Improving Our Ethical Choices: Managing the Imp of the Perverse
Published: September 8, 2008
“Induction, a posteriori, would have brought phrenology to admit, as an innate and primitive principle of human action, a paradoxical something, which we may call perverseness…. Through its promptings we act without comprehensible object; or, if this shall be understood as a contradiction in terms, we may so far modify the proposition as to say, that through its promptings we act, for the reason that we should not.”—Edgar Allan Poe, “The Imp of the Perverse”
“I beg of you to take note that the spirit of mystification, which in some men ensues neither from an effort nor from scheming, but from an accidental inspiration, is akin, if only through the intensity of desire, to that humor …which drives us toward a multitude of dangerous or improper actions.—Charles Baudelaire, “The Mauvais Vitrier (The Bad Glazier)”
The powerful mix of ethical fading, cognitive distortions, and perceptual biases I explored in “Designing Ethical Experiences: Juicy Rationalizations” shows we all carry a bit of the Imp of the Perverse. What can designers do to balance the many factors that lead us to make unethical choices—and see no evil when we do?
Poe says the Imp of the Perverse is “...a radical, a primitive impulse—elementary” and warns, “With certain minds, under certain conditions, it becomes absolutely irresistible.”  Psychologists and ethics researchers, however, say we can take simple steps to align our Want and Should Selves over the three phases of decision making and help keep the Imp of the Perverse in check. They offer four primary recommendations, which I’ll consider from a design perspective: 
- Recognize our multiple Selves.
- Listen to the Want Self during prediction.
- Increase the influence of the Should Self during action.
- Decrease the influence of the Want Self during action.
Figure 1 illustrates these recommendations across the decision cycle.
Figure 1—Recommendations for aligning our multiple Selves
1. Recognizing Our Multiple Selves
Recognizing our multiple Selves means understanding that we all carry a bit of Poe’s Imp of the Perverse. We need to acknowledge our habit of making overly optimistic predictions, and we should identify the biases and cognitive mechanisms that change our solid ethical principles into bounded and malleable viewpoints that change with context. These are fine first steps, but more active measures are necessary.
2. Listening to the Want Self During the Prediction Phase
Ethics researchers suggest two ways of anticipating and managing the Want Self’s ability to trump ethical concerns at the moment of decision:
- highlighting and priming our motivations in advance of making a decision
- rehearsing ethically challenging situations and preparing ourselves to handle them
Highlighting and priming our motivations—the things we want, or our goals—in advance of making a decision makes it more likely we’ll take the ethical path when making decisions . Designers can remind themselves—and others!—that decisions involve ethical concerns as well as business objectives and prime everyone who participates in decision making with the goal of creating the best and most appropriate user experience that matches the values of users.
Rehearsing ethically challenging situations lets designers identify powerful contextual factors that contribute to ethical fading during the action phase—such as tight deadlines or pressure from stakeholders—and prepare themselves to handle them. Simply observing the practice of rehearsing and advocating for ethical choices can offset the temporal separation of our Want and Should Selves and moderate the affects of temporal construal or visceral concerns.
3. Increasing the Influence of the Should Self During the Action Phase
Improving our ethical behavior requires that we increase the influence of our Should Selves during the action phase. Researchers suggest four ways to do this:
- defining difficult decisions in a way that emphasizes a situation’s high-level aspects
- making decisions with ethical implications early
- presenting all the options for an ethically difficult decision at the same time
- envisioning both the ethical and the unethical choices
Casting difficult decisions in a way that emphasizes the high-level aspects of a situation brings out the Should Self.  For example, in the case of social networks’ needing members to prove their growth models or secure additional funding, one could frame the choice of how aggressively to pursue membership goals as a trade-off between rapid, but unhealthy growth in the short term and slower, but healthier growth over the long term.
Changing the timing of ethical choices is another way of increasing the power of the Should Self during the action phase. People will commit to ethically difficult decisions in the present, so long as they will feel the effects in the future.  Changing the timing of such decisions by moving discussions and decisions that have ethical implications to an earlier point in the design cycle lets people to commit to a course of action in advance, when their Should Selves are prominent. For example, product or service designers who are defining social networking capabilities could present the public list of Social Networking Anti-patterns to their team members who define business objectives and set product vision. Securing early commitments to uphold good ethics in all aspects of a product’s design can help a team avoid the unethical practices the patterns describe. 
Presenting all the options for an ethically difficult choice at the same time offsets ethical fading and reduces the prominence of the Want Self. Presenting the options for such a choice sequentially contributes to our making bad choices. Reviewing options simultaneously lets people compare the Want and Should Selves’ choices against their previous commitments or a common framework. More on frameworks later. 
We should envision two kinds of choices when trying to resolve ethical dilemmas: the ethical choice and the unethical choice.  I recommend labeling choices to make the distinction very clear. Direct statements from users or customers about their values, goals, and beliefs—presumably gathered through personal contact with users, user interviews, and user research—make it much easier to label choices as either ethical or unethical, because the choices that clash with their values will inspire conflict.
4. Decreasing the Influence of the Want Self During the Action Phase
Researchers recommend two self-control strategies for decreasing the power of the Want Self during the action phase. Making public commitments to maintaining good ethics is one effective method, because they escalate our commitment to making the ethical choice.  UX designers working on social media could present A Bill of Rights for Users of the Social Web to stakeholders during the initial stages of a project—both to win their early public commitment to these principles and to make them the reference point for later decisions. 
The old-fashioned method of giving people time to think—what researchers call increasing the cognitive resources available to decision makers—also decreases the power of the Want Self.  When making ethically charged decisions, we need to remain focused on the decision and how it will affect the people involved.
Figure 2 maps these strategies to the three phases of the decision cycle—prediction, action, and evaluation.
Figure 2—Recommendations for aligning our Want and Should Selves
Carrots and Sticks: Ethical Standards and Sanctions
In addition to the recommendations I’ve already provided, researchers advise that we carefully consider any standards or sanctions systems—think of the United Nations, for example—we might put in place to encourage ethical choices. Sanctions combine standards of measurement, incentives, punishments, and monitoring mechanisms to encourage decision makers to do the right thing in a situation with ethical implications. When we find ourselves in ethical dilemmas, the presence of standards and sanctions systems can increase the likelihood that we’ll make ethical choices by changing the way we frame our choices and leading us to see ethical choices as sound business decisions. 
However, sanctions systems can also contribute to ethical fading by changing the focus of decision makers from following ethical principals or being ethical to minimally meeting a standard’s specific guidelines or provisions. For decision makers, focusing on meeting a specific goal can suppress the Should Self and instead engage the Want Self in an immediate effort to achieve the goal. Some research findings indicate that ineffective or weak sanctions systems actually decrease ethical behavior! 
Overall, instead of enhancing the ethical quality of our decisions, weak standards and sanctions can actively fade the ethical aspects of our decisions and encourage the predominance of the Want Self.
Frameworks Are Your Friend
Ethical fading, the tension between our Want and Should Selves, and our natural tendency to create juicy rationalizations are powerful obstacles to the making of ethical design choices. As UX professionals, how can we better align our Want and Should Selves, ensuring that we create ethical experiences?
Rather than adding a few sessions of philosophy instruction to your project plan, choose a simple framework that can serve as the basis for the design decisions all stakeholders will face. Use this framework to gauge the ethical impacts of your team’s decisions and compare the choices that are available to you. A framework is a model or set of statements that define how we’ve understood a situation. Whether qualitative, structural, legal, or proscriptive, nearly any descriptive model or set of guidelines can serve as the basis for evaluating and comparing design choices in terms of their ethical aspects.
In combination with the strategy, process, and resolution techniques I suggested in part 1 and part 2 of this series of articles, frameworks offer substantial benefits for designers facing ethical challenges. They can serve as an impartial reference point that allows balanced comparison of design alternatives and options. The elements of the framework provide a common language for all team members who are involved in making design decisions. As Figure 3 shows, frameworks let designers focus on design, because they serve as a neutral mediator of conflicts between stakeholders with differing agendas.
Most important, frameworks counteract ethical fading by ensuring we remain conscious of ethical concerns throughout the design cycle.
Figure 3—Framework in the role of mediator
Choosing Your Framework
Good candidate frameworks for making ethical decisions include familiar user experience and design tools such as the User Experience Honeycomb or the Social Software Building Blocks [5, 6]. We should also consider relevant bills of rights such as A Bill of Rights for Users of the Social Web.  The ethical codes of professional bodies such as the Usability Professionals Association, the Association for Computing Machinery, the American Library Association, and the American Institute of Graphic Arts can also serve as candidate frameworks [7, 8, 9, 10].
Almost any framework is better than the unalloyed play of power. Perfection is less important than effectiveness when striving to encourage ethical behavior and decision making. And once stakeholders accept a framework as part of the design process, it is easier to persuade them to use the same or another framework in making other decisions.
Some essential characteristics of a good framework include:
- understandability by all parties
- applicability throughout the design cycle
All of the parties participating in design decisions must agree to use the framework in order for it to be effective.
Using Frameworks to Guide Decisions
Once a team has chosen a framework, it becomes their central reference for identifying and resolving ethical conflicts that affect design choices. A structural model like the Social Software Building Blocks can reveal the focal points for ethical conflicts. A prescriptive model such as Adam Greenfield’s Ethical Principles for Ubiquitous Computing can identify aspects of a user experience that require improvement—or would suffer as a consequence of unethical design decisions. 
Well-crafted personas, ethnographies, and profiles; quotations and images of customers from user research; and maps of social networks can fill the need for an identifiable victim, allowing designers to appreciate who their ethical decisions will help.
Frameworks are usually broad, so applying them to the specifics of a design project requires some interpretation. They are best used as a guide for the discussion and comparison of design choices in terms of their ethical impact rather than as a checklist of features or capabilities.
Figure 4 highlights the Social Software Building Blocks that were the focus of the ethical conflict Tagged.com experienced. Simply identifying the kinds of ethical conflicts that result from design decisions can help drive their resolution.
Figure 4—Tagged.com and the Social Software Building Blocks
Comparing Ethical Choices
Some decisions require subtle distinctions between good and bad. Try comparing hard-to-distinguish alternatives by quickly assessing the ethical impacts of the different design options versus the stable criteria of your chosen framework. This is easier and much quicker when you use the simplest and most direct kind of comparison that is effective.
In the example of Tagged.com, you can begin by identifying which social elements each design option will impact ethically. This kind of direct contrast is often enough to trigger the resolution of an ethical conflict. As a next step, gauge the degree of ethical impact for each option—that is, whether it will have a high or low impact. For example, if your team is considering three design choices with different, but overlapping ethical challenges, try to estimate the ethical impact of each option, as shown in Figure 5.
Figure 5—Comparing the social impacts design options
You could even try scoring the options for each element in your framework on a numerical scale from 1 to 5. In team settings, designers can share and compare their individual scores for options with the group’s scores. This can be an effective way of involving other stakeholders in decisions. Always remember that scorecards and ranking choices are good tools for starting and informing discussions and comparisons, but they’re not meant for rigorous analysis.
The timing of comparisons is important. Comparing all design options simultaneously during the action phase decreases the influence of the Want Self and increases the influence of the Should Self, as is consistent with my earlier recommendations. Scoring and comparing design options during the prediction, then the action phases helps designers to better understand changes in their own viewpoints, mitigating ethical fading. Those who are diligent can go a step further and revisit their decisions during the evaluation phase, supporting an effective culture of learning within their organization.
The UX designers who are creating the next generation of tools for life must find ways of handling the increasingly complex ethics of integrated experiences that span traditional boundaries such as physical versus virtual and business versus consumer. The kinds of conflicts of values and ethics that are now apparent in the social media arena portend even greater ethical challenges for the fully integrated experiences of the future.
This future will include new socio-technical forms of products with cultural dynamics—ranging from co-creation, hybrids, and spimes to ambient emotions and pervasive interconnection—coupled with holistic, long-horizon concerns like environmental impacts and sustainability. The combination of all these things makes the potential impact of our design decisions great and our ethical responsibility even greater.
By better understanding factors such as ethical fading and cognitive distortions, our competing internal voices, and our native biases, designers can better compensate for the Imp of the Perverse, which inspires us to act unethically, and overcome the juicy rationalizations we craft to justify our unethical choices after the fact. Better aligning our competing Want Selves and Should Selves will help us design ethical experiences for the rest of the people in the world, now and in the future.
 Wikisource contributors. “The Imp of the Perverse.” Wikisource, The Free Library. Retrieved May 1, 2008.
 Tenbrunsel, Ann E., Kristina A. Diekman, Kimberly A. Wade-Benzoni, and Max H. Bazerman. “Why We Aren’t As Ethical As We Think We Are: A Temporal Explanation.” HBS Working Paper #08-012 (2008).
 Microformats.org contributors. “Social Network Anti-patterns.” Microformats.org, Wiki. Retrieved May 1, 2008.
 Open Social Web contributors. “A Bill of Rights For Users of the Social Web.” Open Social Web, Wiki. Retrieved May 1, 2008.
 Morville, Peter. “User Experience Design.” Semantic Studios, June 21, 2004. Retrieved February 11, 2008.
 Smith, Gene. “Social Software Building Blocks.” nForm, April 4, 2007. Retrieved February 11, 2008.
 Usability Professionals’ Association. “UPA Code of Professional Conduct.” Usability Professionals’ Association. Retrieved February 11, 2008.
 Association for Computing Machinery. “Code of Ethics and Professional Conduct.” Association for Computing Machinery. Retrieved February 11, 2008.
 American Library Association. “Code of Ethics of the American Library Association.” American Library Association, January 22, 2008. Retrieved February 11, 2008.
 American Institute of Graphic Arts. “Standards of Professional Practice.” American Institute of Graphic Arts. Retrieved February 11, 2008
 Greenfield, Adam. “All Watched Over By Machines of Loving Grace.” Boxes and Arrows, December 1, 2004. Retrieved February 11, 2008.