Ethics 2003;45:387–401 (July). Kohlberg Lawrence. The psychology of moral development: the nature and validity of moral stages. San Francisco: Harper & Row; 1984. Monga Manjit. Manager's moral reasoning: evidence from large Indian manufacturing organizations. Journal of Business Ethics 2007;71:179–94. Murphy Patrick E, Laczniak Gene R, Wood Graham. An ethical basis for relationship marketing: a virtue ethics perspective. European Journal of Marketing 2007;41(1/2): 37–57. Nishizaki Ichiro, Sakawa Masatoshi, Katagiri Hideki. Influence of environmental changes on cooperative behavior in the prisoner's dilemma game on an artificial social model. Applied Artificial Intelligence 2004;18:651–71. Rallapalli C, Vitell Scott J, Sheryl Szeinbech. Marketers' norms and personal values: an empirical study of marketing professionals. Journal of Business Ethics 2000;24: 65–75 (March).
A. Watkins, R.P. Hill / Journal of Business Research 64 (2011) 922–927 Robin Donald P, Gordon Gus, Jordan Charles, Eric Reidenbach R. The empirical performance of cognitive moral development in predicting moral intent. Business Ethics Quarterly 1996;6(4):493–515. Schroeder Donna J. Morality de-kanted or the biological roots of moral behavior. International Journal of Values-Based Management 2000;13:297–308 (January). Singhapakdi Anusom, Vitell Scott J, Kraft Kenneth L. Moral intensity and ethical decision-making of marketing professionals. Journal of Business Research 1996;36: 245–55 (July). Sullins John P. Ethics and artificial life: from modeling to moral agents. Ethics and Information Technology 2005;7:139–48. Teach Richard D, Christensen Sandra L, Schwartz Robert G. Teaching business ethics: integrity. Simulation & Gaming 2005;36:383–8. Tesfatsion Leigh. Agent-based computational economics: modeling economies as complex adaptive systems. Information Sciences 2003;149:263–9. Trevino Linda K. Moral reasoning and business ethics: implications for research, education, and management. Journal of Business Ethics 1992;11:445–59. Treviño Linda K, Brown Michael E. Managing to be ethical: debunking five business ethics myths. Academy of Management Executive 2004;18(2):69–81.
927
Tuomela Maj, Hofmann Solveig. Simulating rational social normative trust, predictive trust, and predictive reliance between agents. Ethics and Information Technology 2003;5(3):163–76. Vag Andras. First generation multi-agent models and their upgrades. Interdisciplinary Description of Complex Systems 2004;2(1):95-103. Valentine Sean, Barnett Tim. Ethics codes and sales professionals' perceptions of their organizations ethical values. Journal of Business Ethics 2002;40:191–200 (October). Walker Lawrence J, Pitts Russell C. Naturalistic conceptions of moral maturity. Developmental Psychology 1998;34(3):403–19. Watkins Alison, Ronald Paul Hill. The impact of personal and organizational moral philosophies on marketing exchange relationships: a simulation using the prisoner's dilemma game. Journal of Business Ethics 2005;62:253–65 (December). Werhane Patricia H. A place for philosophers in applied ethics and the role of moral reasoning in moral imagination: a response to Richard Rorty. Business Ethics Quarterly 2006;16(3):401–8. Xia Lan, Monroe Kent B, Cox Jennifer L. The price is unfair! a conceptual framework of price fairness perceptions. Journal of Marketing 2004;68:1-15 October.
Contents lists available at ScienceDirect
Journal of Business Research
Morality in marketing: Oxymoron or good business practice? Alison Watkins a, Ronald Paul Hill b,⁎,1 a b
Information Systems in the College of Business at University of South Florida St. Petersburg, United States The Richard J. and Barbara Naclerio Chairholder in Business and Senior Associate Dean of Intellectual Strategy at the Villanova School of Business, United States
a r t i c l e
i n f o
Article history: Received 1 April 2010 Received in revised form 1 June 2010 Accepted 1 August 2010 Available online 11 April 2011 Keywords: Marketing Morality Agent-based simulations Profitability
a b s t r a c t Marketing morality as oxymoron or good business practice is the primary focus of this research and is a topic deserving of attention by the field. Much theory on morality involves issues like salesperson behavior that fails to examine the impact on important goals. For instance, ethicists evaluate and provide recommendations and prohibitions without concern for their influence on the financial well-being of firms. This disconnect has unintended consequences. First, moral dictates often are naive by the standards of marketing managers, outof-touch with complexities of real world firms. Second, scholars discount prescriptions for change since they do not reflect the full range of management requirements for success. Is marketing morality an oxymoron or good business practice? The time is right for marketers to take a serious look. © 2011 Elsevier Inc. All rights reserved.
1. Introduction Marketing morality as oxymoron or good business practice is a rhetorical question. Recent financial and corporate governance scandals notwithstanding, some commentators tie the marketing profession to the “whipping post” to purge in public settings corporations' lack of morals. From deceptive advertising to illicit sales tactics, business reporters traditionally charge marketers with a lack of good faith in the creation and conduct of exchange relationships. Of course, none of this behavior makes sense in light of current thinking about the intersection of moral business practice and relationship marketing (Murphy et al., 2007). Firms and individuals seek trustworthy partners who make and fulfill commitments during transactions. Less is unacceptable and leads to search for other opportunities. Walker and Pitts (1998) attribute some portion of this moral dilemma to the lack of an identifiable moral ideal, something even leading philosophers struggle with over time. Marketers and managers wrestle with the practical implications, asking that they make deals in the best interests of their firms while simultaneously serving the needs of others (Desmond and Andrew, 2004). The potential conflicts are intuitive and obvious. Nonetheless, the significance of developed moral reasoning capability is clear, especially learning the skills necessary to navigate such competing demands (Monga, 2007). This requirement occurs within a variety of
⁎ Corresponding author at: Villanova University, 800 Lancaster Avenue, Villanova, PA 19085, United States. Tel.: + 1 610 519 3256. E-mail address: [email protected] (R.P. Hill). 1 He is thankful for the support provided by the Naclerio family. 0148-2963/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.jbusres.2011.03.009
business settings where testing of different hypotheses takes place (Trevino, 1992). One framework with potential for organizing our thinking is Kohlberg's (1984) cognitive moral development theory. His approach is useful for understanding ethical and unethical behavior in accounting, management, and marketing (Robin et al., 1996). According to this theoretical paradigm, moral reasoning advances along three stages from pre-conventional to conventional to postconventional. The first stage is characterized by extreme egoism that concentrates on maximizing self-gain and minimizing self-loss by all means (Monga, 2007). Using the principal-agent model popular in marketing and management, the only barrier to dishonesty in exchange relationships is fear of getting caught and punishment. Adherence to a set of rules or regulations that arise from reference groups such as peers in a marketing department and/or larger societal or cultural collectives defines the second stage (Falkenberg, 2004). Persons at this developmental level are more consistent in their behaviors, and some stability in their interactions occurs. The third and final stage emphasizes growing moral autonomy rising above external influences to produce an individual set of ethical principles. In an advanced form, selflessness emerges that heightens awareness of others' situations and compassion for helping them achieve goals (Kjonstad and Willmott, 1995). Armstrong (1977) presents such an environment in his discussion of subjects who must represent interests of stakeholder groups, which changes willingness to bring a dangerous product to market. This model suggests that moral reasoning in applied contexts exists along a continuum from pure selfishness to pure selflessness (Watkins and Ronald Paul, 2005). At one end are marketers with poor cognitive skills; these marketers are unable to comprehend or care about the needs of their customers because of personal flaws that
A. Watkins, R.P. Hill / Journal of Business Research 64 (2011) 922–927
predispose them to self-centeredness focusing exclusively on personal interests during exchanges (Ashkanasy et al., 2007). At the opposite end are marketers who operate as moral individuals, empathically and compassionately nurturing their relationships with customers (Walker and Pitts, 1998). They have what Werhane (2006) calls “moral imagination,” requiring awareness of and sensitivity to ethical issues surrounding exchanges. Simulations are increasingly popular methods of exposing students and executives to the complex environment in which morality plays out in businesses. For example, ISLAND TELECOM is a multiplayer, Internet-based game that allows participants to resolve moral problems resulting from globalization (Bos et al., 2006). Further, the SLIPPERY SLOPE simulation captures consequences of acting with integrity, or trying to do what a person believes is right despite negative outcomes (Teach et al., 2005). While games have learning value, they do not examine the impact of specific moral types on key success factors. As a result, other forms of simulation that allow testing are in use for a variety of purposes. Agent-based social simulations for investigating problems that derive from the prisoner's dilemma and the tragedy of the commons show promise (Gotts et al., 2003). These simulations include various sorts of egoism and altruism and lead to research efforts with advanced modeling techniques. Artificial Life reproductions of living systems represent another category of simulation. Creation of moral agents who are capable of developing trust that may guide perceptions and actions are possible in these environments (Tuomela and Hofmann, 2003). Thus, these simulations place autonomous agents with different moral foundations in ethical quandaries where they support (or detract from) the well-being of others (Sullins, 2005). This discussion presents opportunities for research that form the principal basis for the investigation. The present study examines impact of morality among marketers on profitability of their firms. Manipulating moral imagination or the lack there of cannot occur easily in real-world situations; researchers may create business-tobusiness simulations using SUGARSCAPE versions of Artificial Life computer-generated models to resolve this dilemma. Possible contributions of this project are in the next section, as well as additional grounding in the marketing literature. The specific SUGARSCAPE application is in the section that follows, with an emphasis on marketers as moral agents and the cumulative impact of their decisions. The paper closes with implications for understanding simulations and morality in marketing. 1.1. Theoretical orientation The concept of morality refers to perceptions and behaviors that flow from norms or rules that distinguish right from wrong (Brinkmann, 2002). Values, beliefs, and attitudes help determine one's cognitive moral development and one's recognition of and reactions to particular ethical dilemmas (Singhapakdi et al., 1996). Conventional frames that drive personal choices are posited by research on moral decision-making. For example, organizations that people belong to dominate these contexts, and they implicitly or explicitly direct employees to conform to a common set of principles (Gick, 2003). Nevertheless, individuals who deviate from organizational standards but are successful at meeting or exceeding larger goals and objectives of firms may survive or even thrive regardless of moral underpinnings (Schroeder, 2000). Attitudes and actions of individuals in boundary-spanning roles like salespeople often are the basis for application of morality to marketing (Hunt and Vitell, 2006). Marketplace activities among exchange partners optimally are equitable, transparent, and productive as well as stable over time (Klein, 2003). Nonetheless, relationship marketing notwithstanding, this boundary-spanning nature may cause or exacerbate many ethical dilemmas (Valentine
923
and Barnett, 2002). Goolsby. and Hunt (1992) posit that moral problems arise because of the role marketing plays in assimilating wants and needs of various constituencies, especially when demands, desires, and/or expectations are in conflict. Problems are particularly germane with sales professionals because of intense pressures on agents to meet their goals (Cherry and Fraedrich, 2002). Hill and Watkins (2007) address differences in marketer morality within a framework that is oriented toward moral reasoning theory. Their study examines impact of moral philosophies of selling agents in business-to-business settings on firm and industry sales. Philosophies are extremes of the cognitive development spectrum as well as several intermediate positions. “True egoists” are sellers at one end; these sellers act in completely self-serving ways consistent with preconventional morality. “True altruists” are sellers at the other end; these sellers seek to meet the needs of buyers without regard for themselves like individuals at the post-conventional stage of morality. In between these extremes are three additional prototypes that seek to span the gap from one endpoint to the other. The Iterative Prisoner's Dilemma informs the creation of these simulations with variations in moral maturity, payoffs, and decision-making heuristics. Results demonstrate the long-term positive impact of moral maturity on financial performance. For example, simulation firms made up of altruists earn more income than firms containing egoists. This outcome is corroborated by results from other simulations with multiple philosophies where moral maturity advances or declines over time. Under these circumstances, economic success improves from the most egoistic collective to the most altruistic grouping. Sensitivity analyses that manipulate payoffs and heuristics take place, and the results remain robust except under extraordinary circumstances that over-reward more egoistic behavior relative to more altruistic behavior, or hide selfishness from marketing partners for an extended and unacceptable length of time. Investigation of relationships between marketer moral maturity and buyer satisfaction/firm profitability in business-to-business contexts occurs in this study. With a basis in research on embedded morality, selling agents have an array of orientations including egoists, realists, loyalists, or altruists. This method also involves more complex and dynamic simulations of business-to-business exchanges that represent a significant modification beyond previous studies of agent-based computer models of artificial life. In this environment, buying agents seek selling agents to exchange with according to previous experiences and, in some circumstances, wordof-mouth. Market knowledge accumulates over time and reflects tolerance for variation in price offerings relative to actual delivered value. Morality plays an important role within buyer–seller exchanges in a number of contexts, but the area of pricing is most appropriate for the conduct of this research (Xia et al., 2004). Buyers are able to evaluate aspects of transaction quality as a result of marketers' pricing strategies, leading to subjective determinations of ethical and unethical treatments. Such evaluations result from comparisons to internal norms or external standards as judgment bases (Rallapalli et al., 2000). When value (quality-to-price ratio) is lower than expectations, sellers' reputations may suffer through the spread of negative information. High levels of price dispersion within the marketplace are necessary, allowing for enough variation to produce diverse outcomes. 2. Simulating marketing exchange relationships Usage of SUGARSCAPE versions of agent-based social simulations occurs because of its unique approach to the creation of interactive and dynamic environs (Epstein and Axtell, 1996). According to Agar (2005), such models test hypothetical situations rather than represent realworld systems, with the analysis capabilities of quantitative models. Epstein and Axtell (1996) employ SUGARSCAPE to build historical
A. Watkins, R.P. Hill / Journal of Business Research 64 (2011) 922–927
models of civilizations to understand effects on evolution of social phenomena. Subsequent uses include modeling communication and cooperation within animal societies (Buzing et al., 2003), estimating effects of taxation on evolution (Bäck et al., 2002), and studying assortments of cooperative behaviors among agents (Nishizaki et al., 2004). A similar philosophy advances agent-based computational economics (ACE) models, which involve computer science, cognitive science, and evolutionary economics (Vag, 2004). Within ACE models, agents learn from their actions and the actions of others, and modify individual behaviors (Tesfatsion, 2003). For example, Kirman and Vriend (2001) present an ACE model using two groups—buyers who learn loyalty results in lower prices and sellers who learn lower prices bring greater loyalty. In reinforcement learning, participants begin with simple rules about their environment and learn through a system of rewards and punishments which behaviors achieve the desired results. Their research provides a starting point for this study, which concentrates on a more complex model of reinforcement learning that involves multi-agent/environment interactions that synchronize and occur in discrete cycles. The orientations of selling agents capture extremes of moral maturity that are consistent with previous research (Hill and Watkins, 2007). Agents exist within single systems or industries to avoid contamination from extraneous factors or interactions, and each agent affiliates with one buyer or seller firm so that a set of players and their actions are easy to identify for purposes of relationship building. Seller firms contain 100 agents and two types of agent distributions were examined in this study; a uniform approach where agents within a firm had the same moral maturity or a mixed mode consisting of one dominant type and an even mix of the other types. Additionally buyers in the initial simulation did not directly communicate their experiences with one another but in the latter model communication occurred whenever buyers met. Agent Types on the selling side of exchanges include egoists who seek short-term financial gain for themselves with disregard for well-being of their trading partners; altruists who selflessly try to meet the desires of trading partners even if the result is negative impact on short-term profitability; realists who seek to maximize gain but do so with the guise of maintaining relationships with buyers; and loyalists who attempt to advance firm reputations among buyers even if the result negatively impacts personal success. Buyer firms also contain 100 agents in each, whose primary job is to find and consummate exchanges of good value. Sellers' ethical orientations guide price setting decisions. The base price of the simulated product is $750, and sellers can add markups up to $250. Sellers' transactional incomes are ½ base prices plus the full markup; therefore, one might expect more egoistic selling agents with a profit maximization goal to charge highest possible amounts, yet altruistic sellers will find that no markup brings greater personal satisfaction. However, as with human agents, an element of chance exists within the simulation that they might select a lower markup at any given point. Please note that the other orientations are more sensitive to buyer reactions to prices. Buyers move within the marketplace searching for sellers with whom they can have positive experiences. Although buyers find satisfaction when sellers make their offerings at base price, they do have a level of tolerance for markups. Buyers use a uniform system to rate transactions so that sellers get instant feedback on their actions. While real-world buyers may not have a uniform manner for such recognition, most individual appraisals are based on fair treatment and this is reflected in the model. Sellers receive ratings as feedback once exchanges are complete and all but the egoists use these to assist pricing decisions. Buyers, firms, and the industry transform transactional ratings into: (a) an individual buyer–seller relationship rating (B2S); (b) an average firm rating of the seller (F2S); and (c) an overall industry average rating of
Income (in millions) Amount
924
8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 2%
25%
50%
75%
100% 125% 150% 175% 200% 225%
% of Sellers Met by Buyers Egoists
Realists
Loyalists
Altruists
Fig. 1. Mean seller income (in millions) for uniform moral codes without communication.
the seller (I2S). Buyers use ratings and past data on positive/negative experiences to determine whether to complete a purchase. 2.1. Simulation results: uniform and mixed morality without communication Uniform moral maturity within seller firms and no internal or external word-of-mouth among buying agents defines the first baseline simulation model (Fig. 1). As a result, seller firms present a one-dimensional corporate culture, while buying agents are left to accumulated experiences to determine with whom they desire to transact. Buying agents are naive during initial exchanges; however, this scenario does not last long since they learn how to discriminate among selling agents. By the time experience reaches 225%, the point at which all buyers communicate with all sellers at least twice, income rises from egoists ($788 million), to realists ($3.3 billion) to loyalists ($5.5 billion) to altruists ($6.0 billion). Significant differences across seller firms occur (ANOVA: F(3,76) = 163.94, pb0.0001), and t-tests of mean differences are as well. Useful and counterintuitive findings concern intermediate levels of experience between 2% and 225%, as income and order of leadership change regularly during this interval. For instance at 25% (buying agents communicate with one-fourth of selling agents), egoists ($301 million) are in fourth place, altruists third ($482 million), loyalists second ($529 million), and realists first ($571 million). The ANOVA is significant (F(3,76) = 878.25, pb0.0001), as are t-tests. Another change happens at 125%, with egoists ($590 million) still least profitable and the remainder in an increasing order from realists ($2.4 billion), to altruists ($2.8 billion), to loyalists ($3.1 billion). Once again, ANOVA and t-tests are significant (F(3,76) = 198.25, pb0.0001). The capability of SUGARSCAPE to establish theoretical relationships between moral maturity and financial success within a business-to-business sales context occurs with the previous model. Nevertheless, this situation lacks diversity that typically exists within and across seller firms. A completely homogeneous corporate culture where every sales agent has the same moral sense is unlikely. Instead, firms may have a dominant approach that captures most of their employees along with minority perspectives that represent individuals outside the mainstream. Therefore, the model revision includes a majority of one type and a minority of the remaining three. This simulation contains 4 seller firms of 55 egoists, realists, loyalists, or altruists and 15 each of the other maturity levels. Similarities exist between uniform and the mixed morality. Incomes of seller firms at 2% penetration are identical; while the egoist dominant firm falls to fourth place and remains in this position for the rest of the simulation after the 25% rate. Surprisingly, a stable pattern of relative success occurs earlier at the 125% level, and this ordering is distinctive beyond this point in that the predominantly egoist firm ($1.5 billion) makes significantly less than the other three, but the realist ($2.5 billion) and altruist ($2.4 billion) firms earn about
2%
100%
225%
% of Sellers Met by Buyers Egoists
Realists
Loyalists
925
100%
150%
Altruists
Loyalists
Egoists
Realists
Altruists
Loyalists
Egoists
Realists
Altruists
Loyalists
Egoists
2%
Realists
Altruists
Realists
Loyalists
0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 Egoists
Altruists
Loyalists
Egoists
150%
Realists
Altruists
Loyalists
Egoists
Realists
Altruists
Loyalists
Egoists
Realists
Altruists
Loyalists
Egoists
Mean Amount Earned per Agent (in millions)
0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 Realists
Mean Amount Earned per Agent (in millions)
A. Watkins, R.P. Hill / Journal of Business Research 64 (2011) 922–927
225%
% of Sellers Met by Buyers
Altruists
Egoists
Realists
Loyalists
Altruists
Fig. 2. Mean individual seller income (in millions) at intervals for mixed moral codes without communication.
Fig. 4. Mean individual seller income (in millions) at intervals for mixed moral codes without communication.
the same amount while the loyalist firm ($2.6 billion) acquires more by a significant margin (F(3,76) = 213.51, pb0.0001). Finally, ultimate success of the loyalist dominant firm notwithstanding, income totals at 225% penetration supply some interesting findings compared to their complement. The primarily egoist firm earns about $2.4 billion, an increase of about 300% over the purely egoist firm, and the mainly realist firm ($4.3 billion) makes more than its parallel firm at 129%. Conversely, the mostly loyalist ($4.8 billion) and altruist ($4.5 billion) firms do less well in comparison at 88% and 75% respectively. Some basic similarities as well as some distinctive findings are the result of an attempt to provide a realistic representation of the role of morality in sales/financial achievement when compared to the uniform simulation. Early naivety gives way to discrimination powers that doom the primarily egoist firm to a distant fourth place in profitability. In comparison to their uniform complement, relative success that finally emerges over-rewards two least morally mature collectives relative to two more mature firms. The underlying cause is that some less morally mature selling agents are able to prosper in firms that present a mixed but morally advanced corporate culture, potentially masking self-centered intentions and allowing opportunities to transact as a consequence.
Given increasing speed experience accrues relative to previous models, powers of discrimination develop quickly. Egoists, realists, loyalists, and altruists earn the same amounts at low penetration rates within baseline data from the uniform scenario. However, by 25% egoists are at the bottom of the income hierarchy, earning half that of their closest competitor. By 75%, realists relegate to third position and a growth curve that is increasing at a decreasing rate. Finally, profitability of sellers relative to one another stabilizes at 150%, and the same pattern as the previous uniform run emerges as income amplifies from egoists ($348 million) to realists ($2.2 billion) to loyalists ($3.6 billion) to altruists ($4.1 billion) (F(3,76) = 1872.70, pb0.0001/all t-tests significant). Yet, while egoists fall into disrepute early, those who exist where information sharing takes place earn less than half (43%) as much as their no-sharing counterparts at 225% penetration. Further, both realists (78%) and loyalists (85%) are less profitable than their no-sharing complements, with only altruists (125%) gaining ground. This mixed simulation model also includes internal and external communication. While flip-flopping of profit leadership does occur across runs, the final pattern surfaces at the 100% level. Then, the mostly egoist firm ($1.1 billion) remains at the bottom, with the mainly realist firm ($1.9 billion) and altruist ($2.0 billion) and loyalist ($2.1 billion) dominant firms following. (ANOVA/F(3,76) = 1872.70, pb0.0001/t-tests significant between egoists and realists and the other two types.) Seller firms containing chiefly egoists ($2.5 billion) bring in 600% as much as their uniform with communication counterparts, while the principally realists firm produces 150% of their corresponding firm's income at 225% penetration. The loyalist and altruist dominant firms do less well in comparison at 100% and 65% respectively. One distinction is simulation results establish a stable pattern at earlier penetration, although the final result fails to fully discriminate among the stages of moral maturity. Another comparison involves findings from two mixed simulations. Sharing of information within and between seller firms has a modest dampening effect on total income by the end of the run for both egoist and realist oriented firms, with income falling slightly to 94% and 92% correspondingly. Also, the principally loyalist firm experiences a drop to 97% whereas the largely altruist firm expands to 108%. The sharing model causes realist, loyalist, and, to a large extent, altruist agents to improve average profitability as the primary orientation of firms shifts to the more mature in patterns similar to the no-sharing model. However, the altruists eventually stand out in the sharing circumstance, while loyalists tend to do so in the no-sharing situation. Addition of internal and external information sharing impacts results, marginally reducing penetration rate necessary to gain a stable marketplace and shift profit towards altruists. Nonetheless, clear demarcations according to level of moral maturity and selling agents' interest in firms' reputations with or well-being of exchange
2.2. Simulation results: uniform and mixed morality with communication
Income (in millions) Amount
Consistent moral maturity in contrast to mixed seller firms also takes place in the second set of simulations, but internal communication and external word-of-mouth are added to decision calculus of buying agents. When buying agents make contact, they provide each other with their storehouse of seller knowledge on agents that are unknown. Such data are amassed as agent-to-agent, agent-to-firm, and firm-to-firm. Buyer agents who are given access to the quality of exchange relationships of another agent encode data as if their own. 8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 2%
25%
50%
75%
100% 125% 150% 175% 200% 225%
% of Sellers Met by Buyers Egoists
Realists
Loyalists
Altruists
Fig. 3. Mean seller income (in millions) for uniform moral codes with communication.
926
A. Watkins, R.P. Hill / Journal of Business Research 64 (2011) 922–927
partners fails to materialize as they did in the uniform/sharing simulation. Thus, discrimination capabilities among buying agents seem less robust than they do under ideal conditions of onedimensional corporate cultures with widespread information exchange. In the end, buyers may allow sellers with selfish motivations to prosper, especially if they have the good fortune to reside in morally mature firms with largely selfless agents. 3. Discussion and conclusions The real value of models of ethical behavior flows from the ability to capture the complexity of moral decision making (Treviño and Brown, 2004). Thus, this simulation permits examination of the impact of agent moral character on exchange relationships within a business-to-business sales context. Selling agents transact according to assigned levels of moral maturity, which advances from the most selfish to the most selfless in the following order: egoists who maximize their gain from every transaction, realists who extract as much as possible without jeopardizing future opportunities, loyalists who place their firms' reputations above all else, and altruists who meet the needs of exchange partners as their first priority. Buyers respond to these offers from personal experience and word-of-mouth. The industry maintains four firms on each side of the exchange equation, with seller firms containing agents with uniform or mixed moral maturity who operate accordingly. The uniform scenarios provide validation for basic tenets of the artificial life model of living systems as a reproduction of ethical marketing theory. Specifically, buyers and sellers work in an environment where the possibility exists for long-term associations that have a basis in shared understanding of transactional ingredients. Final ordering of profitability meets the expectation that mature agents and their firms prosper, with significant differences in net income rising from firms with entirely egoists to realists to loyalists to altruists. Communication acts as catalyst that speeds identification of sellers by buyers (Fig. 2). While development of a suitable model mirroring this theoretical understanding is important, unexpected nuances often provide the most interesting findings. The inability of buying agents to fathom the intentions of selling agents during early stages of market interactions is clear, but the changes and differences in profit leadership among the levels of moral maturity at various penetration rates are an unexpected twist. Additionally, the reduced capacity of buyers to comprehend the same intentions when seller firms house agents with a variety of moral maturity levels is intuitive, but the speed with which stable patterns of interaction occur and the preferences that emerge suggest considerable marketplace distortion. Finally, ability of less moral sellers to prosper in mature cultures is in contrast to the belief that such corporate environments disallow this possibility (Treviño and Brown, 2004). Marketing managers and other practitioners have several potential takeaways. For instance, the common finding that buying agents lack discriminatory powers in searches for selling agents until the marketplace reaches a critical mass suggests the possibility of widespread manipulation. While certainty is impossible regarding moral makeup of prospective exchange partners until sufficient information and/or experience accumulates, buyers who are aware of this deficit can avoid the pitfalls by employing additional scrutiny and limiting the number of transactions until they are better able to judge seller intentions. This strategy must take into account selfimposed restrictions that lead to shortages in goods or services necessary for ultimate success (Fig. 3). Since discriminatory powers fail to work properly when operable, they lead to suboptimal choices by significant margins. This finding is troublesome given the smaller size and intricacy of the artificial environment and the extent and accuracy of available intelligence relative to real world contexts. Such results beg the question: How can buying agents protect themselves from overconfidence until the
situation truly merits loyalty to particular sellers? While the answer may seem trite, the best remedy involves avoiding premature decisions until a fuller measure of selling agents takes place. The criteria to ensure appropriate selections are made include a wide range of interactions and/or information that triangulates around the same conclusions. The third inference is from mixed morality models, with potential value for buyers and sellers. Preference patterns that arise from SUGARSCAPE simulations pose difficulties in comparison to uniform morality and its theoretical underpinnings. Thus managers in buyer firms, who are likely to encounter at least some heterogeneity within seller firms, must recognize limitations in agents' powers of perception and selection as decision-making guides (Fig. 4). As before, seeking assurance from a larger variety of sources and avoiding early and extensive commitments are the best strategy. Managers in seller firms, who attempt to follow moral directions of the relationship marketing philosophy, can improve chances of success through recruitment and promotion of consistently morally mature selling agents. References Agar. Michael H. Agents in living color: towards Emic Agent-Based Models. Journal of Artificial Societies and Social Simulation 2005;8(1) http://jasss.soc.surrey.ac. uk/8/1/4.html. Armstrong Scott J. Social irresponsibility in management. Journal of Business Research 1977;5:185–213 (September). Ashkanasy Neal M, Windsor Carolyn A, Trevino Linda K. Bad apples in bad barrels revisited: cognitive moral development, just world beliefs, rewards, and ethical decision making. Business Ethics Quarterly 2007;16(4):448–73. Bäck Thomas, Vermeulen Daniel, Eiben Agoston E. Effects of tax and evolution in an artificial society. In: Caulfield HJ, Chen S-H, Cheng HD, Duro R, Honavar V, Kerre EE, Lu M, Romay MG, Shih TK, Ventura D, Wang P, Yang Y, editors. Proceedings of The Sixth Joint Conference on Information Sciences, (JCIS 2002); 2002. p. 1151–6. Bos Nathan D, Sadat Shami N, Sara Naab. A globalization simulation to teach corporate social responsibility: design features and analysis of student reasoning. Simulation & Gaming 2006;37:56–72 (March). Brinkmann Johannes. Business and marketing ethics: concepts, approaches and typologies. Journal of Business Ethics 2002;41:159–77 (November/December). Buzing Pieter, Agoston Eiber, Schut Martjin C. Evolving agent societies with VUScape. Lecture Notes in Computer Science, 2801. Springer Berlin; 2003. p. 434–41. Cherry John, Fraedrich John. Perceived risk, moral philosophy and marketing ethics: mediating influences on sales managers' ethical decision- making. Journal of Business Research 2002;55:951–62 (December). Desmond John, Andrew Crane. Morality and the consequences of marketing action. Journal of Business Research 2004;57:1222–30 (November). Epstein Joshua M, Axtell Robert. Growing artificial societies: social science from the bottom up. Cambridge, MA: MIT Press; 1996. Falkenberg Andreas Wyller. When in Rome … moral maturity and ethics for international economic organizations. Journal of Business Ethics 2004;54:17–32. Gick Evelyn. Cognitive theory and moral behavior: the contribution of F. A. Hayek to business ethics. Journal of Business Ethics 2003;45:149–65 (June). Goolsby. Jerry R, Hunt Shelby D. Cognitive moral development and marketing. Journal of Marketing 1992;56:55–68 (January). Gotts NM, Polhill JG, Law ANR. Agent-based simulation in the study of social dilemmas. Artificial Intelligence Review 2003;19:3-92. Hill Ronald Paul, Watkins Alison. A simulation of moral behavior within marketing exchange relationships. Journal of the Academy of Marketing Science 2007;35(2): 417–29. Hunt Shelby D, Vitell Scott J. The general theory of marketing ethics: a revision and three questions. Journal of Macromarketing 2006;26:143–53 (December). Kirman Alan P, Vriend Nicolaas J. Evolving market structure: an ACE model of price dispersion and loyalty. Journal of Economic Dynamics & Control 2001;25:459–502. Kjonstad Bjorn, Willmott Hugh. Business ethics: restrictive or empowering? Journal of Business Ethics 1995;14(6):445–64. Klein Sherwin. The natural roots of capitalism and its virtues and values. Journal of Business Ethics 2003;45:387–401 (July). Kohlberg Lawrence. The psychology of moral development: the nature and validity of moral stages. San Francisco: Harper & Row; 1984. Monga Manjit. Manager's moral reasoning: evidence from large Indian manufacturing organizations. Journal of Business Ethics 2007;71:179–94. Murphy Patrick E, Laczniak Gene R, Wood Graham. An ethical basis for relationship marketing: a virtue ethics perspective. European Journal of Marketing 2007;41(1/2): 37–57. Nishizaki Ichiro, Sakawa Masatoshi, Katagiri Hideki. Influence of environmental changes on cooperative behavior in the prisoner's dilemma game on an artificial social model. Applied Artificial Intelligence 2004;18:651–71. Rallapalli C, Vitell Scott J, Sheryl Szeinbech. Marketers' norms and personal values: an empirical study of marketing professionals. Journal of Business Ethics 2000;24: 65–75 (March).
A. Watkins, R.P. Hill / Journal of Business Research 64 (2011) 922–927 Robin Donald P, Gordon Gus, Jordan Charles, Eric Reidenbach R. The empirical performance of cognitive moral development in predicting moral intent. Business Ethics Quarterly 1996;6(4):493–515. Schroeder Donna J. Morality de-kanted or the biological roots of moral behavior. International Journal of Values-Based Management 2000;13:297–308 (January). Singhapakdi Anusom, Vitell Scott J, Kraft Kenneth L. Moral intensity and ethical decision-making of marketing professionals. Journal of Business Research 1996;36: 245–55 (July). Sullins John P. Ethics and artificial life: from modeling to moral agents. Ethics and Information Technology 2005;7:139–48. Teach Richard D, Christensen Sandra L, Schwartz Robert G. Teaching business ethics: integrity. Simulation & Gaming 2005;36:383–8. Tesfatsion Leigh. Agent-based computational economics: modeling economies as complex adaptive systems. Information Sciences 2003;149:263–9. Trevino Linda K. Moral reasoning and business ethics: implications for research, education, and management. Journal of Business Ethics 1992;11:445–59. Treviño Linda K, Brown Michael E. Managing to be ethical: debunking five business ethics myths. Academy of Management Executive 2004;18(2):69–81.
927
Tuomela Maj, Hofmann Solveig. Simulating rational social normative trust, predictive trust, and predictive reliance between agents. Ethics and Information Technology 2003;5(3):163–76. Vag Andras. First generation multi-agent models and their upgrades. Interdisciplinary Description of Complex Systems 2004;2(1):95-103. Valentine Sean, Barnett Tim. Ethics codes and sales professionals' perceptions of their organizations ethical values. Journal of Business Ethics 2002;40:191–200 (October). Walker Lawrence J, Pitts Russell C. Naturalistic conceptions of moral maturity. Developmental Psychology 1998;34(3):403–19. Watkins Alison, Ronald Paul Hill. The impact of personal and organizational moral philosophies on marketing exchange relationships: a simulation using the prisoner's dilemma game. Journal of Business Ethics 2005;62:253–65 (December). Werhane Patricia H. A place for philosophers in applied ethics and the role of moral reasoning in moral imagination: a response to Richard Rorty. Business Ethics Quarterly 2006;16(3):401–8. Xia Lan, Monroe Kent B, Cox Jennifer L. The price is unfair! a conceptual framework of price fairness perceptions. Journal of Marketing 2004;68:1-15 October.