What is Preferensism?

Preferensism is a name I coined to describe a society based on individual preferences. The motto is "from each according to his preferences, to each according to her preferences." It is based on social choice, which amalgamates individual preferences in order to formulate an overall societal preference, and utilitarianism, which is based on the greatest good for the greatest number. Pundits have pooh-poohed utilitarianism because "you cannot maximize two independent variables simultaneously." What do they know? It turns out (as mathematicians like to say) that you can maximize the greatest good of a society according to some reasonable measure. Then assuming there are many ways to do this, which would be the case in any imaginable society, you could choose that way which minimizes inequality. Hence you would have the greatest good, and given that accomplishment, the greatest good for the greatest number.

Given a set of individual preferences by each individual in society, the algorithm or method for combining them would try to get each individual an individualized solution as close to his first preference as possible consistent with maximizing satisfaction, utility or happiness (whatever you want to call it) over the whole society. Imagine a society based on the Gross Domestic Happiness (GDH) as opposed to the Gross Domestic Product (GDP).

Preferensism would generalize the concept of voting and merge the political and economic systems. Just as proportional representation is a generalized voting scheme which guarantees that each minority is represented in parliament according to its percentage in the general population, preferensism would obtain economic and political solutions tailored to each individual insofar as is possible. For instance, a set of preferences might entail 100 work -consumption scenarios. First preference might be to work 10 hours a week and receive the equivalent of $10,000.00 worth of purchasing power. Last preference might be to work 60 hours a week and receive the equivalent of $500.00 worth of purchasing power. When amalgamated over the whole society the individual might end up with a solution in which he worked 30 hours a week in return for $2000.00 worth of purchasing power.

Each set of preferences would be treated democratically. That is to say that each would have one unit of voting power. But instead of voting for a solution that applied to the whole society such as the election of a President, the individual would be voting for a solution that applied only to himself insofar as economic issues were concerned. Politically, which is to say for issues for which an individualized solution is not possible, Preferensism would obtain a solution as close to an individualized solution as possible. That is the same solution might apply to all the individuals in a certain geographical area, for example, while a different solution would apply to individuals living ion a different geographical area. Also the society would approach a direct as opposed to a representational democracy since citizens would be "voting" on goods and services consumed collectively as well as those consumed individually. Finally, if only one outcome is feasible and there are only two possible alternatives, voting would resemble the typical election we're all familiar with such as the election of a President.

Therefore, as the "voting" became more political and less economic, outcomes or solutions would apply to larger and larger minorities, and, as the voting became more economic and less political, it would apply to smaller and smaller minorities and ultimately to the individual when it concerned his or her work-consumption situation.

For more information on Preferensism, see my website: Social Choice and Beyond.

Comments

[Ben]

I don't think you take the 'two maximands' critique very seriously. The idea of lexically ordering utility and then equality goes back at least to Sidgwick, if not Mill or Bentham, but it doesn't answer the problem. A more equal distribution doesn't really mean it's for the 'greatest number', and the real problem is whether we can have a lot of moderately happy people or a few really happy people.

It's all the more problematic if population size isn't fixed. Then you basically have a choice between maximising average or total utility. If you go for the latter, you face Parfit's 'repugnant conclusion' (vastly many people leading lives barely worth living) - but if you maximise the average, then you might find the ideal global population to be about 1 billion leading very rich, happy lives.

[John Lawrence]

Ben, when you maximize utility (either total or average), you are not finding the solution which represents "the geatest good for the greatest number." Instead, you're finding the solution which represents the greatest good period. However, assuming there are many ways to do this (which I think would be the case in any real world situation), one can then, subject to the first criterion, find that solution out of the numerous resultant ties which minimizes the average inequality or which maximizes the minimum utility or you could use some other criterion.

I like your idea of finding the solution that maximizes average rather than total utility. I don't see what this has to do with the world's population though.

If you were faced with the result that the vast number of people were living lives barely worth living, then it would be appropriate to use a criterion such as maximizing the minimum utility rather than maximizing average or total utility.

[Ben]

Maximising the total is maximising the good period; but maximising the average isn't - that's why they can differ in implications.

Consider a choice between the three-person world (9,9,9) and the four person world (10,10,10,2). The former has more total good, but a lower average (8 rather than 9)

In this case, it might seem everyone is better off (or, at least, no worse off - I put aside non-identity problems)

But if we go with the total approach then what about a forty-person world where everyone has 1 unit? Or an eighty-person world where they all have 0.5?

If you give maximising total utility 'lexical priority' then you can reach such unpalatable conclusions, even before applying 'greatest happiness of the greatest number' (and note, inequality isn't the problem)

[John Lawrence]

Ben, I would assume a stable population (let's say of size N) although, of course, there are births and deaths on an instantaneous basis. Maximizing utility or total good over the whole population would be the same as maximizing average utility if average utility is calculated as total utility divided by N. There are some other ways to calculate average utility though such as the median utility which would be that utility such that half the population had a greater utility and half had a lower utility. Maximizing median utility would have different implications from maximizing total utility.

I don't see how your example applies if we assume the population is a constant.

[Ben]

Of course, if you assume constant population, maximising the mean and total will converge - but the fact two criteria sometimes recommend the same course of action doesn't mean they're the same criterion, as shown by the fact they diverge when you allow population size to vary.

Of course, future generations and variable populations are significant problems for any moral/political theory, not just utilitarianism.

[Lori]

[[[[Pundits have pooh-poohed utilitarianism because "you cannot maximize two independent variables simultaneously."]]]]

True, but you can negotiate tradeoffs, some of which may be more optimal than others. Even if (unlike me) you believe the Invisible Hand accomplishes this automagically, there must be some satisfaction in gaining an understanding of how, let alone a self-aware sense of one's own participation in the process. Think of it as equilibrium with a manual transmission.

[[[[Imagine a society based on the Gross Domestic Happiness (GDH) as opposed to the Gross Domestic Product (GDP).]]]]

That country is called Bhutan.

I think your notion called 'preferensism' has much in common with my notion called 'pubwan,' especially the 'maxhi schema' for the 'mapping' of personal preferences. I wish to collaborate. I am especially interested in learning about your algorithm for combining preferences. You might start by examining my sample normset and creating your own for comparison. Please feel free, of course, to edit the wiki.

Current list of pubwan topics:

http://scratchpad.wikia.com/index.php?title=Category:Pubwan

[John Lawrence]

Is there a concise write-up of Pubwan? The Wiki site is too scattered for me. How about contacting me by email?

[Lori]

'Pubwan' is a neologism I coined to advance the notion that open source should apply to data as well as programs, especially for advanced uses such as data mining.

'Maxhi schema' ( http://scratchpad.wikia.com/wiki/Maxhi_schema ) is a system I came up with for describing preferences in personal consumption, which appears also to be an area of great interest to you. In this schema, 'maxhi' denotes maximize (some quantitative variable) with high priority. Other designations are maxlo, minhi and minlo. These tags apply to product specifications, for example fuel economy of a car, fat grams in a foodstuff, total harmonic distortion in a stereo; you name it.

A list of specifications with maxhi schema indicators can apply to a product category or to a market basket (or bundle) of products.

I am aware of the disorganized nature of the wiki I created for pubwan. A wiki is not meant to be a solo effort. You need not take that as a recruiting pitch, but 'pubwan', like 'preferensism', cannot go very far as a solo effort. I do not know whether you are working with others on preferensism. If so, I would like to join that group.

My email address is n8chz at yahoo dot ca. I am an infrequent internet user ( see tribes.tribe.net/netizen ) so be patient with correspondence, but preferensism is an idea I would like to help with.