General

Teaching with Cases

Social Impact Analysis

Computer Ethics Curriculum

Curricula Index

Case Materials

Therac-25

Machado

Hughes Aircraft

Ethics in Computing Links

Contact Us

Ethics Tests

Many of the decision-making exercises presented in this website require employing different tests to generate and evaluate alternatives of action. This section is devoted to explaining these tests.

There are certainly many problems with the simplified tests we propose here. They caricature the ethical theories they are supposed to represent. They might lead to simplistic decision making. They leave out some important current approaches to ethics. We will deal with these objections toward the end of this essay.

So, given all these objections, why risk it? Our reply assumes the tests are used carefully, and with awareness of the shortcomings. First, we think these tests actually help to move students' ethical decision making toward an appropriate level of complexity.

Our experience teaches us that students mostly think in terms of outcome-based reasoning, except when they don't. When they don't, they will talk of basic rights. Research supports this observation (see Keefer M. & Ashley, K.D. (2001) Case-based Approaches to Professional Ethics: A Systematic Comparison of Students' and Ethicists' Moral Reasoning. Journal of Moral Education 30: 377-398). Both of these kinds of reasoning are usually at a low level and unanalyzed. Students often do not even know they are switching between them or how the appeals differ.

The ethics tests make the contrasts among the different types of tests explicit and apparent. Students learn that they are different and that occasionally they disagree. Just learning this much is a useful achievement.

In addition, the tests help the student learn that ethical decision making is not simply a matter of applying rules. The rules, in the guise of these ethics tests, often don't establish a definitive answer (or even a set of definitive answers). The most help they can provide is to guide thinking on the issues and to rule out some clearly inappropriate choices.

The tests also provide some structure when evaluating a set of options. We have argued in the section on teaching with cases [link] that some structure is necessary, and this structure is a reasonable one with which to begin. The tests should not be the only structure, but they are a good introductory one.

Finally, for those interested in exploring theories of ethics in more detail, these tests are a door into that exploration. What a critic may call shortcomings in these tests can also be treated as their strengths. We borrowed them from others and developed them because we think they capture essential insights about the sorts of questions major ethical theories propose to ask about decisions.

So, use these tests because they help make students' thinking more complex, they make explicit essential insights about major ethical approaches and make explicit the differences among those approaches, they teach students that the test do not always provide clear answers, and the tests provide some structure to students as they explore their own insights about a case.

The first three tests conveniently encapsulate or telescope three commonly taught ethical approaches. The fourth test presents an important concept in professional ethics, that professionals have special obligations because of their expertise. The final test ask a practical question that has ethical implications: can it be done?

• Harm/Beneficence: Does it do less harm and more good than the alternatives? This test is about the alternatives for all stakeholders
• Publicity: Would I want this choice published in the newspaper? This test is about what the decision would show about your character.
• Reversibility: Would I think this a good choice if I were among those affected by it? This choice is about whether stakeholders are being treated with respect
• Code of Ethics: How does this choice relate to the ethical standards of my profession? This test is about the decision-maker's duties in his or her role as a professional.
• Feasibility: Can this solution be implemented given time, technical, economic, legal and social considerations? This test is about practical issues.

The Harm Test highlights an essential component of the utilitarian ethical approach, the goal of minimizing harm and maximizing benefit. In contemplating an action, we do our best to envision its consequences, especially those likely to take place and those whose occurrence would produce severe harm. We then compare different courses of action in terms of the benefit to harm ratio they are likely to produce. We attempt to maximize this ratio.

1. Identify those who will be affected by your action.
2. Identify the impact your action will have on these people.
3. Determine whether this impact is harmful (Does it produce physical or mental suffering, impose financial or non-financial costs, deprive others of important or essential goods?) or beneficial (does it increase safety, quality of life, health, security, etc.)
4. Repeat these steps for the best available alternatives and compare them in terms of the benefit to harm ratio they produce.
5. Conclude by answering this question: Which alternative produces the best ratio of benefit to harm?

Problem: Students may be tempted to either stop too soon or go too far in their drawing out the consequences of an action. Too much enquiry will produce a "paralysis of analysis" for the student and discourage him or her. Too little may be an issue of a lack of moral imagination, lack of motivation, or a desire to support a predetermined decision.

Solution: One approach is to emphasize the reasonable person standard (or perhaps reasonable computer scientist or reasonable software engineer). But it is possible that students lack the moral imagination to be creative about the possible outcomes for various stakeholders. Some structuring of the process can help here. Ask students to consult the ImpactCS framework to help them get their list of stakeholders complete and to help them think of additional dimensions along which harm or benefit might occur. Ask them to list the elements of the socio-technical system to find additional stakeholders.

In the Publicity Test, a person's actions manifest essential elements of his or her character. What we do reveals who we are. Our actions provide others with a window through which they can view our souls. Under this test, when I contemplate an action, I ask whether I would want to be known as the kind of person who would do this. For example, if the action were cowardly, would I want to be known as a coward? If the action were irresponsible, would I want to be revealed (to myself as well as to others) as irresponsible?

This test encapsulates the approach know as virtue ethics. Virtue ethics is less about how right or wrong a particular action is and more about the character of the person. Thus, in this test, the action is judged in terms of what it says about the person, rather than on any effects that action may have.

1. Consider that the action you are about to perform provides a window through which others can see who you really are.
2. Then take the perspective of those others who are about to judge your character through your action.
3. Ask the following question: Would others view you as a good person for what you are about to do?

Variation:

Problem: Many students reduce the publicity test to the harm test by considering only the consequences of making the action public. For example, blowing the whistle on your company for illegal dumping of toxic wastes would fail the publicity test (under this misconception) because the consequences of making this dumping public would be the loss of your job and the adverse publicity suffered by your company.

Remedy: Any utilitarian calculation would include weighing the risk to your job and your company's image against the benefits brought to the public by revealing to them the illegal dumping. But this is the job of the harm principle, not the publicity principle. The issue here is what your action reveals about you the agent. What would people think about you if you passively went along with this illegal dumping? (Would they consider you a coward?) What would people think about you if you resisted this action, even to the point of going public and putting your job at risk? (Would they see you as a person of moral integrity who strives to do what is right even at the expense of personal sacrifice?) Perhaps your coworkers would look at you as disloyal (lacking the virtue of loyalty) but this would have to be weighed against the way the public you are trying to protect would view your action. Application of the publicity test can get complicated. But students can sort through most of this by keeping focused on the issue of what this action says about the agent as a moral person. Assume that our actions provide a window into our souls. What, then, does this particular action say about you as a person? Do you want to be known as the kind of person who would do this?

The Reversibility Test captures a central idea in Kantian formalism, the idea of universalizing one's actions. It is also the main idea behind the Golden Rule. Positively, it tells us to do to others what we would have them do to us. Negatively, it tells us not to subject anybody to something when we would be unwilling to have them subject us to it. Thus, we are treating reversibility as a key procedure to asking the universalizing question: "would I recommend that all persons in this situation act this way?"

On the worksheet, we add the guidance that the test is about treating others with respect. Again, we do this in the spirit of Kant who sees the issues of treating other as ends (rather than only means) as an issue of respecting the autonomy of all humans. Kant sees the universalizing question and the ends question as essentially the same, and so we combine them into this test. We think it helps make the test a more faithful caricature of Kantian ethics.

1. Determine who is going to be affected by your action.
2. Determine how they are going to be affected.
3. Reverse roles: put them in your place (as the agent or doer of the action) and yourself in their place (as the one subjected to the action).
4. Answer this question: If you were in their place, would you still find the action treated you with respect?

Closely related, alternative tests:

• Does the proposed action treat others with respect? (Does it recognize their autonomy or circumvent it?)
• Does the action violate the rights of others? (Examples of rights: Free and informed consent, privacy, freedom of conscience, due process, property, freedom of expression)
• Would I recommend that this action become a universal rule?
• Am I treating others in this situation only as a means to my own ends? (one is allowed to treat others as means, as in a business transaction, but not only as means)

Problem: Many students misapply the reversibility test in situations where they are being asked to comply with a morally questionable proposal. Take the case of the supervisor ordering you to dump a drum of toxic chemicals in the field behind the plant. A group of students might claim that this would fail the reversibility test because it upsets the supervisor: if you changed places with her, you would be upset when she refused to carry out your order.

Remedy: There are several responses to this. First, would she really be upset that a subordinate refused to carry out an order that was illegal? Second, the issue in the reversibility test is not whether your action may upset someone, but whether it treats those who will be affected by your action with respect. Refusing to carry out your supervisor's illegal order is consistent with treating her with respect if, in your refusal, you make it clear that the grounds of your refusal is not your lack of respect but your concern about the illegality of the order; in other words you are not objecting to the person but to the order. Third, an action may not be reversible with all stakeholders especially if stakeholder interests conflict with one another. In this situation, you must work to honor all the conflicting interests. If this should prove impossible then you must honor those that have the highest moral value. Sacrificing the safety and health of the people living near your plant in order to keep from upsetting your supervisor seems to get it backwards. Finally, have them focus on the action of the supervisor. Is his action reversible in relation to you or to the public whose health and safety is at stake? If his action violates the test of reversibility and you, nevertheless, go along with it, then your compliance would also violate this test.

If students have trouble working with the idea of reversibility, have them substitute other closely related tests. The Golden Rule is familiar and turns on the notion of reversibility. Another alternative would be to have them look at the rights involved. (Formalist ethical approaches argue that reversibility underlies our system of basic human rights and duties.) A third alternative is to have them examine whether their course of action treats stakeholders with respect or upholds their dignity.

Problem: Closely related to the previous problem is the tendency to reduce the reversibility test to the harm test. Often the question, "Would I think this a good choice if I were among those affected?" gets converted into the question, "What impact would my proposed action have on others?"

Remedy: Point out that the reversibility test focuses, not on the consequences of your proposed action, but on whether this action treats others with respect. Does it circumvent their ability to make decisions for themselves? Does it deceive or manipulate them? Is it paternalistic toward others in that it seeks to make decisions for them that they are capable of making themselves? Results enter into the answers to these questions but they are not the central issue; the central issue, again, is whether the proposed action treats others with dignity.

The code of ethics test asks that the agent benchmark the proposed course of action with the recommendations of a professional code of ethics. Engineers, for example, should look at the impact of their decisions on public health, safety, and welfare; almost all engineering codes identify this as the area of paramount responsibility.

1. Identify the provisions in the code that are relevant to the case at hand.
2. Answer the following question: Does your proposed course of action violate any of these provisions?
3. Check for any inconsistencies, i.e., instances where an alternative satisfies some code provision but not others. If there are inconsistencies, look for priority rules. (Example: many codes hold public health, safety, and welfare paramount.)

Problem: The code says nothing specific about the particular set of actions you are considering.

Solution: This is more a characteristic of codes of ethics than a problem in their application. Codes are not about the answers to specific situations, abut more about principles that are valued by the profession. Student will require some moral imagination to connect the principles to specific situations, and even ethicists with lots of moral imagination may not find much in a code that applies specifically. Encourage students to take the stakeholder approach listed in the hint above as a way of opening up their imagination. Encourage students to do the harm/beneficence and reversibility tests before tackling this one. Those tests may produce results that make this one more clear.

Codes of ethics are hard to apply before the students are familiar with the intermediate terms (like intellectual property and privacy). So, if student have trouble with this test early in a class, come back to it towards the end of the term so they can see they have grown in their ability to apply the code.

The feasibility test brings in a series of practical constraints by asking whether the selected alternative can be implemented given time, financial, legal, personal, and social constraints. By focusing the decision-maker on these constraints, the feasibility test helps to integrate ethical considerations with other aspects of a decision.

This integration of the ethical and the social is a central point of the ImpactCS approach to computer ethics. It is also an important issue for whether or not we can hold a person responsible for an action. The more an action is infeasible, the less one may have an obligation to do it. There are hard cases where this is not true (where, for instance, one may be require to try to do something, knowing it may fail).

Consider each of the following practical constraints that might bear on the proposed action:

• Time: Is there a deadline within which your solution has to be enacted? Is this deadline fixed or is it negotiable?
• Financial: Are there cost constraints on your solution? Are these fixed or are they negotiable?
• Legal: Does your proposed alternative violate any laws or regulations? Are the legal constraints in line with the results of your ethical evaluation? If not, what can you do to align them?
• Personal: Do the personalities of the people involved offer any constraints? For example, would your supervisor be open to persuasion, negotiation, or compromise? Or is he or she a dogmatic, close-minded, and inflexible person?
• Social, Cultural, or Political: Consider where your solution is being implemented. How would its impact be viewed through the social, cultural, and political milieu in which it is being enacted? Think of these issues using the several levels of analysis in the ImpactCs framework.

Problems with the Feasibility Test

Problem: Students think that the legal requirements trump ethical ones.

Solution: Often, student fall into this trap because of a lack of moral imagination. That is, they see the legal rule and simply say "Well, there's your answer." This may be used when wanting to follow a course of action that is shady in ethical terms, but "perfectly legal." This is done as a way of ending the search rather than beginning it. There are two things you can try in this case. First, emphasize the other tests and what they say, independent of the legal test. These tests are way to determine if a law is unjust. Second, ask students if they can think of an instance where a legal rule was clearly morally wrong (e.g. slavery in the US).

Problem: Many students treat these tests as representing different, competing theories. Thus, they go on to import several preconceptions about scientific theories into their thinking about the ethical approaches these tests encapsulate. They reason that ethical theories–like the scientific theories of creationism and evolution–represent exclusive accounts of the same phenomena. As a result, they will employ only one ethical test (usually the harm test), reasoning that it is best suited to the situation and then assume that a different test would lead to a different, even contradictory, conclusion.

Remedy: Ethical approaches are more like different perspectives on a multi-dimensional object than mutually exclusive accounts of the same phenomena. Instead of contradicting one another, they complement each other; each compensates for the limitations of the others.

An analogy will help here. When we go to buy a house and the view from the outside perspective is different from the view from the inside perspective, we don't conclude that one view contradicts the other; rather we seek to synthesize the two different, partial views into a complete and comprehensive view of the whole house. It is the same when we turn to ethical approaches. Each approach offers a view of a different aspect of the action; reversibility focuses on the internal dimension of the action (the formal characteristics of consistency, reversibility, and universality), harm on the outer dimension (its consequences or results), and publicity on the agent (the action provides a window into the agent's soul). This analogy helps convey the importance of employing all approaches, using each to address the limits of the other, and harmonizing any differences that may emerge.

Problem: Sometimes, students will make their decisions based on non-ethical grounds such as expedience (a failed use of feasibility) or self-interest (a failed use of harm/beneficence). Having made the decision on these grounds, they will try to dress it up using the ethics tests.

Remedy: Provide a more probing analysis of the ethics of their solution. Often other students in the class will help.

Here are four uses for these ethics tests:

1. They can be used to uncover ethical problems embedded in concrete situations. For example, the reversibility test asks us to consider whether any rights are placed at risk by the contemplated course of action. The harm test prompts us to look for risks that are embedded in a given situation, and those likely to bear them. The publicity test helps us to see if the integrity of the agent is at risk in a given situation. Hence, the ethics tests help us to formulate ethical problems and to identify stakeholders and their stakes.
2. The ethics tests can also be used to evaluate alternatives of action. Consider the following scenario: Your supervisor asks you to dump a drum full of toxic chemicals out in the field behind the plant. Should you do it, refuse to do it, refuse to do it and turn your supervisor in to the local authorities, or resign and get a new job? The ethics tests provide us with standards we can use to evaluate these alternatives. For example, doing as your supervisor asks could (1) harm the environment and the health of people who live near the plant, (2) violate the reversibility test since it ignores their rights (especially if done secretly and without their consent), and (3) be viewed as cowardly by others were it to be publicly displayed.
3. The ethics tests also play a fundamental, constitutive role in constructing solutions to ethical problems that arise in the real world. For example in the above scenario, you would envision a solution that minimizes harm, is reversible with all stakeholders and preserves your integrity. With this in mind, you could inform your supervisor that dumping these chemicals is illegal and could be readily traced back to the company. Then you could provide concrete suggestions for redesigning the manufacturing process so that it would not produce toxic byproducts. The manufacturing process, through this new design, would become informed by ethical considerations: it would minimize harm, respect the public's right to a safe environment, and maintain the integrity of the agents involved. Here the ethics tests would play a constitutive role by shaping the very nature of the solution.
4. Finally, these tests provide reasons that can be appealed to in making ethical arguments. In the above scenario, you have reasons for refusing to carry out your supervisor's order: for example, it would produce harm (violate the harm test), place others at risk (which is not reversible), and expose you as a cowardly person who callously exposes others to severely diminished environmental and personal health. Students, armed with these standards, can use them to justify their arguments. Hence, the ethics tests structure and guide class discussion.

We have still dodged the question about whether the ethics tests are good versions of the more complex ethical theories they represent. A simple answer is to capitulate on this question. Of course they do not represent the full complexity of these theories.

For example, the Harm/Beneficence test ignores many large problems with consequentialist theories. Including beneficence ignores the difficulty of requiring perfection in decision and action (is one obligated always to do the best thing for all humankind?). The test is phrased as an act-utilitarian approach, one that has been shown to be unworkable in practice because it requires (perhaps infinite) deliberation before every act. Rule-utilitarian approaches skirt this difficulty, but have their own problems

We might do this exercise with the other two ethics tests and come to similar conclusions. Thus, if the only goal of a course is to teach students the complexities of these ethical approaches, we would not recommend using these tests (though our perverse imaginations suggests ways even this might be done). These tests are designed for applied courses whose goals are to help students learn to think ethically within their profession (see our comments on goals in ethics courses). [link]

There is good reason to believe that it is a bad idea to treat these ethical approaches as foundational rules from which correct answers can be derived (see for example, Stephen Tolumin "The Tyranny of Principles," The Hastings Center Report, 1981; Stephen Tolumin, "How Medicine Saved the Life of Ethics, " Perspectives in Biology and Medicine, 25, 4, Summer 1982). As you can see from this essay, this is not the way we think these tests should be used. And therefore we think that their direct correspondence to the ethical approaches becomes less of an issue. These are guidelines, not first principles in a deductive solution.

Finally, we do not think these test should be used alone. Much research on moral reasoning suggests that expert in professional ethics spend much of their time thinking at an "intermediate" level about the ethical issues at stake in a case (see for example Keefer M. & Ashley, K.D. (2001) Case-based Approaches to Professional Ethics: A Systematic Comparison of Students' and ethicists' Moral Reasoning. Journal of Moral Education 30: 377-398; Bebeau, M.J., & Thoma, S.J. (1999) Intermediate Concepts and the Connection to Moral Education. Educational Psychology Review, 11(4): 343-360.). So, we encourage instructors to mix these tests up with exercises that emphasize these intermediate concepts.

Another reason these tests should not be used alone is that they are not even exhaustive of the possible approaches to thinking about the ethics of a situation. In the socio-technical analyses of the cases, we encourage students to think about how power is used in the systems that the cases involve. This concentration on the appropriate (just, careful) use of power is a central element in feminist ethics. We did not add a "power test" to the list because we felt the list was already too long. But the careful instructor will make sure students think about their use of the power that comes with being a professional.