The Trolly Problem

The Trolley Problem is a (now) classic problem that forces people to make a serious moral decision. The general format of the problem is as follows: 

You are alone at the controls of a runaway trolley barreling toward five people who would most certainly die. The brakes and controls (horn, doors, etc) do not work – but there is a lever in the trolley that would divert the trolley to another track that has just one person on it who would, likewise, also most certainly die. What is the right thing to do?

This is a complex ethical problem that has been resurfacing in the discussion of AI-driven cars (Marshal, 2018). A modern and soon-to-be real world version of the question would be if a car was going to collide with a pedestrian and could hit a group of people in the crosswalk or one person on the sidewalk, which should it be programmed to pick? Or if the car was about to hit either a young child or an old person, who would it decide to hit to save the other? These are going to be important questions in the evolving AI self-driving cars and trucks as well as the various other AI-driven software but is also an important question to pose to put anyone in a position where they must stretch their usually unused ethical muscles. 

Responses from most people makes it seem as if a solution for the trolley problem is not only possible, but also easy! Just like all complex ethical questions, the solution requires “one to reason through more than one ethical perspective, and come to reasoned ethical judgments” (Paul and Elder, 25). So before one jumps toward the first seemingly good decision, the correct way to reason this ethical problem is to identify and carefully consider every key concept and principle that cold be applied. So let us first explore the various ethical principles of consequentialism, utilitarianism, contextualism, deontology, kantian ethics, virtue ethics, and justice.

There are two general solutions to this problem: pull the lever and switch tracks – sparing the five people but hitting the one; or do not pull the lever and do no switch tracks – killing the five and sparing the one. It is also important to note that one does not really know what they would do in out-of-the-ordinary or high-stress situations no matter how much they think about it. Take Chidi Anagonye from The Good Place for instance when he was put in a trolly problem simulation where he had to decide whether to pull the lever. He was an educated ethicist but froze when put on the spot. In the same manner, we would not know what to do, but given the hypothetical scenario, I would confidently say I would pull the lever and divert the cart killing one person. This problem is, in essence, and exercise of utilitarianism versus deontology. For this reason, one should explore the consequentialist reasoning first.


Consequentialism is an ethical theory that stresses the consequences of action or inaction holding that “all actions are right or wrong in virtue of the value of their consequences” (Honderich, 154). So, in general, “consequentialism provides a very simple theory” that states that “an act is morally right (or morally permissible) if and only if it produces the best consequences”. Additionally, consequentialist would hold that an actor is “morally required to perform the act with the best consequences” meaning not only is it good to switch the lever, it is morally required to be ethical.

The first problem with the consequentialist reasoning is that the standard at which a community or culture judges the consequences and the actions to reach those consequences. People and groups of people have and do use consequentialism to justify horrible atrocities especially when they embrace superstition, hatred, fear to solve their problems. One such example is the witch trials of the 17th century. Those witch hunters and supporters “truly believe that witches cause disease, crop failures, sickness, catastrophes, and accidents” to such an extent that, to them, “it is not only a rational act to burn witches, it is a moral duty” to kill those few in order to save the many (Shermer, 2018). This is just one example of the evils that can be committed with consequentialism as the justification, but it is also important to acknowledge the effects of the actions in the name of consequentialism. No one really knows, or could know exactly the consequences of our actions. Not through reason nor through past experiences because, as Burtrand Russell explains that “even if experiences have told us that past futures resembled past pasts, we cannot conclude that future futures will resemble future pasts” with total certainty because our beliefs “are the product not of reason but the imagination” (Honderich, 378). 

Consequentialists in general would decide that moving the trolley to spare the five and kill the would is not only moral and ethical, but it is morally required of the operator to do. And, although consequentialism can be used to commit atrocities and is flawed in that no person can, for certain, know the consequences of their actions, in this situation – they seem to be onto something. To dig deeper into this, one must understand the consequentialist principle of utilitarianism.


Along the same vein as consequentialism, utilitarianism also focuses on the consequences of actions rather than rules and ethical sentiments – but also asserts, “actions are right in proportion as they tend to promote overall human happiness”. John Stuart Mills, famous for his work Utilitarianism, works with a theory of life that says that there is only one thing intrinsically and naturally desirable: pleasure. Also known as qualitative hedonism, Mills thought some types of pleasures are outweigh other types and should hold greater weight in determining the correct way to act. Just as consequentialists believe we have a moral obligation to act, hedonists believe that “it is our duty to judge all such cases by measuring pleasures against pains, with a view to their respective assets and liabilities” (Epicuris, 64). 

In this way, utilitarians value certain pleasures more than others and certainly the amount of pleasure and, by extension, the avoidance of pain. So, in general, this utility, that is to be weighed “holds that actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness” – this Greatest Happiness Principle continues to say “the ultimate end … is an existence exempt as far as possible from pain, and as rich as possible in enjoyments, not in point of quantity and quality”. 

There are more terms to get into the reasoning behind utilitarianism. The first is ethical hedonism which asserts that the only ultimate good is pleasure. The second is hedonistic utilitarianism which defines utility “in terms of pleasure and pain”. These two distinctions in consequentialist utilitarianism came together under John Stuart Mill who claimed that “actions are right in proportion as they tend to promote happiness.

Mill’s utilitarianism is often bent to mean that one must act in favor of the majority so that the greatest number of people are benefitted, but this is an incorrect and rash jump. This mistake is what leads to the suffering of a minority for the benefit of the majority. But it needs to be stressed that “utilitarianism simply directs us to perform the act that will result in the outcome with the greatest total amount of well-being, and this might well be an outcome that benefits a few individuals a great deal rather than many individuals very little each”. Another way to think about it is that utilitarianism considers the pleasure, pain, and wellbeing of everyone in an absolute value – that is, if an action benefits a group of people (majority or otherwise), one must also take into consideration any action that lowers pleasure or even increases pain in others (majority or otherwise).

With a firm understanding of the basics of utilitarianism, it is clear that they hold that pleasure is the highest intrinsic good and that people have a responsibility to act in a way to bring about the most pleasure. This requires the actor to measure the consequences and utility, calculate how much utility one is obligated to produce, and figure out what things are to be judged in their consequences. 

Utilitarianism seeks to weigh the pleasures (hedons) and pain (dolors) and, much like Epicurean hedonism, this weighing and judging is not always easy, but rather part of the human condition, “the best that we can do is to try to increase our ability to foresee consequences” but it is not possible or reasonable to “hope that we can eliminate uncertainty completely” – we must do the best we can. Some actors would hold certain ideals higher than others such as justice, equality, liberty, etc but this ideal utilitarianism can bring in an abundant amount of subjective decision making by the actors involved because who is to say what ideal is better than another? Of course there are universal morals that most people in most places most of the time hold as good and moral, but with culture and ideological considerations, it is difficult to absolutely know what ideals should be held above any other. Besides this general weighing of pleasures, there is preference utilitarianism which seeks to assign rankings to actions and behaviors in an effort to mathematically weigh actions. But this effort does little to curb the subjective nature of deciding what actions are result in better or more pleasure. 

As far as deciding what things are to be judged, there are two general answers: acts and rules. In act utilitarianism, the actor must review their actions in terms of whether it promotes the most good by their consequences. Rule utilitarianism is a bit different in that it takes into consideration the rules (moral, societal, cultural, etc) stating that people should, “act in accordance with those rules that will produce the greatest overall amount of utility for society as a whole”. This seems to be a way one can claim to be in favor of the best possible good while also following rules – a pseudo-deontology of sorts. But, as we will see later with deontology in general, rule utilitarianism, and writing down morals in stone in general, will not lead to ethical decision making, but rather people acting in a way that simple follows the rules for the sake of following the rules – the same fallacy one can find rampant in religious and legal ethical reasoning (Paul and Elder, 15).

The problems with utilitarianism include many of the same included with consequentialism. No one knows the consequences of their actions, they do not know exactly the amount of pleasure or pain will result from an action, and immoral actions could be justified under the banner of utilitarianism. But Mills does a better job in defense than consequentialists do in that he holds that it is necessary to take everyone’s well-being in consideration regardless of the majority/minority numbers being affected. Additionally, when critisim arose that it would be easy to justify murder, genocide, and the like under utilitarianism. Assassinations, mercy kills, and defense could definitely produce more good so doesn’t that mean murder is permissable? Of course not. Unlike consequentialists, utilitarianists and Mills argue that people must assess “the actual, expected consequences of an action, only if we hypothetically consider that all would act in the same manner” – so a Kant-like universal maxim applied in a way where the actor must think what society would be like and how much good or pleasure would be produced if everyone was permitted to do the same thing. So if murder was permissable in a way where the actor assumes they are producing more good, there would be something of a murderous rampage.

The trolley problem is in essence a utilitarian exercise so it is definitely what reasoning most would use and seems to be the vastly more moral route. But, again, before we assume the right rational, we need to consider all points of view including deontology – which provides an alternative solution as well as good criticisms of consequentialism.


Deontology is a much different view of ethics and is a type of moral absolutism that asserts that “certain kinds of actions are intrinsically right or wrong – right or wrong simply because they are that kind of action” (Honderich, 2). So, where consequentialism focuses on the consequences of actions, deontology focuses on the act itself. So something that is ruled as wrong will always be wrong no matter the context it is in just because it is that act. Take killing for example –  consequentialists can justify killing if the consequence is good (killing a murderer, self-defense, honor killings, etc). Deontologists, by comparison, would say killing is always wrong because, well, killing is wrong!

In the context of our Trolley Problem, it is easy to see where deontologists stand: they would definitely would not choose to switch the tracks because they would be knowingly killing a person. To fully understand the flaw in deontology in regards to this Trolly Problem, let us explore a few conclusions a deontologist would consider in this, or other, ethical problems. First, of course, no rational person would agree that killing someone is right, but it is easy to see the flaw in this situation. By not switching the cart, they are allowing five people to die. Second, as we have see above, no one knows the consequences of their actions for certain. The five men may see the trolley and move, or diverting the trolley may some how kill many more later down the tracks. I have heard a deontologist simply assert that the trolley driver does not even have the right to move the trolley because they do not have the right to break the moral law of killing or even judge the stranger’s lives – they then leaned back and exclaimed “no dilemma at all!” 

It reminds one of the Parkland high school shootings where an armed security guard chose not to act. This was a trolley experiment of sorts where he could have killed the one shooter, or allow the shooter to kill. Any reasonable or virtuous person would say he had a duty to stop the killer. In the same respect, we would expect the driver to kill the one in favor of the five. 

In conclusion, we have a moral duty to help others, but we should refuse to participate in life tradeoff experiments. Those that do participate anyway, are liable for their actions. Yes, rules are rules and humans naturally value certain universal moral principles, but in the context of the problem, deontologist fails to realize we do not operate in a vacuum – we live in a community of people and our actions affect them. People generally have a moral obligation to produce the most good possible while avoiding the most possible pain.

Kantian ethics

Along a similar vein of deontology, Immanual Kant also held that actions in themselves are right and wrong. The genius of Kant extended beyond that principle in two extensions. First, Kant would also say to never use people as a means to an end – only as ends in and of themselves. That is to say, by “saving” the five, we are killing the one, using them as a means to save the five – a deontologist may also assert that the actor is not exactly “saving lives anyways” – they are simply deciding to eliminate the one life.

The other extension of Kant’s is his universal maxim which says that we should only act in a way we would wish everyone else to act – and would we want everyone to make the decision to kill the one? Again, the argument against consequentialism turns to a murderous rampage. This is where the maxim has unintended consequences. Like suicide or mass killers, the actor is in a unique situation where they would indeed pick to apply the rule to everyone. And in this case, those who choose to kill the one would, based on that reasoning, would definitely want others to make the same decision. 

So, although Kant provides some powerful and useful additions to deontology, the stringent rule-follow is a bit absurdist and sets a standard in which no one can live up to. Following rules like the Ten Commandments or other culturally inspired commandments become absurd in situations like this where people do not need to morally reason for themselves, instead, they follow rules to strive to an out-dated and sometimes immoral standard of which they cannot ever wish to achieve. Some hedonists like Epicuruse or even Mills would claim this standard would create more pain and decrease pleasure.


As mentioned before, when reasoning morally, one has to consider all facts in order to put the decisions in the context of the lives at stake. Even when murdering or breaking any other obvious social/ethical rule, one must put context into perspective. Lying, for example is a good thing when in the right context like when hiding a Jewish family in your attic and the Gestapo comes knocking. In the same way, choosing to kill is usually bad, and yet, it is good when to saves (saves is a bad word for this situation) or rather spare the other five. This situation happens too often and is always difficult. Since we are on the topic, the end of World War 2, the United States had to decide to invade or drop the atomic bombs and save lives on both sides.

Contextualism, or rather “moral contextualism holds that moral judgments involving expressions such as ‘good’, ‘acceptable’, and ‘required’ have context-sensitive truth conditions” where the judgement of an action as good “may be true in one sense and false in another” (Montminy, 2). Additionally, contextualism asserts that there really is not such a thing as being morally good at all. In fact, this idea holds that an action can only good judged as good relative to the moral standards that prevail in that specific situation (Montminy, 3). The benefit of this ethical principle is that it provides extreme flexibility to compensate for extreme circumstances. Moral contextualism is unique in that is places actions in terms of the context so its very nature, contextualism disagrees with deontology completely. This principle is very much applicable in most ethical problems and definitely in the trolley problem. Additionally, context matters in consequentialist views and absolutely agrees with utilitarianism because it holds that actions are not judged in blank and white lenses, but rather if they are more acceptable than another action. 

To finish with this principle, lets look at two types of broad behaviors. First, Preservationism is the idea that “the dominant response is correct” and what most people would do in that same situation is what is correct in anyone’s situation (Montminy, 4). The second is liberationism where a person has a duty to help people and do good no matter the situation or consensus. The former is often reliably useful in common low-demanding situations where the latter is often best in high-demand situations. Moral contextualism utilizes these two principles to put ethical and moral problems in the context of general consensus and need. This requires that the decision maker, in order to act morally, be aware of their moral values and ought to be moved by the two values (Montminy, 5). 

In the context of the Trolley Problem, contextualism is definitely applicable to consider in that the high-stakes situation and actions warrant it. Contextualism along with utilitarianism provides a suitable and well-rounded tool to acceptable reason the demanding and lethal decision making at hand.

Virtue Ethics

The virtuous person acts not because something is right or wrong, but because they know that is what a virtuous person would do. Trying to use virtue to solve a situation like the Trolley Problem is where the weaknesses of virtue ethics really comes to light. First of all, who is to say what a virtuous person would do? Virtue ethics use mostly habits to determine a virtuous person and provides no directional advice as to how to actually act. Additionally, virtues are largely subjective. Of course, there are universal morals and standards that most any reasonable person can agree to in most societies most of the time, but when put in a situation like this, the response as to how to act would vary widely. Virtue ethics is somewhat applicable, but extremely limited and flawed. It would be the last “tool” I would use to reason this problem, but ultimately, a virtuously ethical person would say they have a duty to act in a way that they would expect anyone of virtue to behave. In our society, most would choose to kill the one and be seen as a hero. Again, like those who choose to kill mass shooters in order to save more lives.


The decision to kill the one is definitely the best decision for most and most assuredly for consequentialists, but that is not to say that this decision is fair or justice. Justice requires that everyone gets their fair say in a decision. Many types of justice requires one to be apart of the decision making process, be represented in the process, or feel that they are being treated fairly, and in the case of the run-away trolley, no one has any sort of say in the decision making (Baldwin, 1-4). People have a natural need for justice – not just for ourselves, but for others. This can be described as an “animal desire to repel or retaliate a hurt or damage to oneself” or others and is founded on the “human capacity of enlarged sympathy”. Considering the nature of the exercise and that none of the would-be victims can not be consulted or represented before making a decision, there is no way that the end result can be fair or just to anyone involved.


I am sorry to say that the Trolley Problem is one that can never really be “solved” in a way everyone can agree. But if one were to consider and weigh the pleasure and pain involved, it is clear that the best solution is to pull the lever and switch the tracks. Consequentialism, and certainly utilitarianism, hold the best principles to provide the best outcome for the Trolley Problem. As mentioned before, these ethical principles are not always the best solutions for all ethical decisions. Indeed, consequentialism has been used to justify far too many atrocities, but so has deontology and any other ethical principle people seek to miss use to justify their actions.  It is not enough in serious ethical situations to simply follow rules and obey orders. It is the responsibility of everyone to act in a way that benefits everyone around them – and if not everyone can benefit, one must choose the act that results in the most happiness and pleasure. In this situation, utilitarianism with a healthy use of contextualism, is by far the best solution because it is leading to the best outcome and avoids the most pain.


Baldwin, Susanna, “Organizational Justice” Institute for Employment Studies (2006): 1 – 13.

“The Ethics of Consequences: Teleology” accessed September 23, 2019 from

Honderich, T. (1995). The Oxford Companion to Philosophy. Oxford: Oxford University Press.

“John Stuart Mill: Ethics” accessed September 23, 2019 from

Marshal, Aarin. “What can the Trolley Problem Teach Self-Driving Car Engineers?” Wired October 24, 2018. Accessed September 25, 2019 from

Montminy, Martin, “Moral Contextualism and the Norms for Moral Conduct” American Philosophical Quarterly (2007).

Paul, R., & Elder, L. (2006). The Thinker’s Guide to Understanding the Foundations of Ethical Reasoning (2nd ed.). Tomales, CA: Foundation for Critical Thinking Press.

Shermer, Michael. “Moral Philosophy and its Discontents” Skeptic Society, May 2018. Accessed September 24, 2019 from

Tang, Qiang. (2013, April 4). Is there a Final Solution for the Trolley problem? ResearchGate Forum. Retrieved September 25, 2019 from

Featured Image: Lionel Walden Cardiff Docks (1896)

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Powered by

Up ↑

Create your website with
Get started
%d bloggers like this: