Business actors often act in ways that affect or may affect other parties.Footnote 1 In many contexts, the law places restrictions or provides for remedies when one party affects another party adversely. At the same time, legal systems do not anticipate all contingencies nor are legal regulations always well-enforced. Moreover, even if fully specified, the law may nonetheless permit actions that affect others negatively. Actors in such situations require guidance as to whether the action is morally permissible or not.
One approach that has been developed to provide guidance to actors who may harm others is what is commonly referred to as the doctrine of double effect or the principle of double effect. This approach finds extensive application in just war theory, for example, in helping to determine when and how to engage enemy combatants when innocent parties are at risk of harm (e.g., Lee, Reference Lee2004; Lichtenberg, Reference Lichtenberg1994; Walzer, Reference Walzer2006). Bioethics is another field in which the approach finds widespread application (e.g., Aulisio, Reference Aulisio and Post2004; FitzPatrick, Reference FitzPatrick2003). In the case of business ethics, however, with a few exceptions (Bomann-Larsen & Wiggen, Reference Bomann-Larsen and Wiggen2004; Hughes, Reference Hughes2019; Masek, Reference Masek2000, Reference Masek2006; Tully, Reference Tully2005; Velasquez & Brady, Reference Velasquez and Brady1997), there has been little discussion of this approach. Given the fruitful application of this approach in other areas of practical ethics, we find this lack of discussion striking and explore its development and application in the context of business ethics in this article. In doing so, we take encouragement from a recent article in this journal that examines and develops its application to the ethical constraints on paying people to accept jobs that risk their life and health (Hughes, Reference Hughes2019).Footnote 2
We refer to the above approach as the logic of double effect (LDE) for two reasons. First, the term “doctrine” makes it sound like double effect is a matter of religious belief. Second, the approach involves more than a single principle (Cavanaugh, Reference Cavanaugh2006: xx, note 9). Our account of LDE is inspired by G. E. M. Anscombe’s Intention (Reference Anscombe2000). However, we do not intend to argue in favor of or engage in debates about virtue ethics, natural law philosophy, or Aristotelian management theory. On the contrary, we wish to show that LDE is independently compelling and compatible with a number of nonconsequentialist approaches. Cavanaugh (Reference Cavanaugh2006), Nelkin and Rickless (Reference Nelkin and Rickless2014), and Hughes (Reference Hughes2019), for example, have all proposed Kantian rationales for LDE.
Two key elements of LDE are: 1) the distinction between intended and merely foreseen effects (the intention/foresight distinction), and 2) the requirement that intended good effects be proportional to the bad effects. In this article, we present a series of cases in which we argue that the intention/foresight distinction usefully provides guidance for business actors about the permissibility of certain actions. Specifically, we argue that by underscoring the wrong in intending harm, LDE explains the exploitative nature of payday lending, distinguishes between permissible and impermissible forms of market competition, and provides a way to understand the potential wrong in imposing risk of harm on others. We also discuss the way that the requirement of proportionality raises the bar for justifying harmful activities, even if the harms are merely foreseen and not intended.
Our aim in this article is to make the case for taking LDE seriously as an approach to help business actors decide whether and how to pursue activities that may carry with them negative effects on others and also to engage with foundational debates about LDE. In addition, we aim to make a broader point regarding the nature of theorizing in the field of business ethics. Here again, we find it useful to refer to the field of bioethics. In tracing the history of bioethics, John Arras (Reference Arras and Zalta2016) makes the case for “midlevel” theorizing. He contrasts this with high-level theorizing, which involves applying some specific moral theory (e.g., utilitarianism) to decisions faced by actors or theorizing within the context of ideal political theory at the level of policies or institutions. LDE, to be clear, is not a complete moral theory. Nor does it assume ideal conditions. Rather it is compatible with a range of moral views and can be applied against a broader institutional framework given the particulars of a specific situation. We believe its fine-grained and particular nature represents a fruitful, but underappreciated, approach for business practitioners. While it is beyond the scope of this article to defend this claim, we offer our arguments on the relevance of LDE to business as an illustration of this approach. In so doing, we aim to convince readers not only of the plausibility of LDE, but of the value of midlevel theorizing more generally.
The article is organized as follows. Section 1 provides an overview of LDE and its origins and lays out our specific account. Section 2 engages with the foundational question for LDE about how to understand intention and how to ascertain what an agent intends as a means or an end from an examination of the agent’s chain of practical reasoning. Section 3 shows how the role of intention can shed light on three areas of debate in business: the exploitative nature of payday lending, the distinction between morally permissible and impermissible forms of competition, and the potential wrong involved in imposing a risk of harm. Section 4 outlines how the fourth condition of double effect, proportionality, applies to business ethics cases where harms are not intended, but merely foreseen. Section 5 takes up potential objections to our approach. Section 6 concludes.
1. ON THE LOGIC OF DOUBLE EFFECT
The Origins and Scope of LDE
The contemporary formulation of LDE is commonly understood to refer to the following set of criteria or conditions to be considered by an agent in deciding to act:
1. The act in question must be morally neutral or good (that is, it may not be intrinsically wrong);
2. The intended end must be good and the bad effects may be foreseen or permitted, but are not themselves intended (the principle of side effects);Footnote 3
3. The bad effects must not be a means for bringing about the good effects (the means principle), and;
4. The intended good effects must be proportional to the bad effects (the proportionality condition).
These conditions can be traced to LDE’s original formulation, attributed to Thomas Aquinas:
Nothing hinders one act from having two effects, only one of which is intended, while the other is beside the intention. Now moral acts take their species according to what is intended, and not according to what is beside the intention, since this is accidental as explained above (II-II:43:3; I-II:12:1). Accordingly, the act of self-defense may have two effects, one is the saving of one’s life, the other is the slaying of the aggressor. Therefore this act, since one’s intention is to save one’s own life, is not unlawful, seeing that it is natural to everything to keep itself in “being,” as far as possible. And yet, though proceeding from a good intention, an act may be rendered unlawful, if it be out of proportion to the end.
(ST, II-II, q. 64, a. 7)The conditions currently understood to make up LDE are at least implicitly present in Aquinas’s original formulation (Aulisio, Reference Aulisio and Post2004; Mangan, Reference Mangan1949). While Condition 1 is not explicitly stated in the original formulation, in order to argue that the act of self-defense is permissible according to the rest of the conditions Aquinas lays out, it must be true that the act of self-defense is itself, at minimum, morally neutral.Footnote 4 Conditions 2 and 3, which make up the intention/foresight central to LDE, can be inferred from the statement that “moral acts take their species according to what is intended and not according to what is beside the intention, since this is accidental” (ST, II-II, q. 64, a. 7).Footnote 5 The roots of condition 4 can be found in the statement that “though proceeding from a good intention, an act may be rendered unlawful, if it be out of proportion to the end” (ST, II-II, q. 64, a. 7), which requires that one’s means be proportionate to one’s ultimate end.
While Aquinas only explicitly applies LDE to self-defense against an unjust attacker, scholars writing after him, such as Thomas Cajetan, extended its application to other contexts, such as indirect killing of innocent noncombatants during war and risking one’s life for a just cause.Footnote 6 Cajetan was also responsible for broadening the understanding of proportionality to refer not only to the relationship of means to end(s) but also to the relationship between the bad and good effects.Footnote 7 Scholastic thinkers writing in the sixteenth and seventeenth centuries follow the same pattern in applying LDE to a variety of examples: They hold either explicitly or implicitly that the action performed ought to be morally good or neutral, and that one’s intention in performing the action must be good; they explain that the bad effect must not itself be intended; they do not allow one to intend bad effects as a means to good effects; and, they all explicitly argue that the bad effects may be permitted “only for ‘a good reason,’ ‘for a sufficient reason,’ ‘for a good cause,’ or ‘for a reasonable cause’” (Mangan, Reference Mangan1949: 55–56). Lastly, in an important shift in the development of LDE, the publication of the Cursus Theologicus in 1647 makes clear that LDE is understood as a generalizable set of conditions. Thus, while some contemporary work on LDE assumes that its applicability is limited to cases of killing (e.g., Scanlon, Reference Scanlon2008), it is clear that from its inception, LDE was laid out in such a way that is consistent with broader application. We see no reason, therefore, to limit LDE to cases of causing death.
The Four Conditions of LDE
We now turn to expand briefly on each of LDE’s four conditions. Condition 1 implies that a good end does not justify a wrongful means. The implication here is that certain acts are wrong in themselves, such that no end, no matter how good, can justify performing them as a means. This condition comes first because it prevents the application of the following three conditions to acts that are intrinsically wrong (Cavanaugh, Reference Cavanaugh2006: 27–28).Footnote 8
Conditions 2 and 3 together make up what is commonly referred to as the intention/foresight distinction. Condition 2, the principle of side effects, requires two things: that the agent intend a good end and that the bad effects not be intended, but merely foreseen or permitted. Naturally, this raises the question of how to distinguish between intended and merely foreseen side effects. We take up this question in the next section by looking at how intention can be discovered through a careful examination of an agent’s practical reasoning.
Condition 3, the means principle, may seem redundant in light of condition 2, seeing as bad effects are intended if they represent the means for bringing about the good effects. Some have, therefore, combined the two principles into one, i.e., “the agent intends the good and does not intend the evil either as an end or as a means” (Cavanaugh, Reference Cavanaugh2006). However, it may be preferable to keep them separate in order to emphasize that there are two ways one can go wrong with regard to the bad effects: by intending them in themselves or as means to the good effects.
To illustrate the underlying rationale for the intention/foresight distinction, consider an example that demonstrates its intuitive appeal.Footnote 9 Your father’s honor has been falsely called into question. You decide to defend him, but, due to his living abroad, you must travel to do so in a non-native tongue. You know that, despite your best efforts, you will inevitably make mistakes in speaking the foreign language. In denying the intention/foresight distinction, one must endorse two contradictory statements: that you intend to speak and make mistakes (P and Q), while at the same time intending to speak and not make mistakes (P and not-Q). Suppose also that you understand this. We are thus meant to believe that you knowingly intend an impossibility: that you intend both to bring about P and Q and P and not-Q while at the same time knowing that doing so is impossible. This cannot be right; it makes more sense to say that you intend only to defend your father, while merely foreseeing that you will make mistakes.
The distinction carries moral implications. Specifically, when certain consequences of an act are intended, a particular moral description of the act will rightly apply, and this description will differ from cases when those same consequences are merely foreseen. There is a difference, for instance, between cutting off a foot to save a leg and cutting off a foot to torture someone. It is this difference in descriptions to which Aquinas refers in saying that acts “take their species according to what is intended, and not according to what is beside the intention” (ST, II-II, q. 64, a. 7).
In intending an end, one acts or refrains from acting in order to try to bring about a particular state of affairs (Boyle, Reference Boyle1980; Shaw, Reference Shaw2006). While one may choose to act in a way that one foresees will bring about a set of side effects, one does not act for their sake. Stated differently, one does not adapt one’s practical reasoning—that is, one does not adopt and execute a plan of action—so as to bring about mere side effects (Finnis, Reference Finnis2011). In some cases, side effects may even present an impediment to bringing about the desired state of affairs, yet one acts in spite of them (Boyle, Reference Boyle1980). Recall the example about defending your father in a non-native tongue. Your lack of fluency may work to the detriment of being able to effectively defend your father, and yet you may choose to speak on his behalf anyway.
This does not mean that foreseen side effects are irrelevant to moral deliberation. Indeed, agents often try to avoid or minimize them. However, while one may act despite the possibility of foreseen side effects, one always acts because of the intended effects (Boyle, Reference Boyle1980: 535). Merely foreseen effects are not chosen in the same way as intended effects. Thus, we are rightly held to a higher standard of justification with regard to what we intend, rather than what we merely foresee, because it is more intimately bound up in the exercise of our moral agency.
Moreover, as Charles Fried (Reference Fried1978) has argued, the categorical nature of the moral injunction “do no harm” can only be understood with respect to the harms that we intend rather than all of the harms that may result from our actions. Otherwise, we would be constantly running afoul of it. If one considers “do no harm” a foundational moral precept, the relevance of intention is inescapable, as we simply cannot avoid imposing harms or risks of harm on other parties in the course of our daily lives.
Some accounts of double effect concern themselves only with the intention/foresight distinction captured by the principle of side effects and the means principle, and this distinction is the subject of most of the debate surrounding the plausibility of LDE.Footnote 10 However, we argue that condition 4, the proportionality condition, adds something important to LDE, namely the ruling out of some actions that may cause bad effects, even if the agent does not intend them.
The contemporary formulation of proportionality stipulates that the intended good effects should be proportional to the bad effects. There is more than one possible way of interpreting this condition. We adopt the following formulation:
4. There is “proportionately grave reason for permitting” the bad effects (Mangan, Reference Mangan1949: 43).
Consider someone who speeds because she gets pleasure out of it. Although one could say that she does not intend the increased risk of harm to others, she does not have proportionately grave reasons for speeding. In contrast, consider an ambulance speeding to get a patient to the hospital: saving the life of the patient is a proportionately grave reason to put other drivers at increased risk of harm.
Proportionality also directs us to take steps to prevent or mitigate the foreseen bad effects.Footnote 11 For instance, imagine you wish to test new software for autonomous vehicles, but there is a chance that real-world testing will put third parties at risk in the short run. If the promise of future safety benefits greatly outweighs those risks, proportionality would seem not to rule out one’s actions, assuming one exercises due caution and tries to take steps to prevent bad effects, such as engaging in testing in real-world driving conditions only after you have tested under controlled lab conditions that adequately simulate real world conditions.
Electing riskier over less risky means is one way to run afoul of proportionality, but it is not the only way. One may also run afoul of proportionality even if there are no alternative, less risky means for achieving the same end. Again, imagine a manufacturer of autonomous vehicles who cannot design controlled lab conditions to adequately simulate actual driving conditions before moving onto real-world testing. If high risk of harm to drivers and pedestrians is foreseen, then the choice to do real-world testing may be impermissible, even if there are no alternative means available for perfecting the technology.
In summary, the proportionality condition requires us to scrutinize the bad effects and ask whether they are warranted given the end one hopes to achieve. Its function is to rule out acts that may cause merely foreseen bad effects when agents do not have proportionately grave reasons or fail to take measures to prevent or mitigate those effects. As we will discuss in the fourth section, understanding proportionality in this way introduces the possibility of considering other relevant moral principles, because it implies that LDE is not meant to explain permissibility or impermissibility solely in reference to itself. LDE is, therefore, compatible with a number of nonconsequentialist moral theories. This pluralism is part of its appeal.
Understanding LDE
We now clarify two points about how we understand LDE more generally. The first point concerns the distinction between absolutist and nonabsolutist versions of LDE (Stuchlik, Reference Stuchlik2017: 69). According to the absolutist version, intentionally causing harm or death is absolutely unjustifiable. According the nonabsolutist version, intentionally causing harm or death may be permissible, but faces an increased burden of justification relative to merely foreseeing harms. There may be good reason to opt for the absolutist version of LDE in the context of cases that involve intentionally killing the innocent. Nonetheless, because we argue that LDE applies more broadly to cases of harm, it makes sense to leave open the possibility that the nonabsolutist version is more appropriate.
Secondly, whereas LDE is typically presented as specifying conditions that need to be met for permissible action, we favor conceptualizing LDE as excluding acts from permissibility. Stated another way, LDE serves to make a presumptive case for ruling out acts that intend bad effects as a means to one’s good ends. In conceptualizing it this way, we are extending the logic of exclusion that Anscombe (Reference Anscombe1982) argues characterizes the principle of side effects to LDE more generally. One reason for doing so is that, in considering real cases with all of their attendant quirks, we may have more certainty about what is impermissible, as compared to what is permissible. Thus, not being excluded from permissibility should be interpreted as a signal to proceed with caution, rather than as a guarantee of permissibility.
2. The CONCEPT OF INTENTION
Given the central role of the intention/foresight distinction in LDE, we now outline and illustrate what we take to be the relevant conception of intention. We do not define intention in terms of some other phenomenon, such as “desire” or “commitment.” Instead, we rely on what we call an Anscombian-Aristotelian approach, according to which intention is described by the answers to the question “Why are you X-ing?” where it is true that the agent is X-ing and there is an answer to the question “Why?” in the relevant range (Anscombe, Reference Anscombe2000: §5). The relevant range is defined by answers that provide the chain of means-end reasoning the agent considers to be relevant to the action she undertakes. Thus, intentions are revealed in the explanations an agent gives for acting in a given way.
To understand what we mean by a chain of means-end reasoning, consider the following pair of paradigmatic cases from just war theory, to which LDE has been traditionally applied:
Strategic Bombing: targeting a munitions factory in order to more quickly win a war while foreseeing that a number of civilians will die, versus,
Terror Bombing: carrying out a bombing that targets the same number of civilians in order to more quickly win a war.
LDE has been invoked to defend the claim that Strategic Bombing may be permissible, but Terror Bombing is not. How does LDE draw this distinction, given that the ultimate end of both cases to hasten the end of the war is permissible?
The fuller picture of the intended end and means in Strategic Bombing can be captured by asking about the chain of practical reasoning:
“Why are you dropping the bomb?”
“To wipe out the munitions factory.”
“Why are you wiping out the munitions factory?”
“To hasten the end of the war.”
The bomber, therefore, does not intend the civilian casualties, but only to take out the factory; the death of civilians is not a necessary part of the plan to hasten the end of the war. Contrast this to Terror Bombing:
“Why are you dropping the bomb?”
“To kill civilians.”
“Why are you killing civilians?”
“To decrease civilian morale.”
“Why are you trying to decrease civilian morale?”
“To put pressure on the enemy to surrender and hasten the end of the war.”
Thus, in Terror Bombing, civilian deaths are a necessary means for achieving one’s intention of bringing about the end of the war.
The chain of practical reasoning can be reversed in order to ascertain the agent’s means by starting with the end(s) and asking the question “How?” This second method is particularly helpful for determining whether the agent actually intends the bad effects as a means to the end and thereby runs afoul of the means principle. Supposing we know that the terror bomber’s ultimate intention is to hasten the end of the war, we can ask “How?”
“By putting pressure on the enemy to surrender.”
“How do you intend to achieve that?”
“By decreasing civilian morale.”
“How do you intend to decrease civilian morale?”
“By dropping bombs to kill civilians.”
By examining the chains of practical reasoning, one can ascertain that killing civilians does not form part of the intention in Strategic Bombing, whether as a means or an end, but is intended as a means in Terror Bombing. The actions in the latter case are therefore ruled out as impermissible. However, supposing that the strategic bomber’s plans have a reasonable chance of success in hastening the end of the war, and that doing so means fewer civilian and military casualties when compared to a protracted war, one may have proportionately grave reason to allow the civilian casualties.
It is important to note that this approach excludes cases in which the agent responds to the question “Why?” by saying “I wasn’t aware I was X’ing,” or “I didn’t mean to; it was involuntary.” Such answers reveal a lack of intention; actions performed for these reasons do not count as intentional in the relevant sense. To understand why, imagine seeing someone slump down and asking why he did so. Compare two answers he might give:
“I didn’t mean to; I fainted.”
“I was trying to hide from someone, because I owe her money.”
Of these two explanations, only the second one reveals an intention. What is in question is not simply a matter of causation, but the agent’s truthful explanation for why she chooses as she does. Intention is identified by the class of answer that, from the point of view of the agent’s own practical reasoning, reveals the reason “that actually gets the agent to do the act in question” (Shaw, Reference Shaw2006: 205). The first explanation (and other such answers) does not count as intentional, because it reveals that the act was not the result of an agent choosing a plan of action to realize an end.Footnote 12 While it is beyond the scope of this article to give a lengthy defense of this conception of intention, we do think it is relevant to point out why it is a natural choice for our purposes, namely to help business actors think about the impermissibility of harmful actions. Leadership, it has been said, involves determining the company’s vision and setting strategies for achieving it, and management involves creating plans, that is, “specific steps and timetables to implement the strategies” (Kotter, Reference Kotter2012: 73). Intention, conceived in terms of a present- and future-oriented chain of practical reasoning that explains why one chooses to adopt a particular plan, naturally captures or maps onto this kind of reasoning in business.Footnote 13
3. INTENTION AND BUSINESS ETHICS PROBLEMS
In this section, we discuss three contexts in which intention helps to make sense of widely held judgments about the permissibility of certain actions in commonplace business activity: the exploitative nature of payday lending; distinguishing between morally permissible and impermissible forms of competition; and, incorporating risks of harm, as opposed to certain harms, into business decision-making.
Payday Lending
The business ethics literature, and the philosophical literature more generally, have been occupied with developing an account of exploitation and explaining what is wrong with it (e.g., Snyder, Reference Snyder2010; Wertheimer, Reference Wertheimer1996; Zwolinski, Reference Zwolinski2007). In this discussion, the goal is not to cover this terrain but to use the example of payday lending—a practice that has often been put forward as an example of exploitation—to illustrate how taking the role of intentions seriously can help us make sense of the claim that the practice is indeed exploitative.Footnote 14
Typically, payday loan borrowers take out loans with a term of one to two weeks for an amount ranging between $100 to $500 (Pew Charitable Trusts, Reference Trusts2012). In the United States, these loans have a median annual percentage rate charge (APR) of 339 percent; on a $100 fourteen-day loan, the median fee charged is $15 per $100 borrowed (Consumer Financial Protection Bureau, 2013). In storefront payday lending operations, the median income of borrowers is $22,476 (Consumer Financial Protection Bureau, 2013).
In theory, payday loans are supposed to operate as an emergency stopgap measure to tide borrowers over until their next paycheck when they can get back on their feet. However, given the high fees and interest rates associated with such loans, many borrowers wind up in a debt cycle, with 67 percent of borrowers taking out more than six loans in a one-year period and 48 percent of borrowers taking out eleven or more loans over the same period, where the rolling over of the original loan counts as a separate transaction (Consumer Financial Protection Bureau, 2013: 21–22).
How might we characterize the lender’s intention here? Consider some hypothetical chains of practical reasoning in response to the question, “Why are you lending money at an APR of 325 percent?” One possible answer, which the industry itself gives in its lobbying efforts, is that they are merely intending to help borrowers who would not be able to access credit otherwise. One could ask, in reply, “Don’t very high borrowing costs harm the customers by subjecting them to increasing levels of risk over time of not being able to pay off their loans?” or, “Don’t the very high interest rates and fees keep borrowers in a debt cycle?” A lender could respond, “It cannot be helped if I am to make money from the loans.”
Here, one might think that the lender has rejected the premise of our question on the grounds that she has not answered the question “Why?” with the kind of “to …” answers that show that the actions in question were intentional.Footnote 15 This would be incorrect. In the “… if we are to …” statements that follow the assertion that “It cannot be helped,” one learns about the agent’s intention: “to make money off of the loans.” We can then ask the agent about her means: “How does charging an APR of 325 percent make you money off the loans?” Here, one needs to investigate the payday lending model, with the following question in mind: does any aspect of the plan to make money fail if most borrowers paid off their first loan instead of rolling them over repeatedly? Or, does any aspect of the business model fail without keeping a large subset of customers in a cycle of debt? To find out the answer, one would have to understand the profit-making model of payday lenders. For instance, upon which customers do they rely for profits?
There is strong reason to believe that keeping customers in a cycle of debt is an integral part of the industry’s business model, as new fees are generated every time a borrower rolls over the loan. In the United States, 75 percent of all loan fees come from borrowers who take out eleven or more loans in a year, whereas only 11 percent of loan fees come from borrowers who take out fewer than seven loans in a year (Consumer Financial Protection Bureau, 2013). In the United Kingdom, where payday lending is a newer, but growing, phenomenon, more than 25 percent of all loans are rolled over. Rolled over loans and default charges account for roughly 50 percent of the UK industry’s profits (Church Action on Poverty, 2013).
Consider also the following statements from those who work in the industry. Phil Locke, ex-payday lender in the United States, has stated that it is precisely those borrowers who roll loans over repeatedly that are the most profitable: “The cycle of debt is what makes these stores so profitable” (Church Action on Poverty, 2013). Even more telling is what Daniel Feehan, ex-CEO of Cash America, a payday lender, stated at a 2007 industry conference: “The theory in the business is you’ve got to get that customer in, work to turn him into a repetitive customer, long-term customer, because that’s really where the profitability is” (Center for Responsible Lending, 2014).
In response, lenders may admit that they intend to turn someone into a repeat customer but deny that they intend to keep customers in a cycle of debt. But moral agents cannot simply pare their intention in whatever way they want. The cycle of debt is a necessary component of the payday lending business model. Without it, there are no repeat customers and the business model does not work. The lenders’ intention, therefore, runs afoul of the means principle; the bad effects—in this case, a subset of vulnerable customers kept in a cycle of debt—are intended as a means.
In determining an agent’s means, what is relevant is not whether the agent ultimately achieves her end by way of the bad effects. Rather, the question is whether any part of her plan fails absent those bad effects. Furthermore, the impermissibility of using the bad effects as a means to the good effects does not hinge on whether the agent achieves the intended end. Imagine a case in which the lender does not succeed in making profitable loans because a tight employment market drives up local wages, so she is unable to yield enough repeat customers. Or, perhaps the lender is operating a bricks-and-mortar operation and finds that she is losing market share to online lenders. Even so, lenders intend to make profits by keeping customers in a debt cycle; the determination does not hinge on whether they actually succeed in turning a profit.
Understood this way, what is exploitative about the practice of payday lending is the intention to put some borrowers into a cycle of debt. While one may think it is permissible because of consent, what we have shown is that the moral work behind the claim of exploitation is not done by a defect in consent but rather by the plan for some people to be put in a difficult situation as a necessary part of one’s business model. Note that this is more than merely taking advantage of the borrowers or their situation; rather, it is the necessary reliance in one’s plan on their present and future vulnerability for one’s gain.Footnote 16
A question that may come up in this context is why we should believe that LDE heightens the justificatory bar for both consensual and nonconsensual harms rather than only nonconsensual harms.Footnote 17 Imagine, for example, a patient who suffers from a life-threatening virus. The only treatment available requires a doctor to cause him intense pain; only this will stimulate his body to release the requisite antibodies and hormones to kill the virus (Kagan, Reference Kagan1989: 168). How can LDE accommodate this?
The justificatory bar set by LDE may be met in such cases, as long as the consensual harm is undertaken for the sake of preventing an even greater harm from coming to pass to the person being harmed (Anscombe, Reference Anscombe1982). Thus, for instance, one may inflict pain on the patient to prevent her death, or amputate her foot to save her leg.
Characterizing LDE in this manner is consistent with the conclusions drawn in the payday lending example. In that case, the consensual harm is not being undertaken for the sake of preventing a great harm to borrowers. Firstly, the lender’s ultimate end is profitability, so the harm is imposed for the sake of the lender’s, not the borrower’s, benefit. Secondly, even if we grant that the lender was making the loans for the benefit of the borrowers, e.g., to help them meet their short-term financial needs, the lender does so by putting their long-term financial stability at risk. Unlike the case of the doctor inflicting pain on the patient to save his life, this is not a case where one is imposing a harm in order to prevent an even greater harm from coming to pass to the borrower. The conclusion LDE draws is one we should welcome. After all, as the debates on exploitation show, some cases demonstrate that it can be a struggle to make consent do all of the moral work in explaining our intuitions about how we can permissibly wrong one another.
Competition
A second context in which we find LDE promising is in helping managers to distinguish between morally permissible and impermissible forms of competition. To make things concrete, we consider the practice of exclusive distribution relationships: contractual arrangements in which producers allow distributors to sell their products on the condition that they do not carry their competitors’ products. Examples include automobile dealerships, which sell new cars from only one manufacturer, or arrangements that beverage companies, such as Coca-Cola and Pepsi, have with the transportation companies that deliver their products.
One way in which to decide the permissibility of such practices is in terms of aggregate consumer welfare. With this approach, if aggregate consumer welfare is reduced by such a practice—say, because of limited alternatives or an increase in prices owing to monopoly power—it would be judged impermissible or at least face a higher degree of scrutiny. This is roughly the approach adopted by regulators in many countries, including the United States (Federal Trade Commission, 2013). While there is reason for public regulators to adjudicate the legal permissibility of exclusive distribution relationships in terms of aggregate consumer welfare, there is a challenge to adopting such an approach for individual business actors. Namely, the use of markets is often justified on the grounds that because the informational requirements to realize aggregate welfare are too great for any one actor to incorporate into her decisions, markets increase aggregate consumer welfare by coordinating individual decisions based on more limited information. Hence, it may be overly demanding to ask individual business actors to determine the impact of their decisions on aggregate consumer welfare.Footnote 18
We argue that LDE provides individual business actors a more intuitive and useful way to distinguish morally permissible forms of competition from impermissible ones, by using the case of McWane, Inc. v. Federal Trade Commission. Footnote 19 To be clear, we are not attempting to explain the law or defend the decision that was reached in the legal case we discuss below. Rather, we take this case to illustrate the sorts of decisions that business actors are likely to face with regard to competition. In light of the uncertain legal landscape, managers will need sound moral advice, which we believe LDE supplies.
In 2008, McWane was the only US manufacturer of ductile iron pipe fittings (DIPFs), which direct the flow of pressurized water in municipal water pipeline systems. In response to the financial crisis of 2008, Congress enacted the American Recovery and Reinvestment Act (ARRA), which included $6 billion for water infrastructure projects and required all fittings purchased with ARRA funds to be manufactured domestically. Star, an importer and supplier of a broad range of waterworks products and fittings, announced it would produce DPIFs in the United States. Until it could build its own foundry, Star contracted with six existing domestic foundries to produce DIPFs on its behalf. By June 2010, Star offered the most commonly ordered DIPFs. McWane introduced the Full Support Program (FSP), which required its distributors to carry its products exclusively. The Federal Trade Commission (FTC) filed a complaint alleging that McWane violated section 5 of the Federal Trade Commission Act that declares “unfair methods of competition in or affecting commerce” to be unlawful.
McWane argued it needed to prevent Star from producing only the most popular fittings, which would leave McWane to incur the costs associated with satisfying orders for obscure parts. While there were more than 2,000 distinct DIPF configurations, 100 fittings satisfied approximately 80 percent of demand. Moreover, McWane argued that the FSP was pro-competitive because it helped keep McWane’s domestic foundry profitable. However, in subsequent litigation, the FTC found internal McWane documents confirming that the FSP was specifically designed to harm Star in “[raising its costs and impeding] it from becoming a viable competitor” by forcing it to provide a full line of products.Footnote 20 For instance, one executive wrote: “We need to make sure that they [Star] don’t reach any critical market mass that will allow them to continue to invest and receive a profitable return.” Footnote 21 In the end, the courts ruled against McWane in favor of the FTC.
We take the facts of this case to illustrate the ways in which intentions can help individual business actors distinguish between morally permissible and impermissible forms of market competition. While many forms of competitive behavior may harm competitors, one ought to distinguish between intended and foreseen harms. For example, a company may develop a better product that draws away customers from its competitor, or in the case of an exclusive distribution relationship, the relationship may be intended to avoid brand confusion (e.g., having a truck with a Pepsi logo distributing cases of Coca Cola). If the loss of customers that one’s competitors suffer in these scenarios are unintended harms—that is, if the competitors’ loss of customers is not a necessary component of one’s plan—then the foreseen effect of luring away a competitor’s customers may be permissible.Footnote 22 In contrast, while McWane no doubt sought to ensure its continued profitability through the FSP, the stated intention of the FSP was to harm Star by excluding it from the domestic market, rather than, say, to serve customers better. Moreover, even if nothing were explicitly stated in company documents about excluding Star, it is difficult to see how this exclusion would not figure into the chain of logic involved in explaining how the FSP would secure benefits for McWane. LDE helps us to explain how McWane’s actions differ from causing harm to competitors in unintended ways, such as simply making better products that lure away a competitor’s customers.
Risk of Harm
A third area in which intention is not only helpful, we argue, but necessary for giving managerial guidance, is in the context of making decisions that risk harm to others. Given the wide-ranging potential for serious harm caused by business activity, this is among the most important questions for the field. Online banking, for example, brings with it a risk of exposing sensitive customer data to hackers. Oil spills may result from breaches in the hulls of oil tankers, harming wildlife and damaging whole ecosystems, and so forth.
Before continuing, we should note that in speaking of “risk of harm,” we are referring not solely to situations in which the probability of harm is known to the agent. Rather, we use the term to encompass both conditions of “risk” and “uncertainty” as they are defined in the finance and economics literature. That is, we use the term to refer to potential harms about which probabilities may be known or unknown.
Nonconsequentialist moral theory typically distinguishes between how one ought to deliberate about certain harms, on the one hand, and risks of harm, on the other, often marshaling different theoretical resources to deal with these separately. One example of treating harms and risk of harms with separate theoretical tools can be found in an institutional approach, which argues in favor of allowing for some risk of harm on the grounds that everyone benefits from a society in which there is risky activity. One might imagine that all affected parties could theoretically agree to certain risks because the distribution of benefits and burdens is regulated by principles defined from a position of procedural fairness (Hansson, Reference Hansson2003). While this solution may be plausible against the backdrop of a well-functioning state with strong background institutions, if one is operating in a weak or dysfunctional state, this argument is less plausible as it would make imposing even modest risks on others impermissible.Footnote 23 Even in a well-functioning state, new technologies may give rise to previously unconsidered harms—witness, for example, the current debates about the regulation of self-driving cars, the risks associated with artificial intelligence, and the addictive nature of many apps and social media platforms.
Another nonconsequentialist approach is to invoke a right against risk imposition, but this leads to the problem of paralysis. That is, given the ubiquitous nature of risk in human activity, a right not to be harmed would render many uncontroversial human activities, such as driving, impermissible (Hayenhjelm & Wolff, Reference Hayenhjelm and Wolff2012). In order to deal with this problem, a variation is to argue for “a right that other people not impose risks of harm upon us above threshold R” (Holm, Reference Holm2016: 920). In addition to determining R itself, a problem with this variation is that it permits the distribution of risk of harm across a large number of individuals, so that the risk of harm to any one individual falls below the threshold R, even though it would be impermissible to impose the same harm with greater than probability R on a single individual (Holm, Reference Holm2016: 922). Another problem with this approach is that it only works in cases in which one knows the numerical probability of the potential harm in question. As in the case of exploitation, this is a large debate and we cannot engage it extensively here. Rather, we bring up these points in order to illustrate that it can be difficult to make sense of the intuition that there is something wrong in putting someone at risk, while still avoiding the problem of paralysis.
We argue that LDE can help to avoid these difficulties, because it corresponds to how moral agents experience decision-making as a present- and future-oriented phenomenon, particularly with regard to foreseen consequences. Specifically, in real life decision-making, there are few circumstances in which we have certainty about what will happen. Activities aimed at causing certain harm may fail. Often, the best we can hope for is sound knowledge regarding the probability of the consequences involved. However, the vast majority of our deliberations about what to do occur under conditions of uncertainty.Footnote 24 For instance, if we decide to speed moderately one time, we know it is not very likely that we will run over someone’s pet or rear-end another driver, but our knowledge is imprecise. In short, given how we, in fact, experience decision-making, the distinction between certain harms and risks of harms strikes us as somewhat arbitrary.
Because virtually all activity contains some amount of risk of success or failure, we propose treating risk akin to a bad effect under LDE and suggest that this approach is more demanding than it might initially appear. In particular, the principle of side effects and the means principle focus one’s attention on an agent’s intention to determine whether the risks of harm of a given action are intended through an examination of the agent’s chain of practical reasoning.
Recall the example of payday lending. We have argued that the risk of a debt cycle is intended as a means to profitability. This has implications for other kinds of risk, as well. Consider a researcher who releases a very low dose of a toxin into the water supply in order to study its effects (Hughes, Reference Hughes2019). His ultimate end is to make a valuable contribution to the medical literature. The researcher need not intend that the locals actually suffer harm; indeed, he may hope to find that the toxin is safe for human consumption in low doses. Nonetheless, the imposition of risk of harm is central to his plan. Without it, there is no research project and he cannot make a contribution to the medical literature.
Similarly, imagine a banking executive who is deciding on an investment strategy for the portfolio she manages. The bank that employs her is systemically important; it is “too big to fail.” As such, she suspects that, were her investment strategy to put the bank at risk of default, the government would step in to prevent the bank failure. If this were not the case, the executive would adopt a less risky investment strategy. Instead, she opts for a riskier strategy that has more potential upside for the firm (in the form of higher expected profits) and for herself (in the form of a bigger expected bonus). In other words, she makes a riskier bet than she otherwise would, knowing that the bank will not have to bear the downside risk of loss. The externalization of risk onto third parties (in this case, the government, and by extension, the taxpayers and the broader society) is central to her plan of making profits for the bank. This imposition of risk of harm is therefore intended, and not merely foreseen.Footnote 25
An important upshot to our approach to risk is that it requires that managers think seriously about the role risk impositions play in their ends, which in turn requires managers to be more deliberate in their planning and decision-making processes. Of course, if a manager is operating in bad faith, then one may not always be able to get at the truth of the matter, regarding intention. However, it is also true that if a manager is trying to “cover up” his intentions, he may go about things differently. His plans will include steps to give himself cover, which may betray his true intentions. The opposite is also true. If a manager has incorporated steps for reducing risk of harm to others into his decision-making, one would expect to see evidence of this.
At this point, one might ask what kind of guidance LDE offers in cases when harms or risks are foreseen, but not intended. We take up this question in the next section by examining the role of proportionality.
4. Proportionality: Impermissible Harm Without Intention
Our discussion thus far has centered around the ways in which LDE can help guide business actors and make sense of commonly held judgments by highlighting the role of intentions. In this section, we consider the proportionality condition and examine the role it plays in providing guidance to business actors. We show that, even if harms or risks of harm are merely foreseen and not intended, business actors must nevertheless provide a sufficiently compelling reason for the harmful activity. Not only does this condition serve to raise the justificatory bar for harmful activity, it also highlights the more general question of how to characterize the proper ends of business activities.
Consider a garment manufacturer that disposes of the waste product by releasing the it into a river a mile upstream from where villagers collect water for their daily needs. In response to the question “Why are you dumping those chemicals into the river?” a number of answers make clear that the garment manufacturer does not intend to harm the villagers downstream by using toxic chemicals, such as “to dispose of the waste product from the factory’s garment dying processes.”
However, the mere fact that the garment manufacturer does not intend to poison or harm the villagers does not necessarily exonerate her. The manufacturer must meet the burden of justification introduced by proportionality in two ways: with regards to her ends and her choice of means. One could imagine a variety of reasons given for the manufacturer’s choices, such as “I had no other means of disposal available to me,” or “I used these means because otherwise my operations would not be profitable,” or “These are the chosen means because I wanted to maximize profits.”
In the case of the first response, one might push back on whether this were really true. If so, as might have been the case historically, one might ask whether the benefits from the garment dyeing were disproportional given the cost to human life, particularly since garment dyeing was considered, until relatively recently, a luxury. In the case of the second and third responses, one must explore whether the end profit for the garment manufacturer is a proportionately grave reason for imposing harm on the villagers. One could ask, “Are there cost-effective means available to reduce or eliminate the risks to the villagers? Did you look into them?” If the garment manufacturer did not look into those measures while having foresight that toxic chemicals could introduce risk of harm to the villagers, then she would fail to meet the burden imposed by the proportionality condition. Our aim here is not to decisively defend the claim that profit does not give a proportionately grave reason, but to point to how proportionality introduces a burden of justification for business managers.Footnote 26
Finally, consider an example of a CEO directing the development of smartphone encryption to enhance privacy features, even though she foresees the risk that terrorists will try to make use of such technology to evade law enforcement surveillance. The end that the CEO seeks—enhanced privacy for consumers—is at least morally neutral, if not laudable. The terrorists’ use of the technology is not necessary for any part of the CEO’s strategic plan to succeed; therefore, the CEO cannot be said to intend to aid terrorists in their evasion of law enforcement by the addition of encryption features. Considerations of proportionality direct us to reflect on the CEO’s ends and whether they are sufficiently grave to override the interests that are threatened by the foreseen side effects. Thus, the CEO must consider whether privacy enhancements are sufficiently grave to permit the unintended safety risks. We argue that, when one considers the possibility of governmental abuse of unencrypted mobile phone technology in contexts rife with human rights abuses, there may be sufficiently grave reasons to decide in favor of privacy enhancements despite the potential safety risks. Indeed, in light of the human rights risks, failure to include encryption may be disproportional.
5. OBJECTIONS
In this section, we outline three objections that have been raised in regard to LDE and advance arguments to respond to the three objections from the vantage point of practical ethics.
Gerrymandering
The approach to intention proposed here is psychological and, therefore, subjective. One objection to such an approach is that it allows agents to give such a fine-grained account of intention as to allow them to escape moral responsibility for causing harm when they should be held accountable. The worry, referred to as “gerrymandering,” is that the agent might pare, shave, or withhold intentions, whether before, during, or after the act in question, in order to escape moral responsibility for an action. Consider, for instance, the McWane executives who argued in court that they were merely trying to protect the profitability of McWane’s domestic foundry.
Gerrymandering, we argue, does not constitute a serious objection to the intention/foresight distinction, because what is at stake in describing an agent’s intention is the truth of the matter, not what she says in order to exonerate herself. Indeed, the very term, “gerrymandering,” points to a mistake in conceiving the process of identifying intentions as a trial, or inquisition, in which the agent asserts false descriptions of her intention in order to get off the hook (Finnis, Reference Finnis2011). An agent lying or playing coy about her intention(s) does nothing to change the truth of the matter, and it is the truth—and not what a savvy agent could claim—that determines whether the agent’s actions ought to be deemed impermissible on the basis of her intention.
Moreover, we argue that, in business and management, the epistemological worry about getting to know agents’ true intentions is less pressing than in theory. In business, decision- and policy-making are often undertaken by groups and include discussions in meetings in which objectives and the means to achieving them are laid out and agreed upon. As we saw with McWane, business decision-making often includes the writing and dissemination of memos and emails that describe the policies and principles, according to which decisions are made on behalf of the firm. This reality lessens the concern that motivates the gerrymandering objection.
What about making an interior act of intention prior to or during the act in question in order to get away with doing something one otherwise could not? Even if one makes such an interior act of intention, this paring is nonetheless subject to the question “Why (or with what intention) do you pare, shave, or withhold your intention?” (Anscombe, Reference Anscombe2000: §27; Cavanaugh, Reference Cavanaugh2006: 86–87). Because one can examine second-order intentions the same way one examines first-order intentions, this strategy cannot be used in order to escape responsibility.
What about cases in which business practitioners purposefully make themselves ignorant of potentially harmful side effects to avoid taking responsibility for their actions? In these cases, one can dispute the claim that they do not foresee bad side effects. After all, in choosing to make themselves ignorant of the potential bad effects of their actions, they show that they possess some amount of foresight. One can ask, “Why did you not look into the potential harms from your policy or strategy?” If the intention was to avoid finding out about some issue that could prevent the execution of the plan, this would imply an intention to evade responsibility, which is morally suspect.
In normal cases of ignorance—where there is no foresight of bad effects—an examination of the chain of practical reasoning may uncover reasons like “We didn’t even think about that,” or “We were in too much of a rush.” If so, one must look beyond LDE to other moral concepts, such as negligence, in order to rule out their plans as impermissible. To claim negligence, one may have to establish that there is a pre-existing duty of care to those harmed. Business practitioners can also fail to discharge their general obligation to do proper due diligence. If a team, out of an overabundance of enthusiasm, were to move ahead with a project without proper due diligence, this would also be a moral failure on the part of its members.
Closeness
The second objection, closeness, bears some similarity to gerrymandering but represents a more serious problem. It derives its name from a worry that the intention/foresight distinction is so arbitrary that agents could truthfully claim that they do not intend harm in any case, even in cases featuring foreseen harms that appear to be one and the same with—or excessively “close” to—the stated intention.
The closeness problem is often discussed in the context of cases meant to be paradigmatic illustrations of LDE. For instance, it has been argued that the problem of closeness arises in the Strategic and Terror Bombing cases because the terror bomber can respond that she does not intend for the civilians to die; all that matters for her plan is that they appear dead long enough to strike terror into the hearts of the enemy (Bennett, Reference Bennett1995: 210–211). If this characterization of the terror bomber’s intention is admissible, the intention/foresight distinction fails to distinguish between the two cases, making terror bombing as permissible as strategic bombing. In such cases, the intention/foresight distinction at the heart of LDE effectively collapses. Critics of LDE argue that closeness is a serious problem (Bennett, Reference Bennett1995; Kamm, Reference Kamm1989; Nelkin & Rickless, Reference Nelkin and Rickless2013). Even many proponents of LDE concede that closeness represents a challenge that LDE must answer (Cavanaugh, Reference Cavanaugh2006; FitzPatrick, Reference FitzPatrick2006; Stuchlik, Reference Stuchlik2017).
One way of responding to the challenge would be to invoke a suitable account of closeness, such as closeness as constitutive relations (FitzPatrick, Reference FitzPatrick2006). However, accounts of closeness tend to be hotly disputed (Nelkin & Rickless, Reference Nelkin and Rickless2013). Instead, we argue that the closeness problem is most acute in the context of theoretical and highly stylized cases. In many real-life cases, a close examination of the chain of practical reasoning will suffice. For instance, in the Strategic and Terror Bombing cases, “the obvious question to ask is: how does the bomber intend to accomplish” making the civilians appear dead (FitzPatrick, Reference FitzPatrick2006: 590)? The answer is that the terror bomber intends to kill them as a means to making them appear dead.
The same strategy is useful in the context of business ethics cases. Consider, for instance, software programmers who have been tasked with designing apps in such a way to get users to engage as much as possible with them. One could imagine a case in which programmers might admit that they intend to keep user engagement as high as possible, but deny that they intend users’ smartphone or internet addiction. Critics of LDE might argue that such a case illustrates the futility of the intention/foresight distinction: Even though we would want to say that the programmers intend to make users addicted to their apps and thereby users’ smartphones, and rule the programmers’ actions as impermissible on this basis, we could not do so.
One response is to argue that spending large amounts of time on these platforms is constitutive of the harm of addiction (Bhargava & Velasquez, Reference Bhargava and Velasquez2019); that is, one could invoke an account of closeness.Footnote 27 We argue instead that the relevant consideration is that users are put at risk of addiction. As others have documented, apps are often designed to be addictive by relying on facts about brain chemistry and habit formation, even optimizing dopamine reward patterns to keep user engagement as high as possible (Haynes, Reference Haynes2018).Footnote 28 While it may be theoretically possible to conceive of a case in which developers pursue the end of ever-increasing user engagement without relying on the psychological and physiological mechanisms of addictive habit-formation, in practice, this is unlikely. We suspect the same is true about other cases which may appear to invite objections based on closeness: upon a careful examination of the agent’s truthful deliberation and means, one is likely to unearth evidence that the bad effects, or risk thereof, are a necessary part of the agents’ plans to achieve their ends. Thus, invoking an account of closeness may be unnecessary when discussing practical issues in business ethics; the means principle is likely to suffice.
In summary, there may be cases in which LDE does not yield what one might consider to be the right answer because of closeness, but we suspect that these cases are quite rare and far from our everyday experience. Therefore, while the closeness problem may represent a challenge to LDE in theory, as a matter of practical business ethics, where our concern is to address the situation facing the real-life manager given the facts and means at hand, we see it as a less serious objection.
Absurd Impermissibility
A third objection against LDE is that it is wrong to make the permissibility of an act hinge on the actual intentions of a specific agent. According to critics of LDE, doing so yields the absurd conclusion that acts thought to be straightforwardly permissible become impermissible solely in virtue of a bad intention of the agent (Rachels, Reference Rachels, Steinbock and Norcross1994; Scanlon, Reference Scanlon2008; Thomson, Reference Thomson1999). Consider a variant of the McWane case. Suppose that rather than harm Star, the McWane executives sincerely intended only two things: 1) to defend McWane against Star’s strategy of producing only the most popular fittings, which would leave McWane to incur the costs associated with satisfying orders for obscure parts; and 2) to promote competition by keeping McWane’s domestic foundry profitable. Why should adopting the FSP with this set of intentions change its permissibility? According to critics of LDE, this implication is absurd, and we should conclude that while intention may be relevant for assessing the character of moral agents, it is irrelevant for determining moral permissibility.
There are two responses to this objection. The first is to deny that LDE ought to be interpreted as hinging on an agent-dependent intention. One could argue that LDE ought to be understood as asking “whether a good agent could perform the action in question for the reasons given in that justification without intending” harm to innocent persons (FitzPatrick, Reference FitzPatrick2012: 192). On this understanding, LDE would rule out only those acts that no one could perform without intending harm. Under this response, McWane’s adoption of the FSP to keep Star out of the domestic market would be morally permissible, because possible justifications exist for the FSP, which involve pursuing a good end, such as promoting competition. This conclusion would contradict the analysis of the case we offered in the third section.
The second response is to draw a distinction between thin act-types (“implementing the FSP”) and thick act-types (“implementing the FSP to harm my competitor”), which contain the intention in their description (Wedgwood, Reference Wedgwood2011). If the thick act-type betrays an intention to act badly, then it makes sense that it should be avoided, which supports the conclusion that it is impermissible (McMahan, Reference McMahan2009; Wedgwood, Reference Wedgwood2011). On this account, McWane’s adoption of the FSP to exclude Star from the domestic market would be ruled out as impermissible, which is consistent with our analysis.
Independently of the analysis of the McWane case, however, why should one believe that R2 is the correct account of impermissibility? Does it not make more sense only to exclude actions that cannot be performed by an agent acting well, rather than raising the bar to exclude all actions that are performed by agents acting badly?
Besides the argument that the second response proscribes us from acting badly, which strikes us as fitting, the more general appeal is that it captures the commonsense intuition that one should do things for the right reasons. Note that the second response need not only yield the judgment that McWane should refrain from implementing the FSP, although that is one possible conclusion. The other possible conclusion that may fall out of hinging permissibility on particular agents’ intentions is the directive that the McWane executives ought to change the intentions with which they implement the FSP.
This last point leads us to a more general observation about the objection of absurd impermissibility. The objection arises in situations when two different intentions (one permissible and the other impermissible) can be attributed to the exact same act. We close this section by raising doubts about how realistic such situations are in practice. Again, consider the case of McWane. Suppose we grant that the intention of executives was to defend against having to incur the costs associated with providing obscure fittings and to promote competition by keeping McWane’s foundry open. In reality, if this had been true, it likely would have led to implementing means that were better suited to achieving these ends. Indeed, as the FTC pointed out, there was an alternative to the FSP—namely, McWane could have lowered the price on the popular fittings and then charged more for the less popular fittings to cover the cost.Footnote 29 This underscores the view of thinking about intention in terms of a forward-looking chain of practical reasoning, according to which we tailor our means to our intended end.
6. CONCLUSION
Double effect has a long history, having been extensively developed, applied, and debated in a variety of practical contexts. To date, however, business has not been one of them. Given that business actors frequently face situations in which their pursuit of positive benefits—whether in the form of profit, customer satisfaction, or employment—can negatively affect others, we find the absence of debate about LDE in the context of business to be striking. Business seems to be exactly the sort of activity that lends itself to theorizing about and applying LDE.
In this article, we have made the case for taking seriously LDE as an approach to help business actors decide whether or not to pursue activities that carry with them negative effects on others. Through a range of business ethics cases, we have illustrated how focusing on intentions can provide guidance to business actors, as well as make sense of commonly held judgments about the permissibility of certain business practices. We also discuss the way that the requirement of proportionality raises the bar for justifying harmful activities, even if the harms are merely foreseen and not intended. In making the case for the usefulness of LDE in business, we also engage with foundational debates about LDE more broadly.
To be certain, much work remains to be done if LDE is to play a more prominent role in business ethics. For example, as noted above, when discussing the issue of proportionality, LDE requires us to articulate the appropriate ends of various business activities. Our hope is that this article stimulates further research along these lines, as well as encourages further “midlevel” theorizing more generally. Such an outcome would not merely be a side effect, but intended.
Acknowledgements
We wish to thank participants in the Susilo Symposium 2017 at Boston University, the Fellows-in-Residence Seminar at the Edmond J. Safra Center for Ethics at Harvard University, a Wharton Legal Studies & Ethics doctoral student and alumni workshop at the University of Pennsylvania, the Adversarial Business Ethics Conference at Yale University, a West Coast Business Ethics Workshop at the University of British Columbia, as well as Rob Hughes, Daryl Koehn, Christopher Michaelson, and Katherina Pattit for their helpful feedback on earlier versions of this article. We are also grateful to three anonymous reviewers and Bruce Barry for their highly instructive and detailed comments and suggestions.
Rosemarie Monge is assistant professor in the Ethics and Business Law department at the Opus College of Business at the University of St. Thomas in Minneapolis–St. Paul. She received her BA in international studies, BS in economics, and PhD in legal studies and business ethics from the University of Pennsylvania. Her research examines questions of managerial responsibility from the perspective of nonconsequentialist morality.
Nien-hê Hsieh is professor of business administration and Joseph L. Rice III Faculty Fellow at Harvard Business School. His research and teaching aims at helping business actors and organizations determine and deliver on their responsibilities. He has written on a variety of topics, including worker participation in the management and governance of firms, the responsibilities of companies for human rights, the ownership of productive property, and choices involving incommensurable values. He taught previously at the Wharton School of the University of Pennsylvania.
