Parfit: Equality or Priority

The author aims to discuss Nagle's paper 'Equality' as a "kind of egalitarian reasoning."

Stephen Kershnar June 21, 2020 at 1:21 PM

RATIONING IF EQUALITY IS UNIMPORTANT

I find the following argument convincing.

(1) If equality is intrinsically good, then, other things being

equal, leveling down is good.

(2) It is false that other things being equal, leveling down is

good.

(3) Hence, equality is not intrinsically good. [(P1), (P2)]

Even if the argument were not sound, the intuitive distaste we have for leveling down in every case I can think of tells us that equality, if it is valuable at all, is almost always outweighed by the value of well-being. That is, equality is (intrinsically) unimportant.

The previous papers we read about rationing principles and fault-based penalties all assumed that equality is a good thing. If equality is not intrinsically good or is at most a very minor good, then those articles are implausible.

Similarly, the leveling-down objection shows luck egalitarianism to be false. If this is correct, why is this position still so widely held?

Best

Replies 

Stephen Kershnar June 21, 2020 at 1:29 PM

PRIORITARIANISM ABOUT THE RIGHT IS WORSE THAN PRIORITARIANISM ABOUT THE GOOD

Let us assume that prioritarianism is a theory of the good is false. If this is true, then there is no reason to adopt prioritarianism as a theory of the right.

Consider this version of it: The intrinsic duty to make someone’s life go better is more stringent for an individual whose life goes worse than one whose life goes better.

Unless a gain in intrinsic value justifies a sliding-scale duty, nothing makes it true. I can’t see any connection between the usual suspect of duty-justifiers (autonomy, dignity, equality, respect, and utility) and the sliding-scale duty.

Neil Feit June 22, 2020 at 10:14 PM

I find the argument at the top convincing as well. Two quick follow up comments here. (1) The argument would show that what Parfit calls "Telic Egalitarianism" is false. On the rationing issue, do those principles really need this view, or would Deontic Egalitarianism suffice? The argument at the top is not directed at this view. (2) See my comment below if interested in a sketchy defense of prioritarianism as a theory of the good. I need to think more about the other point, that IF it's not a theory of the good then nothing grounds a more stringent duty to benefit those who are worse off.

Jim Delaney June 23, 2020 at 8:50 AM

The leveling down objection seems to me also to be a serious problem for telic egalitarianism. But Steve, I wonder what you think about Parfit's distinction between strong and moderate versions of the view. If I understand it correctly, moderate egalitarians just hold that increases in utility that decrease equality are only in some way worse, but they are not all things considered worse. Conversely, in the leveling down case, the change is only in one way better, but is not all things considered better. I am not convinced that equality is an intrinsic good, but the moderate version (again, assuming I understand Parfit correctly), seems a much more plausible position. In these cases, they would just say that while the change is all things considered better, there is still something regrettable about the change. Does this still seem "incredible" to you?

Stephen Kershnar June 23, 2020 at 10:29 AM

Neil:

Great point. I agree that deontic egalitarianism would suffice, although as you point out than the deontic duty seems to lack a ground.

Your non-comparative account of the axiology of prioritarianism might be the way to go. Perhaps this could be a basis of intrinsic value, one that is summed up with desert-based value. This would allow for some priority for the worst off, but require this compete against desert-based value.

A concern is that with a non-comparative function for prioritarianism, an inefficient transfer of well-being units from someone who is well off to someone who is badly off might be all things considered intrinsically good even though it decreases both the total and average well-being. This is odd.

Example

A is at a +40 well being level and we take 10 well-being units from him. This results in +8 well-being units being given to B who is at a -20 well-being before the transfer. On prioritarian axiology, the transfer makes the world better even though the total and average are lower (total: 20 to 18 and average: 10 to 9).

Best.

Stephen Kershnar June 23, 2020 at 10:32 AM

Jim:

Excellent point.

My concern is whether the equality-related value is so weak as to merit no attention. It is hard to imagine any case in which a Pareto inferior move (a change in which at least one person doing worse and no one doing better) is all-things considered good. If you share this intuition, and perhaps you do not, then equality is at best an incredibly weak value. If so, then we can safely ignore it.

I wonder if you can think of a case in which a Pareto inferior move is all-things-considered good. I can't.

Best.

Jim Delaney June 23, 2020 at 1:43 PM

Steve,

The Pareto inferior move is definitely a tough case. I guess if there were any such case that I could see being all-considered good, it would be one in which both people were very well off (with one doing very slightly better than the other), and the better off one was reduced to the same level as the other. So something like the following:

(1) A=1000 B=1001

(2) A=1000 B=1000

I confess that I myself don't have the intuition that the move from 1 to 2 is all things considered better. But it's not a super strong intuition, and if someone wanted to say that the small utility loss was outweighed by the gain in equality, I could see where they were coming from.

But suppose that in these cases we agree that the move is never going to be all things considered better. Would that show that equality is so weak that it merits no attention? Maybe not. Even if it were always outweighed by a utility gain, it might nevertheless be a morally important feature of the world. Inequality would be be something regrettable. And it's something that cannot be captured just by the amount of utility in the world. I guess it comes down to whether you think a value is worthy of our attention even if it never "tips the scales." Once again, I'm not sure I think it is, but perhaps this is a line the moderate egalitarian could take.

Lastly, isn't it the case that we are not limited to Pareto inferior move cases? What about moves in cases in which some people are doing better and other people are doing worse? Recall Parfit's other example:

(1) All at 100

(4) Half at 100 Half at 200

(5) Half at 70 Half at 200

In terms of utility, the move from 1 to either 4 or 5 would be all things considered better. For the moderate egalitarian the move to 4 would be better, but the move to 5 could be considered worse because of the resulting inequality. I imagine you don't think it reasonable to say the move to 5 is worse? Or if not, do you think something besides equality is doing the work?

Neil Feit June 23, 2020 at 4:09 PM

Hi Steve,

You say this above:

"A concern is that with a non-comparative function for prioritarianism, an inefficient transfer of well-being units from someone who is well off to someone who is badly off might be all things considered intrinsically good even though it decreases both the total and average well-being. This is odd."

I don't think it's odd, if desert is what underlies this kind of prioritarianism. For example, if a sinner is flourishing then we might think it's intrinsically better if he has less well-being. So, it's better with lower total and lower average well-being. I think the same could go for the example you gave above -- it could be a change for the better to take 10 away from someone with a high well-being level and give 8 to someone much less well-off (the "steepness of the desert curve" here does it).

Stephen Kershnar June 25, 2020 at 8:50 AM

Jim:

You make an excellent point in that the issue is whether a moral factor that never tips the scales is unworthy of our attention. I think this is true. If something is so unimportant that even in extreme cases designed to showcase it, it has no effect [for example, your original (1) & (2)], we should ignore it. It's less important than a rounding error.

On your second example, the intuition that (1) is better than (5), I think prioritarianism as well as egalitarianism would justify the ranking, if either one is justified. Given that egalitarianism is not justified (assuming it is at most a very weak value), this leaves prioritarianism.

I have two explanations of the intuition in your second case.

First, consider desert. I think we assume that the people going from 100 to 70 are getting less than what they deserve whereas the people going from 100 to 200 are getting more than they deserve. If this is correct, then it is desert that justifies why we should prefer (1) over (5). If desert and prioritarianism are independent, and I think they are, this suggests the background assumptions are doing the work here.

Second, I think we are hard wired to prefer fair or equal distributions. Capuchin monkeys prefer equality and fairness.

https://news.berkeley.edu/2015/03/11/frans-de-waal-greater-good/

I suspect this has an evolutionary advantage even if it does not track moral reality. I suspect, then, that the preference for (1) over (5) is like other evolutionary preferences that are deeply felt but not part of moral reality (for example, heterosexual and in-group preferences).

Best.

Stephen Kershnar June 25, 2020 at 9:00 AM

Neil:

Great point. I wonder, though, whether desert can underlie prioritarianism. The two seem independent of one another because their are cases when desert calls for transferring well-being from a person whose life is going poorly to a person whose life is going well. It also justifies the goodness of transfer by desert rather than well-being level.

If prioritarianism and desert are independent, I wonder if we should add the prioritarian- and desert-adjusted value to get an overall value or do some other function.

I also wonder whether our intuitions are dependable in these sort of population ethics case. Looking at utils/life is not a good measure of how well someone's life goes if intrapersonal repugnant conclusions are a problem (e.g., century of ecstasy vs. drab eternity). But if we switch to utils/life/time, then intuitions about egalitarianism and prioritarianism become noticeably less strong, at least for me.

Best.

Stephen Kershnar June 21, 2020 at 1:23 PM

PRIORITARIANISM IS FALSE BECAUSE IT MAKES THE GOOD DEPEND ON IRRELEVANCIES

Derek Parfit asserts that benefitting people matters more the worse off these people are. Those who are worse off in their lives as a whole.

A theory of the good is comparative when it says that what makes the world intrinsically better (or worse) involves a comparison of two or more things. In this context, it involves a comparison of two or more individuals’ well-being.

Consider the following argument.

(1) Prioritarianism is comparative or non-comparative.

(2) If prioritarianism is comparative, then it is false.

(3) If prioritarianism is non-comparative, then it is false.

Proposition (1) is trivially true.

Proposition (2) rests on Parfit’s Egyptology and Reverse Egyptology problems. Egyptology says that how valuable a person’s life is depends on how well Egyptians’ lives went (because this will affect the value or priority of someone’s life). Reverse Egyptology says that how valuable a person’s life is depends on how well people’s lives go thousands of years in the future. Facts about Egyptian king Djoser’s life (2650 BC) do not make someone’s life (for example, Neil Feit’s life) more or less valuable.

Proposition (3) rests on the notion that how valuable a person’s life is should not depend on how well merely possible people’s live go. Possible but non-existent people’s life (for example, King Arthur’s life) do not make someone’s life go more or less valuable than does Djoser’s life. If we are not comparing someone’s life to possible people, then it is hard to see what motivates prioritarianism.

Replies

Neil Feit June 22, 2020 at 10:24 PM

Steve, I'm not sure that I get your reasoning for premise (3) above. Doesn't Parfit stipulate that on prioritarianism, only absoute (non-comparative) levels matter -- in other words, it's irrelevant to the fact that benefiting the worse off matters more that they are, in fact, worse off than others. I guess I don't see how merely possible people figure into this. I'd suggest that one way that prioritarianism could be a non-comparative theory of the good has to do with desert. If those who are worse off gain well-being, this will make more value for the world than if those who are better off gain well-being. (We can even assume with Parfit that all relevant individuals do not deserve more than they in fact (currently) get.) On this view, how valuable a person's life is doesn't seem to depend on how well merely possible people's lives go; it depends only on what they get and what they deserve.

Stephen Kershnar June 23, 2020 at 10:08 AM

Neil:

I think this is an excellent point. I think you're right that Parfit interprets prioritarianism in a non-comparative sense in way analogous to non-comparative desert. If so, then possible people are irrelevant because this would make it comparative.

I wonder though whether the non-comparative approach fits as well with prioritarianism as it does with desert. Consider people who have lifetime well-being levels: -20, -10, 0, 10,20. Consider a graph with well-being on the horizontal axis and intrinsic value on the vertical axis. What would the chart look like with non-comparative prioritarianism but without desert? Perhaps it would be convex or asymptotic.

My concern is that the graph either takes desert into account or it does not.

(1) If it takes desert into account, I'm not sure what room is left for prioritarianism. Perhaps it is a different ground of intrinsic value that is added to the desert-based value.

(2) If it does not take desert into account, I suspect it gets our intuitions wrong. On a sufficiently concave function, small increases in well-being to a person at -1,000 will outweigh massive increases to a person at +10.

Best.

Neil Feit June 23, 2020 at 4:21 PM

Steve,

I'd go with your horn (1) and offer this response. Desert has to do with the good. And Parfit mentions that the Priority View can take telic or deontic forms. So, if we take the deontic form, then priotiarianism has to do with the right, and not the good. The idea is that, other things being equal, the more worse off one is, the stronger the moral reason to benefit him. A theory of the good that makes use of desert-adjusted well-being doesn't say anything about what we ought to do; prioritarianism does.

Stephen Kershnar June 25, 2020 at 9:10 AM

Neil:

Interesting point.

Note, though, that if the good depends on desert and the right depends on prioritarianism, and desert and prioritarianism are independent, then the good and the right start to diverge in a way that intuitively seems to be unsatisfying. Here is a principle I have in mind.

(X) The right intrinsically depends on, and only on, some or all of these factors: well-being, the good, and dignity/respect (e.g., rights, Kantian duties, or fairness).

Unless prioritarianism is filled out in terms of dignity/respect, it is not part of the intrinsic right.

Perhaps it is part of them, maybe fairness. Still, this is putting a lot of weight on an incredibly vague reason and one that has to do a lot of different types of work (e.g., prohibiting exploitation, torture, and degrading fantasies). I am doubtful it is coherent enough to do this sort of work. Perhaps I'm mistaken.

Best.

Stephen Kershnar June 21, 2020 at 1:26 PM

THE INTRAPERSONAL ARGUMENT AGAINST EGALITARIANISM AND PRIORITARIANISM

No one thinks that egalitarianism or prioritarianism is true for intrapersonal value. Consider, for example, how much periods of someone’s life adds to (i) how well his life goes or to (ii) how much value his life adds to the world.

Here then is my argument.

(1) If egalitarianism (or prioritarianism) counts for interpersonal value, then it counts for intrapersonal value.

(2) Egalitarianism does not count for intrapersonal value.

(3) Hence, egalitarianism does not count for interpersonal value. [(1), (2)]

(4) Hence, egalitarianism or prioritarianism are false. [(3)]

Phil Reed June 26, 2020 at 4:38 PM

Thanks for letting me eavesdrop on the interesting conversation above.

I find these matters really difficult and I have not thought a lot about them. I think I learned from this paper that telic egalitarianism is dumb. And because, unlike Parfit, I don't have the intuition that equality is better than inequality in the Divided World, I'm inclined to accept deontic egalitarianism. But I am not sure that deontic egalitarianism needs to be distinguished as sharply from prioritarianism in the way Parfit wants to. This is because, if I understand the view correctly, prioritarianism need not accept that equality is better than inequality in the divided world: perhaps 100 is an acceptable absolute level of well-being.

I think that talking about inequality so abstractly can be problematic. For example, Nagel seems to think that we can have a view about appropriate levels of inequality among our children that translates to the level of burdens and benefits in society. Parfit seems to go along with this. But parents have special obligations to their children that could and probably should make a difference to how parents think about distributing benefits and burdens among their children that do not apply when we think about these distributions at the level of society.

Replies

Stephen Kershnar June 27, 2020 at 4:17 PM

AGAINST DEONTIC EGALITARIANISM

Phil:

Always great to hear from you.

You say the following, “I'm inclined to accept deontic egalitarianism. But I am not sure that deontic egalitarianism needs to be distinguished as sharply from prioritarianism in the way Parfit wants to. This is because, if I understand the view correctly, prioritarianism need not accept that equality is better than inequality in the divided world: perhaps 100 is an acceptable absolute level of well-being.”

I have a couple of questions about deontic egalitarianism.

(1) Duty to Level Down. Do you think a deontic egalitarianism entails that we have a duty – all things considered or other things being equal - to level-down well-being, at least within a single society?

If so, isn’t this implausible?

(2) Duty-Maker. What would explain or justify the duty in deontic egalitarianism? Telic egalitarianism – whether strong or moderate - is a theory of the good that can connect cleanly with consequentialism.

But what would explain a roving duty to go out a promote or protect equality even when it

(a) does not maximize the good and

(b) extends beyond those we owe duties in virtue of contract (actual or hypothetical), prohibitions on harm, rights, and special relations?

Perhaps a better approach is to reject egalitarianism and prioritarianism – whether as theory of the good or the right – and reject any rationing principle that includes them as a basic feature of the good or the right.

Best.

Phil Reed June 29, 2020 at 2:21 PM

Hi Steve,

Thanks for your comment.

(1) No, I don't think deontic egalitarianism commits one to a duty to level-down. I believe Parfit was saying that only telic egalitariansim does. But if it does, then I would abandon deontic egalitarianism.

(2) What could explain a duty to promote equality is some kind of theory of justice.