The ethics code for behavior analysts describes the professional standards set forth by the field to protect consumers of behavioral services and practitioners of Applied Behavior Analysis. By acting in compliance with these standards, behavior analysts demonstrate their commitment to the principles of beneficence, respect, integrity, and competence, and embody the goal of applied behavior analysis (ABA), i.e., to produce meaningful changes in the quality of life of individuals and communities (Baer et al., 1968; Schwartz & Baer, 1991; Slocum et al., 2014; Wolf, 1978).

The development of this code of ethics, however, followed years of unregulated practice characterized by poor application of behavioral principles, unnecessary use of aversive control procedures by untrained staff, and overall misuse of behavioral techniques (Bailey & Burch, 2016; Friman, 2021). Unfortunately, these events reaffirmed popular mischaracterizations of behavior analytic practices as restrictive, manipulative, cruel, or dehumanizing (Critchfield & Reed, 2017; Foxx, 1996; Friman, 2021; Poling, 2010).

To this day, despite the development of high quality standards for practitioners, many of these negative views prevail (Leaf et al., 2021). Social media posts, reports by autistic individuals who have received behavior analytic services, and even peer-reviewed publications continue to describe ABA interventions as aversive, inhumane, and traumatic for clients and families (e.g., Kupferstein, 2018; McGill & Robinson, 2020; Sandoval-Norton et al., 2021; Sandoval-Norton & Shkedy, 2019). These reports generally portray ABA as a unified set of practices that disregards the histories, preferences, and needs of clients and stakeholders (Leaf et al., 2021). These perceptions and claims are contrary to the goals of ABA and may indicate, among other things, problems of adherence to the ethical standards of the field.

In response to these claims, recent publications have underlined the importance of establishing high-quality therapeutic relationships for treatment effectiveness and provide insightful discussions regarding the impact of consumer satisfaction on the success of our field (e.g., Callahan et al., 2019; Rohrer et al., 2021; Taylor et al., 2019). These efforts serve as contemporary reminders of Wolf’s (1978) critical points about social validity: (a) that it is consumers who ultimately validate goals, interventions, and outcomes, and (b) that the primary goal of a science of behavior is to contribute to the well-being of others.

Nonetheless, despite familiarity with the ethical standards of the profession, most practitioners are likely to know at least one individual or organization who has failed to behave ethically. In our view, the fact that unethical behavior continues to occur suggests that behaving ethically requires more than learning the rules, knowing the ethics code, and the threat of contacting contingencies for noncompliance. This becomes apparent when ethical behavior is analyzed functionally, i.e., in terms of its controlling variables. In this paper, we suggest a functional analysis of ethical behavior based on empirical findings from choice, delay, and social discounting research. Based on this analysis, we discuss strategies for cultural design that may favor the selection of ethical practices in organizations and communities.

Ineffective Contingencies: Rule Governance

During coursework required for certification, behavior analysts are trained to identify ethical violations, suggest a principled course of action, and demonstrate an understanding of the applicable standard of conduct in a variety of complex hypothetical situations. However, what a person does in the presence of direct contingencies is likely to differ from what they say they would do when merely exposed to descriptions of such contingencies, likely, due to corresponding differences in the controlling variables. In other words, the variables of which verbal behavior about ethical dilemmas is a function are not the same as the variables that control behavior in the presence of an actual ethical dilemma.

Perhaps one factor explaining this is that rules exert weak control over behavior when the consequences for behaving unethically are delayed or improbable, which is usually the case with moral or ethical rules (Malott, 2001). Similarly, rules that describe macrocontingencies are often ineffective because the consequence they specify is produced over time by the action of many individuals (Glenn et al., 2016). For example, the negative impact on the field produced by the actions of some practitioners is not likely to deter an individual from behaving unethically if doing so is immediately convenient and unlikely to be detected by others. To supplement ineffective control exerted by ethical rules, sanctions for lack of compliance are usually added (Malott, 2001; Skinner, 1974). However, the effectiveness of methods of aversive control for evoking ethical conduct is also limited, as unethical behavior under aversive control is likely to be avoided only in the presence of a potential punisher (Skinner, 1971, 1974).

Further, unreported ethical violations pose a problem because the person engaging in the unethical behavior is likely to do it again. Due to ineffective rule control, familiarity with ethical standards and sanctions for violating those standards may not be sufficient to evoke behavior that benefits others, particularly in the presence of competing contingencies.

While some rules exert weak control over behavior, others influence behavior long after the conditions requiring their use have changed, and even when adverse consequences follow (Hayes et al., 1986). An example is the continued use of procedures (i.e., extinction) that have lost social validity and for which other alternatives have gained empirical support. The use of extinction has not only been criticized by neurodiversity experts; a growing body of ABA research has shown that differential reinforcement without extinction is capable of achieving the same results (Athens & Vollmer, 2010; Briggs et al., 2019; Kunnavatana et al., 2018; Vollmer et al., 2020) with similar or even lower rates of resurgence (Brown et al., 2020). However, the over-extended rule that effective behavioral treatment should outweigh sentiment may prevent practitioners from contacting a new set of rules and related direct contingencies.

Further, insensitivity of rule-governed behavior to contingency changes often occurs due to relational responding and derived rule-following. If an individual responds to a rule stating that strict and rigorous practices are effective, and if “compassionate” and “empathic” are in a frame of opposition with respect to “effective” and “rigorous,” this individual may fail to acknowledge evidence supporting the effectiveness of more compassionate and socially valid practices. Thus, as the product of a system of interlocking verbal relations, ideology will sometimes aid and at other times stall social progress toward ethical practice. If ineffective rules are one barrier to ethical conduct, over-extended rules, are another such barrier.

Effective Contingencies Exerting Control Over Ethical Behavior: The Choice to Behave Ethically

When ethical behavior is under the control of competing contingencies, a person can either do what is more convenient and less effortful at the moment or what is in the client’s best interest, the organization, and the field. Thus, most instances of ethical behavior may be described as a choice between behaving for the good of others and behaving selfishly (Borba et al., 2014; Malott, 2001; Rachlin, 2002; Rachlin & Jones, 2008; Skinner, 1987). Billing for services that were not provided, fabricating data, or working outside one’s scope of competence are examples of actions that provide advantages to the individual, whereas the alternatives would produce delayed and less certain gains. In these situations why would an individual prioritize others’ interests over their own?

The delay and social discounting literatures provide empirical evidence that may contribute to answering this question (Locey & Buddiga, 2022). On one hand, research in delay discounting shows that individuals tend to discount the value of outcomes that are delayed in time (DeHart & Odum, 2015; Lattal, 2010; Odum, 2011; Odum et al., 2020). People who are described as impulsive show steeper discounting curves; that is to say, they tend to choose a smaller immediate rather than a larger but delayed reward. Importantly though, the effectiveness of procedures designed to shape less impulsive and selfish choices demonstrates that these choices are the result of learning histories. One such procedure involves giving participants a choice between a smaller and a larger reward and providing immediate access to the larger reward when chosen. Gradually adding increased delays to the larger reward has shown to maintain self-controlled choices in individuals with a history of impulsive choices (e.g., Binder et al., 2000; Dixon et al., 1998; Dixon & Holcomb, 2000).

Social discounting on the other hand involves examining the extent to which people are willing to share resources or give them in full to another person. In the traditional social discounting procedure, participants are asked to make an imaginary list of people, ranking them by degree of social distance. Thus, the person closest to the participant would be first on their list, and a mere acquaintance would be last. Participants are then asked to make a series of choices that involve keeping an amount of money for themselves or giving a fixed amount to one of the individuals on the list (Jones, 2021; Jones & Rachlin, 2006; Rachlin & Jones, 2008). Findings from these studies consistently show that the likelihood of forgoing a reward decreases as a function of social distance; that is, more selfish choices tend to be made as social distance increases. Note that whereas in delay discounting, the choice is between a small immediate and a larger delayed reward, in social discounting, the choice is between keeping a reward for oneself or giving it to others. Because behaving ethically is the result of a choice in a social context – a choice that affects others – this social component of choice behavior may be essential to understanding and predicting self-controlled and selfless (i.e., ethical) choices (Locey & Buddiga, 2022).

Ethical behavior may thus be conceptualized as forgoing individual rewards to behave for the good of others when doing so produces delayed and long-term reinforcers for all involved. This circumstance is particularly well recreated in experimental preparations using the prisoner’s dilemma game (Borba et al., 2014; Locey et al., 2013; Locey & Buddiga, 2022; Morford & Cihon, 2013; Yi & Rachlin, 2004). Variations of this procedure involve groups of participants playing a game in which points can be earned over iterated opportunities to make choices (e.g., Locey & Rachlin, 2012). As in the traditional prisoner’s dilemma game, in these experimental preparations, the highest benefit is obtained for behaving selfishly. However, if most participants behave selfishly, everyone in the group is affected negatively, while if all cooperate, everyone benefits in the long run (Borba et al., 2014; Locey & Rachlin, 2012; Rachlin, 2002; Toledo & Avila, 2021; Yi & Rachlin, 2004).

Findings from these studies may have important implications for arranging contingencies that favor ethical behavior (Borba et al., 2014; Locey & Rachlin, 2012; Rachlin, 2002). When a person engages in unethical behavior, they are choosing between two response alternatives: one that results in immediate positive consequences for the individual and negative consequences for others, and a second one, that results in larger benefits for the community and smaller or less immediate benefits for the individual. When practitioners choose the selfish option, the communities involved (the practitioners, the agency, the field, the community) are negatively impacted in the long run. Notably, however, experimental research findings have consistently shown that cooperative behavior or ethical self-control can be shaped over repeated iterations of contrived choice-making situations (Borba et al., 2014; Locey et al., 2013; Locey & Rachlin, 2012; Morford & Cihon, 2013; Toledo & Avila, 2021). Given this set of contingencies, individuals may learn over time that benefits are maximized when everyone chooses the good of all (Morford & Cihon, 2013; Rachlin, 2002). A repertoire of generalized ethical choice-making may be selected if obtaining the best outcome for behaving for the good of all is a contingency that is available, frequent, and reliably contacted.

Several related research studies have described more specific circumstances that may facilitate behaving for the good of others. For example, Locey and Rachlin (2012), and Locey et al. (2013), found that increasing the magnitude of rewards for cooperating correlates with larger numbers of people in a group who will cooperate. However, when the reward remains the same, the rate of cooperation tends to decrease as the number of participants in a group increases (Morford & Cihon, 2013; Yi & Rachlin, 2004). Further, Borba et al. (2014) showed that impulsive choices are more likely when group members do not have access to each other’s choices and are unable to talk among themselves, whereas cooperative choices are more likely when participants are allowed to talk about their choices. Similarly, findings from the social psychology literature show that decreased self-control, increased likelihood of aggression, and decreased prosociality are related to social exclusion and social distance (Zhang et al., 2019).

These findings are consistent with the predictions from social discounting and seem to support the role of trust in effective group interactions. In a group, an individual is more likely to behave selfishly if they predict that others are also likely to behave selfishly (De Cremer & Stouten, 2003; Irwin & Berigan, 2013) and if there is a low probability of reciprocation (Locey & Buddiga, 2022). Behaving for the good of all becomes unlikely if an individual predicts that everyone in their group will act to pursue their interest only (Atkins et al., 2019).

Taken together, these findings suggest that in most cases behaving ethically may require a history of reinforcement for making choices that benefit others and for establishing meaningful, relations with team members, consumers, and with the communities in which they live. People are likely to make the choice that is followed by delayed positive consequences for themselves and others if they have learned to wait for larger delayed rewards, and if those others are perceived as less socially distant. A practitioner who is often at their desk analyzing data and writing behavior plans without having frequent, meaningful interactions with the client and/or their families or guardians, may be unlikely to forgo an immediate gain to favor the best interest of their clients. Further, when most members of an organization behave selfishly (as a result of predicting that others will do so as well), behaving for the good of all is not likely to pay off.

Establishing close social interactions and selecting meaningful but delayed consequences that benefit all are repertoires that can be shaped using behavior analytic interventions. Perhaps in addition to merely learning the rules for ethical behavior, what we need is to intentionally arrange contingencies that will sustain such behaviors. Translational studies evaluating the extent to which the functional relations observed under controlled laboratory conditions are replicated in the complexity of organizational settings will be needed to empirically support this approach. Nonetheless, a compelling description of the functional relations that may account for ethical behavior may be drawn from the cogency and consistency of the findings from the EAB literature. Further, because pathways toward change seem to inevitably follow description and predictability of lawful relations, a discussion of applied implications seems warranted. The following section discusses some strategies for behavior change that correspond to an analysis of ethical behavior in terms of an acquired repertoire of self-controlled choices shaped by group outcomes.

The Design of an Ethical Culture

In a culture that favors selfish choices, knowledge of ethical rules or the threat of sanctions for non-adherence to these rules may not be sufficient to produce behavior change. As Skinner (1948) discussed, engineering environments that support self-controlled ethical choices may be the answer. The challenge for the behavior analyst is to design such a culture based on the empirical findings of the circumstances under which individuals make ethical choices.

The effect of functional relations that operate in larger time spans is usually seen much after their effects are apparent, and any measures to reverse those effects or to produce new ones require significant time and effort. For this reason, education, information, and awareness are inadequate measures to produce a sufficient impact on ethical behavior, especially if prevailing contingencies that maintain behavior differ from those specified by verbal descriptions. The solution to changing behavior in these situations was elegantly proposed by Skinner almost 70 years ago: We need to engineer environments based on the scientific evidence of the circumstances that impede and facilitate doing what is in the best interest of the group (Skinner, 1948). In the absence of such an intentionally designed culture, prevailing institutional contingencies will continue to select selfish choices (Skinner, 1987).

Training people to behave ethically by discussing individual occurrences may be impractical, however, because the frequency of reports or opportunities to make selfish or unselfish choices may not be controlled. Hypothetical situations, on the other hand, may not be accurate predictors of choices in the presence of direct contingencies. To shape and maintain individual ethical behavior, programming contingencies at the group and individual levels that limit the recurrence of selfish behaviors is necessary (Atkins et al., 2019). Group-based strategies need to consider the control exerted by competing individual contingencies and incorporate measures to outweigh those effects. More specifically, the behavior of each member of the organization or team needs to come under the effective control of a group contingency that, compared to the competing individual contingencies, produces higher benefits for all members in the group. Thus, in a situation where individual and immediate reinforcers compete against reinforcers that benefit all, the question of why an individual would prioritize the benefit of all can be easily answered: the benefit of the group systematically results in the highest benefit for the individual. Further, this outcome tends to be more likely when group members are trusted and perceived as socially close. This answer albeit not new, is empirically now supported by a wealth of research findings.

When conceptualizing change at the group level, it is crucial to consider how individual behavior may contribute to the functioning of the whole system. An alternative to increase ethical behavior in a group is to use self-control training procedures such as progressive delay, in the context of macro or metacontingencies. For example, cooperative behaviors could be shaped in small groups by contriving reinforced opportunities to behave for the good of all while forgoing immediate individual benefits. While initially making this choice is immediately reinforced, delays to reinforcement upon choosing the behavior that benefits the group are gradually introduced. Conditioned reinforcers may also be used as delays to reinforcers increase. In this strategy, the natural reinforcing contingency maintaining self-controlled ethical choices could be a product that requires the coordinated participation of its members as in a metacontingency, or a cumulative effect resulting from individual choices as in a macrocontingency. Eventually, repeated exposures to group contingencies with consequences that benefit everyone in the group will contribute to building a naturally maintained and generalized cooperation repertoire.

To illustrate with an example, small teams may be told that if they volunteer an extra hour after work to clean up and put together work materials for clients’ sessions, they could earn incentives such as parking privileges or free lunch at the end of the month. If all team members chose to stay the extra hour, they would be given the reward immediately. In future opportunities, contrived rewards for choosing to cooperate can be slowly and gradually delayed as natural contingencies (e.g., social recognition, promotions, clients’ progress) are more frequently contacted.

Variations of the iterated PDG procedure could also be used as exercises to help team members contact contingencies that benefit all and facilitating the selection of ethical choices. As discussed, cooperation is a more probable outcome in smaller groups because establishing relations with less perceived social distance is more likely and because behavioral outcomes are more clearly linked to individual behavior. The goal of these group-level approaches is to increase equitable interactions within small self-directed teams by programming contingencies for acting in the best interest of the group. Once this is achieved, higher-level selection contingencies may then be implemented whereby each small team, conceptualized as a unit in its own right, cooperates with other teams to seek the benefit of the organization (Atkins et al., 2019).

In our view, an integral part of the design of an ethical culture is involving the community. This entails viewing clients and stakeholders as valuable group members who contribute to the success of the organization and not merely as the recipients of a service and external to the group. For example, a behavior analyst may program reinforcement for cooperative behaviors among a group consisting of a pair of BCBAs, the client, and a parent. The resulting increase in behaviors that serve the benefit of all will inevitably start contacting natural reinforcers and modeling cooperation in others. For example, Szabo (in press) describes the use of an expedition narrative tailored to meet treatment goals for an individual receiving services and cooperation skill attainment goals for members of the treatment team. This narrative would describe the BCBAs, the client, and parents, as Hobbits of Middle-Earth traveling though rich and magical forests to find Smaug, the dragon that stole their ring. Along the way, the Hobbits become entangled in an argument that no one can win. Somehow, they must work together to become like the ring they seek to retrieve – they must become an unbroken, untangled ring of Hobbits. The narrative is exciting to young members and metaphorically evocative for parents and professionals in the group. The group members are asked to form a circle, outstretch their left hands, and grab hold of the hand of someone not immediately to their left or right. They are then told to do the same with their right, after which they are instructed to work collaboratively to disentangle themselves without breaking hands; that is, they must form a perfect ring without losing their familial Hobbit connection. A valuable extension of this activity is to invite those who typically lead in the group (e.g., BCBAs) to follow others’ instructions, and those who commonly follow the rules supplied by others (e.g., children) to try and lead the group. Upon successfully disentangling, the group earns a pastry ring and unlocks the next level of the expedition. This interdependent contingency and expedition narrative serves to promote cooperation. The activity is physically and intellectually challenging enough to engender laughter, confusion, frustration, and innovative behavior. The resulting contact with natural reinforcers for cooperative effort immediately serves to springboard discussion of the benefits of acting for the good of all.

In everyday practice, examples of cooperative behaviors that benefit each individual and the group as a whole could include establishing treatment goals, defining the behaviors required of each person in the group to attain those goals, self-recording occurrences of these behaviors, meeting with the group to share progress data, and troubleshooting barriers to success. Self-selected and/or peer-delivered rewards may be contingent on engaging in the agreed upon behaviors. The behavior analyst could lead group discussions reflecting on what would happen if one person failed to follow through and highlighting the natural contingencies contacted when each member does their part.

An increase in behaviors that serve the benefit of all will inevitably start contacting natural reinforcers and modeling cooperation in others. Other likely effects may include increased treatment adherence, increased accountability, and trust among group members. In sum, building meaningful relationships with families is not only a good idea to establish rapport, buy-in, and to enhance treatment effectiveness it also increases the likelihood that practitioners will choose to behave in the best interest of the client.

Another strategy to increase self-controlled ethical choices may involve augmenting the value of delayed benefits or bringing them to the present by way of function transformation of verbal or other stimuli (Locey & Buddiga, 2022). For example, Castro et al. (2016) enhanced engagement in direct care workers by asking them first to identify the things they valued when working with their clients and then to write them on a paper T-shirt. Incorporating systems of peer-delivered reinforcement or systems of self-recording may also add to the effectiveness of this intervention. For example, once organizations or teams have identified common values such as caring, making a difference, or showing kindness, they could arrange for members to frequently contact and commit to these values by printing them on a T-shirt and have peers tally and reinforce each other for behaving in line with those values. Counting greetings, smiles, high-fives, or occurrences of verbal praise is a simple yet effective approach that targets the quality of the interactions between members of small groups and brings to the present natural reinforcers for value-oriented actions.

A group-level approach, however, does not preclude setting up contingencies at the level of the individual. These could be embedded within and supplement group contingencies. Programming frequent reinforcement for making ethical choices may have limited effectiveness because opportunities of observing practitioners making un/ethical choices are not controlled. However, as a practitioner would do with a client, a supervisor or a team leader may want to catch group members engaging in ethical behavior. Graphing data daily and accurately, sharing data with consumers, having a positive interaction with parents or guardians, referring out a case that is outside the scope of competence of the behavior analyst, protecting the confidentiality of clients, etc., may be often overlooked as part of the daily routine of practitioners; nonetheless, they are also opportunities to reinforce actions that benefit others. One way to do so in a way that is behavior specific, measurable, and conspicuous is collecting social validity data and publicly sharing the extent to which consumers are satisfied with the specific services provided.

Importantly, when instances of unethical behavior are detected and measures have been taken to protect the safety and the best interest of the client, supervisors should be mindful of responding as a non-punishing audience. Doing so will require pausing before responding and switching from a personal view (i.e., blaming) to a behavioral view that involves understanding the behavior of the practitioner and identifying its controlling variables (Friman, 2021). This entails acknowledging that choosing the selfish option does not mean that the person has bad intentions or is immoral or unethical, but simply that their environment was not conducive to making an ethical choice and that a stronger ethical repertoire needs to be shaped. Behavior analysts can then identify and address the contingencies controlling the selfish choice, which may involve aversive EOs, cultural biases, or simply a lack of opportunities to contact the reinforcing effect of behaving for the good of others.

Leading a group of BCBAs or RBTs involves designing cultural contingencies with the awareness, humility, and compassion that inevitably comes from taking the circumstances view (Friman, 2021). This requires moving away from habitual practices of rule governance, enforced by contingences of aversive control, self-checking to prevent attributing personal responsibility for actions, and viewing actions as the result of contingencies experienced by the practitioner and arranged by the supervisor. Simply put, it requires acting like a behavior analyst. Doing so will most likely transform a culture of selfish choices into an ethical culture.

Summary and Conclusion

Rule familiarity, even with added punishment contingencies for non-compliance, is not likely to affect an individual’s pattern of culturally reinforced selfish choices. An alternative is to analyze ethical behavior functionally, that is, in terms of a choice response and its controlling variables. As such, most instances of ethical behavior can be described as a type of cooperative behavior that requires self-control. Understanding ethical behavior in this way allows the behavior analyst to identify the variables that predict selfish and cooperative choices and manipulate those variables to favor increased occurrences of the latter.

Unfortunately, our cultural environment has arranged contingencies that predominantly reinforce selfish behavior (Skinner, 1986, 1987). In most cultural groups (schools, families, organizations, and larger communities), contacting the benefits of behaving for the good of all is prevented by immediate reinforcement for behaving selfishly. Further, in a cultural environment where group members predict that others (including group leaders) will act to protect their interests, behaving cooperatively will be recurrently punished. Culturally we have learned to live fast, behave fast, and expect fast outcomes (Skinner, 1986); but the selection of practices and groups who behave for the good of all is only evident through their cumulative effects over time (as it occurs in macro-contingencies). Equally, the negative consequences of unethical practices slowly but inevitably impact agencies, organizations, and the professional community.

Nonetheless, ethical self-control can be learned and maintained as people experience the more significant benefits it produces over time. This will not happen, however, unless behavior analysts program contingencies specifically designed to select ethical self-controlled choices and arrange frequent opportunities to contact these contingencies. This view, which Skinner instilled as early as 1948 and throughout his career, has been widely supported by recent empirical studies in the experimental analysis of human behavior. These studies systematically show that acting for the good of the group is observed when experimental arrangements produce the benefit of all as a reliable and frequent outcome that is more advantageous over repeated iterations than the selfish choice.

No amount of training in our ethical standards may be enough to change the perception of consumers who have been negatively impacted by unethical practices. All it takes is for service providers and leaders of organizations to be cultural designers. The design of a culture sustaining ethical-self-controlled choices may involve components such as: arranging group contingencies, training others to identify and observe instances of ethical behavior, reinforcing ethical behavior on a dense schedule, modeling making self-controlled choices, identifying and creating shared values within small work teams, building in self-management strategies and peer-administered consequences, using rules to augment the value of delayed consequences, avoiding aversive control and punitive consequences, implementing data collection systems to track progress and to prompt implementation of program components, and remaining present to the guiding value shared by the ABA community: to produce behavior changes that improve the quality of life of others.

Although examples of group contingencies with recurrent benefits for engaging in self-controlled choices were suggested, applied systematic research evaluating the effectiveness of cooperation strategies in organizations is needed. Future research may focus, for example, on examining changes in client’s progress and parental adherence when service providers and recipients of behavioral services work as a consolidated group to identify values, establish what it takes from each member to commit to those values, outline the costs of behaving selfishly, and use data collection, monitoring, and accountability systems. Using the Prosocial matrix (Atkins et al., 2019) to self-record and map occurrences of selfish covert and overt behaviors that are not in line with the benefits of all may be another strategy to support sustainable systems of cooperation that requires further systematic research. Pilot data from the implementation of interdependent contingencies in organizations such as those described in other sections would also provide empirical support for the present analysis and further the development of technologies of cultural design.

In sum, a program of cultural change should be informed by an analysis of group-level contingencies that requires: (a) understanding that as a type of self-control, behaving ethically involves the shaping of a behavioral repertoire in a group setting, (b) acknowledging individual competing contingencies and cultural learning histories, and (c) targeting change in smaller groups, communities or organizations; as increasing connectedness and familiarity between group members is likely to decrease selfish (unethical) choices in the presence of competing contingencies.

Ethical choices may not be prevalent unless cultural change is programmed within organizations. We need to engineer environments where people choose to do what benefits others, not because they are punished if they do not follow the rules but because they have reliably contacted the value of doing so. However, because we have built cultures that typically operate under a different set of contingencies this will not happen naturally or effortlessly. Achieving this end will take more than becoming familiar with the ethics code. It will take a commitment to work on cultural design. A wealth of data supporting effective technologies to teach others (employees, children, clients, communities) to behave in ways that are meaningful for all is within our grasp. Using this knowledge to shape behavior analytic practice and its impact on the community is our challenge and responsibility. Unethical behavior is a concerning issue that requires evaluation, prevention, and action. As team and organization leaders, behavior analysts are responsible for engineering environments that are unequivocally characterized as ethical cultures.