The Agent-Age Problem for Consequentialism

Suspend your disbelief for a moment, and imagine the 6-year-old daughter of a major world leader travels with her father to a major nuclear launch site. She is left unsupervised, and happens to wander into the launch room. There, out of curiosity, she presses the big red button.

This launches a nuclear weapon that immediately kills millions of people. Before the weapon has even detonated, other nations have launched missiles of their own. A single launch rapidly escalates to nuclear war. Billions of humans and nonhumans are killed, and the planet is left barely habitable.

This scenario is clearly implausible to the point of impossibility – the big red button, after all, doesn’t even exist. However, it is a useful archetype that raises serious questions for consequentialism. Consequentialism inadequacy in certain moral issues is clear when the accidental action of a small child leads to immense suffering. I’ll add another example that deals with similar issues, but that is far more likely.  

A young boy happens to find a few matches on the floor of his family’s garage. While playing with them and scraping them across the rough floor, one of them ignites. In panic, the child rushes to the garbage and throws the match in. Then, losing interest, he walks inside and finds something else to do. The match lights a fire into the garbage which spreads into the house. The house burns to the ground, killing everyone inside. The fire spreads to nearby houses and kills or injures several more people.

This type of counterexample to consequentialism is demonstrably plausible, as there is empirical documentation of similar cases. According to the Washington Post, at least 265 Americans were accidently shot by children in 2015. Many of these shootings resulted in tragic deaths. Meanwhile, the number of American fatalities due to terrorism in 2015 was about 20, depending on certain counting methods.

In a truly consequentialist atmosphere, accidental shootings by children would be discussed far more than terror attacks – precisely 13.25 times as much. Moral deliberation on an action would be indexed to the amount of pain or happiness caused by the action. But in reality, the ethical issue of terrorism is discussed prolifically, while accidental shootings by children are virtually ignored. Why is this the case? I argue that while the amount of discussion on terrorism doesn’t reflect consequentialism, it does reflect our moral intuitions. We assign greater condemnation to actions not based on the numerical impact of these actions, but based on the intention of the actor, the nature of the action, and the emotional impact of the action.

The probability of child accidents will only increase in our modern world, as dangerous technologies proliferate and become more available, the population expands, and systems become more interconnected. A single accidental action by a child can result in unfathomable pain. However, our moral intuitions indicate that accidental actions by children are not blameworthy. Can consequentialism reconcile this problem?


I will use this definition of consequentialism, based loosely on the one from Stanford Encyclopedia of Philosophy:

“The belief that consequences are the only normative property that affects the rightness of an action.”

Or, in simpler terms, an action is made right or wrong only by its consequences. Consequences are the only morally relevant consideration.

Thus, if this essay shows that there are non-consequential normative properties that affect the rightness of an action, then consequentialism is false. I will use primarily an intuitionist approach to prove this claim – that is, showing that consequentialism is incompatible with clear moral intuitions. I will not touch on whether intuitionism is true; I will just discuss the consequences of its assumed truth.

Normative properties are defined as any ethical aspects of an action. This is a simple and non-rigorous definition that would be considered inadequate by many metaethicists, but it works well for the essay. For example, “rightness of intent” is a normative property, as it is an aspect that could impact the ethics of the action. Furthermore, this would be a non-consequential normative property.

What are the relevant normative properties in the examples above? I will consider the following:

  1. Intent – the actor’s purpose or intended goal in a certain action.
  2. Actor – the individual who commits the act.
  3. Consequence – the morally relevant impacts of the action.

Different moral theories place different emphases on these properties; consequentialism is the theory that only the third property is relevant to the rightness of an action.

In the case of the child pressing the red button, I believe we have clear answers as to the ‘value’ of these properties. The consequences are certainly bad. The intent is morally indifferent, as the child did not intend for anyone to suffer nor for anyone to benefit from her pressing the button.

The most interesting property is the second. Our moral intuitions agree that the age of the actor is morally significant. If a child commits a crime, they are considered less morally responsible than adults. This intuition is ingrained in law – individuals are not usually morally responsible until the age of 18. Some religions have an ‘age of accountability,’ which makes people accountable to God for sins after it is reached. Since children are less capable of complex moral reasoning, they are less responsible for mistakes in this reasoning.

Furthermore, there are also arguments for the moral relevance of the age of the actor that are not based on intuitions. For example, the following deductive argument:

P1. One is not morally responsible for what one does not know.

P2. If one is not morally responsible for what one does not know, then people who know less than others are less morally responsible.

P3. Children know less than adults.

P4. Children are less morally responsible than adults.

Thus, when the child presses the red button, and she does not know that this will fire a nuclear weapon, she is not morally responsible for the nuclear war that ensues. This argument attempts to prove that children in general are less responsible, but it can also be applied in any case where lack of knowledge is involved. If someone does not know the consequences or nature of an action, they are not morally responsible for this action.

Based on clear moral intuitions and the above deductive argument, the action of pressing the button is either (1) less wrong or (2) morally indifferent when the actor is very young or when the actor does not know the consequences. Either case means that the actor – a non-consequential normative property – affects the rightness of the action, disproving consequentialism.

Leave a Reply

Your email address will not be published. Required fields are marked *