We appreciate the work of Lefkowitz (Reference Lefkowitz2021) to encourage more reflection regarding the structural forms of ethical dilemmas faced by those in industrial-organizational (I-O) psychology. As professionals, there are few situations where decisions are entirely free of ethical ambiguity or negative consequences. This notion is made particularly salient by the focal study’s findings that highlight the depth and breadth of ethical dilemmas, defined as a choice between unattractive alternatives that the actor “does not want to make” (Lefkowitz, Reference Lefkowitz2021), faced by Society for Industrial-Organizational Psychology (SIOP) members. Given this definition, we extend the focal study’s contribution by using the principle of double effect (PDE), a decision-making approach that serves to guide actors on the moral permissibility of their actions, to argue that I-O practitioners would recognize more situations as presenting a genuine ethical dilemma if they evaluated unintended, but reasonably foreseeable harms that may result from their decisions.
In addition to providing a working definition of ethical dilemmas, Lefkowitz (Reference Lefkowitz2021) sees these dilemmas as made up of three parts: the necessity of making a decision, the need to apply some form of moral/ethical principle to this decision, and the reality that the decision will have significant consequences for others. Moreover, an implicit assumption of ethical dilemmas is the belief that individuals who are faced with these situations are motivated to do good and have no desire to cause harm.
Although we agree with the definition of ethical dilemma brought forward by the author, we suspect that this definition is somewhat conservative and may underestimate the extent to which practitioners experience ethical issues on a daily basis. For example, it may be possible that practitioners make a particular decision without fully thinking through the full spectrum of consequences of their actions, particularly as: “Professional ethics is a rarely considered subject in industrial-organizational (I-O) psychology, which might imply that it is not of much concern to the field” (Lefkowitz, Reference Lefkowitz2021, p. 297). Given this observation and notion that the structural forms of ethical awareness rely on awareness and intention (e.g., opportunity to prevent harm), it makes sense that practitioners may be unaware or have not fully considered the unanticipated, albeit reasonably foreseeable, consequences of their actions and, therefore, do not initially classify their actions as constituting an ethical dilemma. This limits the scale of the problem and our ability fully to realize the extent to which ethical dilemmas affect I-O practitioners.
As a remedy, we propose that, in addition to observing the taxonomy provided by Lefkowitz (Reference Lefkowitz2021), practitioners adopt PDE as part of a more systematic and detailed decision-making approach to help actors evaluate decisions that may reasonably lead to or contribute in some way to negative outcomes for others. This, we argue, augments our understanding and classification of ethical dilemmas in I-O psychology, as it expands the conversation to include both the context of the choice and the unintended, but reasonably foreseeable, effects of decisions on others. Thus, if the situation has not yet been determined to be an ethical dilemma according to the strictures set forth by the focal article, PDE encourages practitioners to exceed these notions by encouraging them to consider the foreseeable, yet formerly unforeseen, effects of their decisions. It also encourages actors to consider the potential harms that may be felt by a wider set of stakeholder groups, particularly those who are typically underrepresented and, therefore, more likely overlooked (for examples, see Langford et al., 2020 Messick & Bazerman, Reference Messick and Bazerman1996).
PDE has been widely used to facilitate reasoning and improve decisions for hundreds of years (Cavanaugh, Reference Cavanaugh2006) and has spread to a wide range of fields from bioethics to military studies (e.g., Lee, Reference Lee2004; Lindblad et al., Reference Lindblad, Lynöe and Juth2014). At its core, PDE provides a series of conditions that practitioners should consider when deciding to act (Cavanaugh, Reference Cavanaugh2006). The first of these conditions include the requirement that practitioner actions must be good or at least neutral. This is consistent with the taxonomy presented in the focal article, as this condition immediately rules out any intentional unethical or corrupt behavior. Practitioners must also intend only good outcomes, and any adverse effects that result from the decision must not be intended or be required to bring about a desired outcome. Finally, any anticipated positive outcomes must be proportionally good to offset any negative consequences, and practitioners should do everything in their power to minimize any foreseeable harm (Cavanaugh, Reference Cavanaugh2006).
As an applied ethical decision-making framework, PDE makes a clear distinction between what is intended and what may be reasonably foreseeable by taking “a careful examination of the agent’s practical reasoning” (Monge & Hsieh, Reference Monge and Hsieh2020, p. 365). By encouraging practitioners to consider the negative effects of their decision (both intended and unintended), individuals who integrate the aforementioned conditions of PDE into their decision-making process are better able to systematically assess the potential negative consequences of their actions that were foreseeable but inadvertent. This, in turn, improves one’s ability to consider areas of potentially unintended but reasonably foreseeable harm that may have been overlooked. Ultimately, we believe that deliberate consideration of unintended but reasonably foreseeable harm may convert what would be an anodyne and straightforward decision into one that is, following Lefkowitz’s (Reference Lefkowitz2021) definition, an ethical dilemma. In this way, PDE acts as an effective means of purposefully scanning stakeholder groups to identify any additional points of unintended but reasonably foreseeable harm. The practitioner can then incorporate a more comprehensive set of stakeholder outcomes into their decision-making process while simultaneously working to minimize harm.
To help distinguish the contribution of our commentary, consider the current use of big data and algorithms to make work-related decisions (Newman et al., Reference Newman, Fast and Harmon2020). This is a growing area of concern for I-O psychologists, as organizations attempt to use recent advances in artificial intelligence to inform personnel decisions (Chamorro-Premuzic et al., Reference Chamorro-Premuzic, Winsborough, Sherman and Hogan2016). If one were to examine the decision, for example, to adopt an algorithm to screen applications for future organizational openings, it may not meet all three stated criteria for an ethical dilemma. Specifically, practitioners who evaluate their past actions based on the taxonomy presented in the focal paper may view their decisions as innocuous and not invoking ethical principles. This is evident as experts have readily adopted algorithms in the recruitment and selection of candidates that were later found to adversely impact members of underrepresented groups in both the corporate world (Burke, Reference Burke2020) and in academia (Dastin, Reference Dastin2018). In these cases, the potential unintended but reasonably foreseeable effects of building an algorithm using biased foundational assumptions and/or data play a secondary role in the desire to implement recent advances in technology to improve the selection process. In this case, practitioners could have taken a more systematic and detailed review of the likely consequences of using these data in an effort to better identify the potential negative outcomes associated with their actions.
In this example, we are not suggesting an additional structural form of ethical dilemma; rather, we argue that ethical dilemmas should be more commonplace upon further consideration of the unintended but potentially foreseeable negative consequences of one’s actions. By adopting the PDE lens, we argue that the decision to adopt algorithms for personnel decisions would be more likely to be viewed as an ethical dilemma. Thus, thoughtful and deliberate consideration of these decisions, using PDE’s emphasis on intention, is likely to unearth a number of fundamental ethical and moral concerns around the introduction of a potentially biased algorithm that may disproportionately affect employment outcomes for underrepresented group members (see Bogen, Reference Bogen2019; O’Neil, Reference O’Neil2016).
In sum, we applaud the critical contribution of the focal paper by Lefkowitz (Reference Lefkowitz2021), which makes an important step in underscoring the importance and prevalence of ethical dilemmas in I-O psychology. With this commentary, we leverage PDE, a long-standing, applied ethical decision-making framework, to address ethical dilemmas in I-O psychology. It is our view that by including an emphasis on the unintended yet reasonably foreseeable effects of our decisions, we, as I-O practitioners and researchers, are better equipped to take a reasoned approach that attempts to minimize harm by considering adverse effects on a wider set of stakeholders. Although this does not make the decision-making process any easier, it is likely to expand both the breadth and depth of our understanding of ethical dilemmas and result in better quality decisions, particularly those that may affect others adversely. Further, we argue that the often-overlooked voices of minority or marginalized stakeholder groups will be better represented when practitioners adopt a PDE view of ethical dilemmas.
We appreciate the work of Lefkowitz (Reference Lefkowitz2021) to encourage more reflection regarding the structural forms of ethical dilemmas faced by those in industrial-organizational (I-O) psychology. As professionals, there are few situations where decisions are entirely free of ethical ambiguity or negative consequences. This notion is made particularly salient by the focal study’s findings that highlight the depth and breadth of ethical dilemmas, defined as a choice between unattractive alternatives that the actor “does not want to make” (Lefkowitz, Reference Lefkowitz2021), faced by Society for Industrial-Organizational Psychology (SIOP) members. Given this definition, we extend the focal study’s contribution by using the principle of double effect (PDE), a decision-making approach that serves to guide actors on the moral permissibility of their actions, to argue that I-O practitioners would recognize more situations as presenting a genuine ethical dilemma if they evaluated unintended, but reasonably foreseeable harms that may result from their decisions.
In addition to providing a working definition of ethical dilemmas, Lefkowitz (Reference Lefkowitz2021) sees these dilemmas as made up of three parts: the necessity of making a decision, the need to apply some form of moral/ethical principle to this decision, and the reality that the decision will have significant consequences for others. Moreover, an implicit assumption of ethical dilemmas is the belief that individuals who are faced with these situations are motivated to do good and have no desire to cause harm.
Although we agree with the definition of ethical dilemma brought forward by the author, we suspect that this definition is somewhat conservative and may underestimate the extent to which practitioners experience ethical issues on a daily basis. For example, it may be possible that practitioners make a particular decision without fully thinking through the full spectrum of consequences of their actions, particularly as: “Professional ethics is a rarely considered subject in industrial-organizational (I-O) psychology, which might imply that it is not of much concern to the field” (Lefkowitz, Reference Lefkowitz2021, p. 297). Given this observation and notion that the structural forms of ethical awareness rely on awareness and intention (e.g., opportunity to prevent harm), it makes sense that practitioners may be unaware or have not fully considered the unanticipated, albeit reasonably foreseeable, consequences of their actions and, therefore, do not initially classify their actions as constituting an ethical dilemma. This limits the scale of the problem and our ability fully to realize the extent to which ethical dilemmas affect I-O practitioners.
As a remedy, we propose that, in addition to observing the taxonomy provided by Lefkowitz (Reference Lefkowitz2021), practitioners adopt PDE as part of a more systematic and detailed decision-making approach to help actors evaluate decisions that may reasonably lead to or contribute in some way to negative outcomes for others. This, we argue, augments our understanding and classification of ethical dilemmas in I-O psychology, as it expands the conversation to include both the context of the choice and the unintended, but reasonably foreseeable, effects of decisions on others. Thus, if the situation has not yet been determined to be an ethical dilemma according to the strictures set forth by the focal article, PDE encourages practitioners to exceed these notions by encouraging them to consider the foreseeable, yet formerly unforeseen, effects of their decisions. It also encourages actors to consider the potential harms that may be felt by a wider set of stakeholder groups, particularly those who are typically underrepresented and, therefore, more likely overlooked (for examples, see Langford et al., 2020 Messick & Bazerman, Reference Messick and Bazerman1996).
PDE has been widely used to facilitate reasoning and improve decisions for hundreds of years (Cavanaugh, Reference Cavanaugh2006) and has spread to a wide range of fields from bioethics to military studies (e.g., Lee, Reference Lee2004; Lindblad et al., Reference Lindblad, Lynöe and Juth2014). At its core, PDE provides a series of conditions that practitioners should consider when deciding to act (Cavanaugh, Reference Cavanaugh2006). The first of these conditions include the requirement that practitioner actions must be good or at least neutral. This is consistent with the taxonomy presented in the focal article, as this condition immediately rules out any intentional unethical or corrupt behavior. Practitioners must also intend only good outcomes, and any adverse effects that result from the decision must not be intended or be required to bring about a desired outcome. Finally, any anticipated positive outcomes must be proportionally good to offset any negative consequences, and practitioners should do everything in their power to minimize any foreseeable harm (Cavanaugh, Reference Cavanaugh2006).
As an applied ethical decision-making framework, PDE makes a clear distinction between what is intended and what may be reasonably foreseeable by taking “a careful examination of the agent’s practical reasoning” (Monge & Hsieh, Reference Monge and Hsieh2020, p. 365). By encouraging practitioners to consider the negative effects of their decision (both intended and unintended), individuals who integrate the aforementioned conditions of PDE into their decision-making process are better able to systematically assess the potential negative consequences of their actions that were foreseeable but inadvertent. This, in turn, improves one’s ability to consider areas of potentially unintended but reasonably foreseeable harm that may have been overlooked. Ultimately, we believe that deliberate consideration of unintended but reasonably foreseeable harm may convert what would be an anodyne and straightforward decision into one that is, following Lefkowitz’s (Reference Lefkowitz2021) definition, an ethical dilemma. In this way, PDE acts as an effective means of purposefully scanning stakeholder groups to identify any additional points of unintended but reasonably foreseeable harm. The practitioner can then incorporate a more comprehensive set of stakeholder outcomes into their decision-making process while simultaneously working to minimize harm.
To help distinguish the contribution of our commentary, consider the current use of big data and algorithms to make work-related decisions (Newman et al., Reference Newman, Fast and Harmon2020). This is a growing area of concern for I-O psychologists, as organizations attempt to use recent advances in artificial intelligence to inform personnel decisions (Chamorro-Premuzic et al., Reference Chamorro-Premuzic, Winsborough, Sherman and Hogan2016). If one were to examine the decision, for example, to adopt an algorithm to screen applications for future organizational openings, it may not meet all three stated criteria for an ethical dilemma. Specifically, practitioners who evaluate their past actions based on the taxonomy presented in the focal paper may view their decisions as innocuous and not invoking ethical principles. This is evident as experts have readily adopted algorithms in the recruitment and selection of candidates that were later found to adversely impact members of underrepresented groups in both the corporate world (Burke, Reference Burke2020) and in academia (Dastin, Reference Dastin2018). In these cases, the potential unintended but reasonably foreseeable effects of building an algorithm using biased foundational assumptions and/or data play a secondary role in the desire to implement recent advances in technology to improve the selection process. In this case, practitioners could have taken a more systematic and detailed review of the likely consequences of using these data in an effort to better identify the potential negative outcomes associated with their actions.
In this example, we are not suggesting an additional structural form of ethical dilemma; rather, we argue that ethical dilemmas should be more commonplace upon further consideration of the unintended but potentially foreseeable negative consequences of one’s actions. By adopting the PDE lens, we argue that the decision to adopt algorithms for personnel decisions would be more likely to be viewed as an ethical dilemma. Thus, thoughtful and deliberate consideration of these decisions, using PDE’s emphasis on intention, is likely to unearth a number of fundamental ethical and moral concerns around the introduction of a potentially biased algorithm that may disproportionately affect employment outcomes for underrepresented group members (see Bogen, Reference Bogen2019; O’Neil, Reference O’Neil2016).
In sum, we applaud the critical contribution of the focal paper by Lefkowitz (Reference Lefkowitz2021), which makes an important step in underscoring the importance and prevalence of ethical dilemmas in I-O psychology. With this commentary, we leverage PDE, a long-standing, applied ethical decision-making framework, to address ethical dilemmas in I-O psychology. It is our view that by including an emphasis on the unintended yet reasonably foreseeable effects of our decisions, we, as I-O practitioners and researchers, are better equipped to take a reasoned approach that attempts to minimize harm by considering adverse effects on a wider set of stakeholders. Although this does not make the decision-making process any easier, it is likely to expand both the breadth and depth of our understanding of ethical dilemmas and result in better quality decisions, particularly those that may affect others adversely. Further, we argue that the often-overlooked voices of minority or marginalized stakeholder groups will be better represented when practitioners adopt a PDE view of ethical dilemmas.