The minimum complaint theory and maximum autonomy

The minimum complaint theory

The minimum complaint theory is a new moral theory that avoids a lot of problems in welfare ethics and population ethics. It is probably the strongest ethical theory in the sense that it can be most preferred by everyone and is the easiest to defend against objections and complaints. According to this theory, we should choose the option that generates the least amount of valid complaints whereby a complaint measures how strong an affected person prefers his or her most preferred option above the actual chosen option.

Let’s express this in a more rigorous or technical way. In ethics we have to choose between different options. Options can be for example situations, behaviors or moral rules. The option set Oi consists of all possible situations Sj that we as moral agents can choose: Oi = {S1, S2, … Sj,…}. If we choose situation Sj, there will be a number N of sentient beings or persons who have personal preferences or utilities.  Sentient beings are by definition able to value situations and subjectively want certain situations. They feel their personal preferences for different situations. Hence, the personal preferences are to be distinguished from unconscious interests. The preference or utility of a person P for a situation Sk is dependent on the option set Oi and the actual situation Sj in which this person has the preference for the possibly different situation Sk. So we can write the utility as a function UP(Oi,Sj,Sk). This is the preference of a person in situation Sj for a possibly different situation Sk.

Note that the person P who has the utility UP(Oi,Sj,Sk) for situation Sk does not have to exist in situation Sk. For example I can prefer a situation or a world where everyone is maximally happy, even if in that situation I do not exist. Furthermore, if the person P does not exist in the actual situation Sj, his or her utility UP(Oi,Sj,Sk) equals 0. It is as if that person exists but has no preferences.

Given the option set Oi and the situation Sj, we can look for the possible situation Sk* which is an element of Oi that has the highest utility for person P. So person P in situation Sj has the highest preference for situation Sk*, where this preferences equals UP(Oi,Sj,Sk*). The utility UP(Oi,Sj,Sj) is the preference of person P in situation Sj for situation Sj itself. The difference between UP(Oi,Sj,Sk*) and UP(Oi,Sj,Sj) is the complaint of person P in situation Sj. It is the difference in preference between the most preferred and the actual situation. In other words, we can write the complaint CP(Oi,Sj) as the maximum of UP(Oi,Sj,Sk)-UP(Oi,Sj,Sj) over all possible situations Sk. The total complaint in situation Sj is the sum of all CP(Oi,Sj) over all persons in situation Sj. We have to choose the situation Sj where the total sum of everyone’s strongest complaints (the total amount of complaint) is minimized.

So far the abstract theory. What are its implications and how is this compatible with maximum autonomy?

Maximum autonomy

The basic elements of the minimum complaint theory, are the personal preferences of persons. The personal preferences should be distinguished from projected preferences: I can project my preferences on someone else, comparable to anthropomorphism of non-human animals. The personal preferences of a person are the preferences that the person holds himself or herself, not the preferences that someone else projects on that person. Projected preferences violate the autonomy of a person.

A traditional consequentialist welfare ethic tries to improve everyone’s well-being. This requires an answer to the question what is well-being. With some thought-experiments we can try to figure out this notion of well-being. Take for example Robert Nozicks thought-experiment of the experience machine: a kind of virtual reality machine that can generate all positive experiences that you like. If well-being is simply the collection of positive experiences that you like, then it would be good (in terms of increasing well-being) if you plug yourself into this machine. If you are reluctant to do so, you may have other preferences that are not satisfied with this experience machine. We can argue whether those other preferences are real or rational, but the minimum complaint theory avoids this discussion: it is up to the persons themselves to decide what they prefer. It is up to the persons themselves to decide whether they want to maximize their well-being and how they interpret or define well-being. This is an example where the minimum complaint theory respects the autonomy of all persons.

Suppose people prefer to increase their well-being. Suppose also that a person in situation Sj has a well-being WP(Sj) and that if that person would be in any other possible situation Sl, that person still believes that he or she will have a well being WP(Sj) in situation Sj. That means the utilities or preferences are functions of their levels of well-being, independent from the actual situation Sl. For all situations Sl we get: UP(Oi,Sl,Sj)= UP(Oi,Sj,Sj)=UP(WP(Sj)).

Now we can ask the question what people prefer in uncertain situations: suppose in situation Sj, person P has a 50% probability of having well-being 10 and 50% probability of having well-being 0. The expected well-being is 5.However, if the person is risk averse, that person might prefer another situation Sk in which that person has probability 100% (i.e. certainty) of having well-being 4. With this risk aversion, the utility function UP(WP) becomes a concave function of WP. This means that well-being has a decreasing marginal utility: the more well-being you have, the less utility an extra unit of well-being generates. The difference in utility between well-being 5 and well-being 0 is bigger than the difference in utility between well-being 10 and well-being 5.

If every person has the same utility function (the same risk aversion), we arrive at a prioritarian ethic. This prioritarianism says that we have to increase everyone’s well-being, giving a higher priority to increasing the well-being of the worst-off people, the persons who have the lowest levels of well-being. If risk aversion is 0 and every person’s utility function is linear in well-being, we arrive at a traditional interpretation of utilitarianism that maximizes the sum of everyone’s well-being.

We can argue whether risk aversion is morally relevant or rational, whether the utility function is concave or linear, but the minimum complaint theory avoids these discussions: it is up to the persons themselves to decide what they prefer, what their own utility functions look like. It is up to the persons themselves to decide their preferred level of risk aversion. Some people may have a linear utility function resembling zero risk aversion, some may have a highly concave utility function corresponding to a high risk aversion. This is another example where the minimum complaint theory respects the autonomy of all persons. Whether we arrive at a prioritarian, a traditional utilitarian or a mixed ethic is completely up to the persons themselves to decide. There is even more autonomy: the utility function of a person does not even have to be a function of a quantity called well-being.

The above discussion about well-being assumes that a person P in situation Sl is the same person as (or can identify with) a person in another situation Sj. However, this is not always self-evident: it leads us to a metaphysical problem of personal identity over different situations. Suppose I can choose between two situations: in situation S1 I will be tortured next year, in situation S2 I will be happy next year. Is my future state in situation S1 the same person as my future state in situation S2? How can I compare persons between different situations? It may even be possible that different situations have different numbers of persons. The minimum complaint theory avoids this problem. It is up to a person P in situation Sj to decide whether a person in another situation Sl is the same person. Persons can decide with whom they prefer to identify themselves. A person P who exists in situation Sj can even prefer another situation Sl where that person does not exist or will soon die. If I could choose a situation where everyone is maximally happy, except for the fact that I will die or I will be brainwashed and become a completely different person, I will prefer that situation, even if there is no-one in that situation I can identify myself with.

The above discussion about well-being also assumes that the preference UP(Oi,Sl,Sj)=UP(Oi,Sj,Sj) for all situations Sl. However, it is possible that those preferences are different. In fact, the person P in situation Sl does not necessarily have to be identical to the person P in situation Sj. It is up to the person P in situation Sl to decide whether he or she identifies with the other person in situation Sj and whether he or she takes that person’s utility UP(Oi,Sj,Sj) as his or her own utility UP(Oi,Sl,Sj).

What about inconsistencies in preferences? There are two kinds of inconsistencies: synchronous and intertemporal. Intertemporal inconsistencies are easy to deal with: suppose I strongly prefer X at time t and no longer prefer X at a later time t+1. If that is the case, I can say that at time t+1 I am another person than at time t. A person can be fully described or defined by the set of all his or her preferences. If preferences change over time, we have to consider this as if there are multiple persons, one person for each moment of time. To be safe, we have to assume that every other moment in time there are different persons. The utility Up of a person P refers to the utility of a person at a specific time t, so we should write Pt to denote a person at time t. At another time, we have another person. It is true that at a moment t a person Pt can identify himself or herself with a single future person Pt+1 at a later time t+1, and this identification influences the preferences of the person Pt. If future technologies allow for duplicating brains or copying minds, it is even possible that a person can identify himself or herself with multiple future persons. The minimum complaint theory respects autonomy again: it is up to the persons themselves to decide with whom they identify themselves with and how that identification influences their preferences.

What about synchronous inconsistencies in preferences? Suppose a friend lies to you, you will never be able to know that it was a lie, and if you know the truth, you would be very unhappy. You prefer to be happy, so are lies permissible in such situations? Not necessarily, because you may also prefer to know the truth. So you may have two conflicting preferences: one to avoid unhappiness and one to know the truth. The minimum complaint theory respects your autonomy: it is up to you what you most strongly prefer. Your preference for a certain situation (e.g. the situation where friends tell the truth) should be based on the most coherent set of your strongest preferences. Only if a contradiction remains, if you insist on both preferring and not preferring a certain situation at the same time, other people can decide for you.

Next we have the problem of choice independence. Suppose I am faced with option set O1 that consists of two situations S1 and S2. I prefer S1 above S2. However, suppose a third option S3 arises, so I have a new choice set O2 that consists of the three situations. It is possible that adding this third option changes my order of preferences and I now prefer S2 above S1. More generally: UP(Oi,Sj,Sk) may be different from UP(Oh,Sj,Sk). This means there is a choice dependence. We can argue whether this choice dependence is morally relevant or rational, but the minimum complaint theory avoids this discussion: it is up to the persons themselves to decide what they prefer given the option set.

Finally we come to the question of validity of complaints. Suppose in situation Sj a person P most strongly prefers the situation Sk*, so UP(Oi,Sj,Sk*) is the maximum and UP(Oi,Sj,Sk*)-UP(Oi,Sj,Sj) is the complaint. We add this complaint to the complaints of all other persons and then we look for the optimal situation S° that minimizes this total complaint. However, it is possible that person P in situation Sj does not prefer this situation S°, so UP(Oi,Sj,S°) < UP(Oi,Sj,Sj). Now it becomes tricky for that person: if that person complains, the situation S° will be chosen and the person P will be worse off. Again the minimum complaint theory respects autonomy: the person P in situation Sj can decide to change his or her preferences, for example to set UP(Oi,Sj,Sj)=UP(Oi,Sj,Sk*). In this way, that person will not complain in situation Sj and the total complaint will no longer include the complaint of person P. That means another optimal situation will be chosen, one that person P in situation Sj can prefer.

Why is this relevant? If we simply want to minimize complaints, we might have to prefer the situation where no-one (no future generation) exists, because existing people can always complain. If only one person exists, that person might feel lonely and prefer a situation with more people. So consider all possible situations that contain at least two persons. Those people have to distribute finite resources amongst themselves, which means in every situation there is always at least one person who can complain. The only situation with no complaints, is the situation where no-one exists, because UP(Oi,Sl,Sk)=0 if person P does not exist in situation Sl. If that person exists in situation Sj and has a positive utility UP(Oi,Sj,Sj), and if some or all other possible situations that have a smaller total complaint are situations in which the person does not exist, that person might decide not to complain in situation Sj. That person could set preferences equal to UP(Oi,Sj,Sj)=UP(Oi,Sj,Sk*).

This can be made more general: the complaint of a person is calculated as the difference between the preference for a reference situation Sk*and the preference for the actual situation Sj, whereby the reference situation Sk* is the most preferred situation. However, to respect full autonomy, we can say that the person can decide what counts as the reference situation. It doesn’t have to be the most preferred situation. It can also be for example the least preferred situation, in which case the complaint becomes negative: it becomes an anticomplaint or gratitude. Or it can be the situation for which the person has 0 preference.

If everyone took as the reference situation a situation for which they have 0 preference, we end up minimizing the sum over all persons of -UP(Oi,Sj,Sj), which equals the maximization of the sum of everyone’s preferences UP(Oi,Sj,Sj). This is nothing but total utilitarianism. If everyone took as reference situation a situation for which they all have a non-zero, positive preference C, we end up with a so-called critical level utilitarianism that maximizes the relative utilities UP(Oi,Sj,Sj)-C.

Minimizing the sum of UP(Oi,Sj,SR)-UP(Oi,Sj,Sj), where SR is a reference situation chosen by person P in situation Sj, the general minimum complaint theory becomes a maximum relative preference theory that maximizes the sum of relative preferences UP(Oi,Sj,Sj)-UP(Oi,Sj,SR). This theory is a utilitarian theory with a few extra degrees of freedom. Total utilitarianism means we have to maximize the total of utilities UP(Oi,Sj,Sj)-0, summed over all persons P. Critical level utilitarianism maximizes the sum of UP(Oi,Sj,Sj)-C. The relative preferences theory maximizes the sum of the relative utilities UP(Oi,Sj,Sj)-UP(Oi,Sj,SR), where UP(Oi,Sj,SR) is the preference of person P in situation Sj for a reference situation SR. That reference situation can be the most preferred situation, which results in the minimum complaint theory. It can be the least preferred situation, which results in the maximum gratitude theory. It can be the situation for which person P in situation Sj has a preference 0 or C. If all persons in all situations choose as a reference situation a situation for which they had preference 0, then we have total utilitarianism. Basically it means that if everyone is a total utilitarian, then we should maximize total utility. But in the maximum relative preferences theory, nothing prevents them from choosing other reference situations, and those reference situations can be different for different persons and different situations Sj.

Implications of the minimum complaint theory

The minimum complaint theory has many advantages. First, as we have seen, it maximally respects autonomy and therefore avoids many problems about for example what counts as well-being and how much priority should be given to the worst-off.

Second, the minimum complaint theory avoids many problems in population ethics that other theories such as total utilitarianism and average utilitarianism have to deal with. Consider total utilitarianism that wants to maximize well-being or preference satisfaction. As we have seen, the general minimum complaint theory allows for individuals to choose their reference situation in order to calculate their complaints or gratitudes. If everyone chose as a reference the situation for which they have 0 preference, the general minimum complaint theory turns into a total utilitarianism. If that is the case, it means that everyone accepts the implications of total utilitarianism, including a conclusion that is repugnant or counter-intuitive to many actual people. According to total utilitarianism, a situation S1 that contains thousand maximally happy people (who have maximum preference satisfaction) is worse than a situation S2 where those same thousand people are maximally unhappy and where an extra huge number of people exist who have lives barely worth living (i.e. slightly positive preference satisfaction). If the number of extra people is large enough, the total well-being or preference satisfaction is higher in the second situation, but saying that this situation is better than the first is very counter-intuitive, repugnant or sadistic.

In the second situation, average well-being is much lower than in the first situation, so an average utilitarian prefers the first situation. However, the alternative theory of average utilitarianism faces other counter-intuitive sadistic conclusions. Consider a situation S1 with a million maximally happy people and one person who has a highly negative well-being. The average well-being is close to maximum. We can choose situation S2 in which this unhappy person also has a maximally happy life, but suppose this is only possible if we add a huge number of extra people who have very high but not maximum levels of well-being. The average well-being in the first situation is higher than in the second, but saying that the first situation is better is counter-intuitive: in the second situation every person is at least as good off as in the first situation, and the total amount of well-being is much higher in the second situation.

It is easy to see why the minimum complaint theory avoids these problems (at least when most people choose their most preferred situation as the reference situation to calculate complaints). The persons who are very unhappy have very strong complaints, and the other people cannot complain. If we choose the minimum complaint theory, we can look for the most pressing problems in the world according to this theory. In terms of scale of the problem, the biggest problem is probably the suffering of wild animals in nature. It is probable that a lot (perhaps a majority) of animals have negative preferences for their situations in the wild, because they experience mostly negative experiences. If in future generations of animals there are animals with negative preferences (i.e. lives not worth living, having a preference for non-existence), their total complaints can add up to a very big number. That is why the problem of wild animal suffering deserves an almost absolute priority. We have to figure out what animals are sentient, what their preferences are, how strong their preferences are and what we can do to decrease their complaints. See a rational approach to improve worldwide well-being.

Problems with the minimum complaint theory

We finally have to face the major problems of the minimum complaint theory, which are basically the same fundamental problems of all moral theories. First we have the problem of communication: how can a person or sentient being communicate his or her preferences? Some persons (animals, toddlers, mentally disabled humans) are not able to clearly articulate their preferences. Second we have the problem of interpersonal comparison: how can we compare the preferences and complaints of different persons?

The first problem can in principle be solved through science. For example behavioral economists have experience in revealing a person’s preferences and these methods are also applied to some animals. We can see what chickens prefer when they face different choices. Neuroscience can also help us with this problem. Still, it is difficult and for the moment impossible to know if the preference of an animal or a mentally disabled human for being alive in a specific situation is positive or negative. How can we know if another person’s life is worth living for that person?

The second problem is far more serious. We have no idea, not even in principle, how to compare the preferences or levels of well-being between different persons. There is no universal unit of preference, no measure or scale that all persons can use to compare their preferences with each other. If you say you prefer A twice as much as B, how can I know how big the difference is between your preferences of A and B?

There is only one solution at this moment: all moral agents have to do their best to estimate the preferences of all other sentient beings as unbiased and impartial as possible. The moral agents are forced to do this exercise, because they are the ones who have to make the decisions, they are going to choose the optimal situation. So here we cannot avoid deviating from respecting autonomy. Moral agents have to judge what other individuals want and how strong they complain. If moral agents disagree in those judgments, a kind of democratic parliamentary system is required.

Dit bericht werd geplaatst in Artikels, Blog, English texts en getagged met , . Maak dit favoriet permalink.

Een reactie op The minimum complaint theory and maximum autonomy

  1. Pingback: Variable critical level utilitarianism as the solution to population ethics | Stijn Bruers, the rational ethicist

Geef een reactie

Vul je gegevens in of klik op een icoon om in te loggen. logo

Je reageert onder je account. Log uit /  Bijwerken )

Google+ photo

Je reageert onder je Google+ account. Log uit /  Bijwerken )


Je reageert onder je Twitter account. Log uit /  Bijwerken )

Facebook foto

Je reageert onder je Facebook account. Log uit /  Bijwerken )


Verbinden met %s