What criteria should be used to decide whom to include or exclude from a utilitarian welfare estimation? To use Peter Singer’s metaphor, how far should the circle be expanded?
For example, considering the welfare of the next, soon to be born, generation doesn't seem exaggerated. But considering the welfare of all future generations seems an unreasonable stretch (it would dwarf the consideration of our own welfare). This could be addressed by using a discount rate, but what criteria should be used for estimating the discount rate and why not others?
Similarly, many pay lip service to the idea of considering all humans as part of the circle, but as a matter of fact, that is not the case (i.e. regardless of what they say, many don't actually consider this expansion to be reasonable, as evidenced by their actions). Should we consider (some) animals as well? One inclusion criterion proposed here is the ability to suffer. But this doesn't seem to work because it follows that we should consider all future generations' welfare as equally important to our own (no discount rate).
One possible solution is sociobiological - take genetic relatedness as the basis for the discount rate (both when expanding the circle to existing beings and when expanding it into the future). Care about the immediate family more than about strangers, care about your own children more than about your great-great-grand-children. However, this leads to a very small circle - smaller than what people now normally consider.
We could consider some multiplier to genetic relatedness, but, again, what multiplier should be used and why?
A simpler question is this: if this model is correct, what is the actual multiplier that most people use? Is there just one or are there many?
One possible multiplier that immediately comes to mind is familiarity. This can be applied both to animals (even objects) as well as to humans. Many people care more about their dog than about dying children in Africa, although they are more genetically related to the latter – the familiarity is less however.
From a evolutionary point of view, the importance of familiarity can be understood by considering that, in the original hunter-gatherer environment, familiarity used to be a proxy for genetic relatedness. However, these two factors seem to be independent, as evidenced by the fact that adopted persons care to meet their biological parents (with whom they have no familiarity at all). Thus, genetic relatedness cannot be considered to be simply incorporated into familiarity.
Considering the product between familiarity and genetic relatedness can also be applied in regard to future generations. The prediction however fails. We are less familiar today with the future generations than medieval people were with their future generations, because we expect things to change radically, while they expected things to remain the same. This leads to the prediction that the welfare of future generations should have been much more salient to medieval people than it is to us. However, as it was pointed out by Gregory Clark, medieval people lived in the present to a much larger extent than we do, as evidenced by the size of interest rates then and now.
This problem could be fixed by assuming another multiplier, the present welfare of the person we’re considering. Medieval people were very poor and miserable compared to us, and, thus, they tended to care more about themselves than about others. This multiplier can be justified from an economic point of view: the means one has at one’s disposal are scarce and these scarce means are distributed to the most important goals first. The more means one has, the more goals one is able to fulfill, i.e. increased welfare implies an expansion of the area of interests, including an expansion of Singer’s circle.
What I said so far is still not enough, as it predicts that we care equally about common people and about exceptional people. As evidenced by newspaper and mass media choice of subjects, there is little demand for regular people stories as compared to the demand for stories with “celebrities” and unusual people in general. (Also, applied to objects, what I said so far, fails to account for the concept of “garbage” – something that is so abundant that one is willing to pay to get rid of it.) We can fix this by dividing to the availability of what we’re considering – the more common she/he/it is (including oneself) the less one cares about her/him/it. This might also explain why people are horrified by the idea of cloning themselves – cloning would increase their availability and thus decrease the amount they care about themselves.
Here is thus a possible formula for the amount X cares about Y:
CX(Y) = (GXY + 1) × K × FX(Y) × WX / AX(Y)
GXY = genetic relatedness between X and Y (a number between 0 and 1, 0 corresponding to non-living objects and 1 to oneself and one’s identical tween; parents and siblings 1/2; cousins, uncles and aunts 1/4 etc.),
FX(Y) = X’s familiarity with Y (a number between 0 and 1), which is to be multiplied by some factor K in order to account for the fact that we care about some familiar animals more than about some unfamiliar people (i.e. familiarity and genetic relatedness don’t have the same measuring unit and K is the conversion factor),
WX = X’s current well-being (a number between 0 and infinity),
AX(Y) = the availability of Y to X.
There are further predictions of this formula. For example, the more one knows oneself, FX(X), the more one cares about oneself, CX(X). Thus, the problems of suicidal people [WX small –> CX(X) small] may be compounded if they are self-delusional [FX(X) small –> CX(X) small]. Also, people with low well-being don’t care much about anything [WX small –> CX(Y) for all Y]. And, conversely, the better one feels, the more one cares about everything.
We can also apply this to other things. For example, according to the above formula, the more one learns about a subject (familiarity increases with the subject), the more one starts to care about that subject. People who start studying something without much initial interest, may end up caring a lot about it. I suspect this is what happens to many students who don’t have a clear idea from the start about what they are interested. They think that are “searching” for what interests them, but it may be that in fact the interest is created rather than discovered. And those who keep “searching” never “find” – i.e. develop – any interest because they don’t get sufficiently familiar with anything. The issue of availability may also influence matters: getting familiar with relatively obscure subjects increases the likelihood of ending up caring a lot about them. The prediction thus is that one tends to become passionate about something in direct relation to the familiarity one develops with the subject and the obscure nature of that subject. This also explains why obscure artists and musicians tend to have much more loyal and passionate fans than mainstream artists and musicians.
While the above formula may work at describing how much people care about various people, animals, things and subjects, it surely does not tell what people should care about.