CHAPTER 3: Social Decision Functions





We now turn to a discussion of freedom and equality within a societal context. We consider how decisions are made in a society which take into account individual tastes and preferences but in which the decision does not necessarily result in every individual having the outcome he most prefers. In an individual's private life we assume he is free to pursue whatever course he most prefers. In social or public life compromise may be necessary, and how these compromises are made has a lot to do with how the values of freedom and equality are embedded in the social decision making mechanisms. Also how much individual tastes and preferences, individual values, if you will, are taken into account has a lot to do with whether the society is repressive or expressive and to what extent justice prevails.

The two arenas of social decision making with which we are most concerned are the political and economic arenas. In "Social Choice and Individual Values," Kenneth Arrow states: "In a capitalist democracy there are essentially two methods by which social choices can be made: voting, typically used to make 'political' decisions, and the market mechanism, typically used to make economic decisions. In the emerging democracies with mixed economic systems, Great Britain, France and Scandinavia, the same two modes of making social choices prevail, though more scope is given to the method of voting and decisions based directly or indirectly on it and less to the rule of the price mechanism. Elsewhere in the world, and even in smaller units within the democracies, social decisions are sometimes made by single individuals or small groups and sometimes (more and more rarely in this modern world) by a widely encompassing set of traditional rules for making the social choice in any given situation, e.g., a religious code."1 Other areas of social decision making are the family and groups and organizations of all kinds.

In social decision making processes, the problem is the selection of one or more alternatives by the society taking into account the preferences of the individual members of society. The individual's freedom is expressed in his right to express his preferences, to make choices about what decisions he, himself, prefers and to have a say in the process of the actual selection. The agreement between the individual and society is that he will abide by the actual alternative or alternatives selected by society even though these are ones he may or may not prefer.

In our discussion we will attempt to formalize our thought processes to a considerable degree compared to the foregoing discussion and build on an area of social thinking in which considerable formal analysis has already taken place. The analysis will be mathematical in some places although at an elementary level. We are thus trying to bridge the gap between philosophical treatises which are vague in the precision of their analyses and mathematical analyses which are devoid of motivation and social impetus.

Let us say we have a society composed of N individuals. We will refer to a specific individual as the jth individual. We also consider a set of options, O, containing a number of options, M. Let the particular options be denoted by the set, O = (m1,m2,...mM), and let mi be a variable denoting an option. We assume that each individual has a corresponding set of options which, in general, may not be the same in quantity or quality for each individual. In any particular instance an individual may choose one or more options. An individual's freedom is a function of the quantity and quality of options available to him. We define an individual's utility as a measure of the satisfaction he derives from having chosen a particular option. There will be a certain utility associated with each option in the option set for each individual. For the purposes of the present discussion we will use the terms "utility" and "satisfaction" interchangably.

An individual's criteria for making a choice or choices of the options available to him is based upon his preferences considered over the available options. We define a preference rating function, Pj(m), which represents the jth individual's preferences over the set, O = (m1,m2,...mM). In general, Pj(mx) > Pj(my) means that the jth individual prefers the option, mx, to the option, my.

We can represent an individual's preference rating, in the most general terms, geometrically as in Figure 1. We assume here that the option set is the same for each individual.


Figure 1

Preference Ratings of 3 Individuals



We note from Figure 1 the following:

P1(m4) > P1(m3) > P1(m1) > P1(m6) > P1(m2) > P1(m5)

P2(m4) > P2(m1) > P2(m5) > P2(m3) > P2(m6) > P2(m2)


P3(m2) > P3(m3) > P3(m6) > P3(m5) > P3(m1) > P3(m4)


This represents a complete ordering of the preferences of these individuals. If N=3, so that the society in the example is composed of three individuals, this would represent a complete ordering of preferences over the whole society. We note that the ordering of preferences does not contain as much information as the geometrical representation which contains not only ordering information but also information as to the relative intensity of preferences. For instance, again referring to Figure 1, we see that for the second individual, m5 is greatly preferred to m3i.e.there is a large gap between the two options. Also m5 is greatly preferred to m6. However, he is relatively indifferent between m3 and m6 (small gap between them). So the geometrical representation gives information with regard not only to preference ordering but also with respect to preference intensity. It is this complete specification that we refer to as a preference rating. We also note in passing that it is possible for two options to have exactly the same preference rating. In this case, we say that an individual is indifferent with respect to the two options.

Because of the way that the preference ratings can, without loss of generality, be represented geometrically on a straight line, it can be shown that there is a unique mapping from the set of rational numbers, x, (0≤x≤1), to the preference rating. Therefore, Pj(mi) = ax + b, where a and b are integers and 0≤x≤1 for all i and j. Each option can be assigned a rational number which represents the individual's preference rating of that option and the set of numbers corresponding to the rating of each option becomes the individual's preference rating of the option set. The preference intensity between the two options, mx and my, can be defined as follows:

Preference intensity = Pj(mx) - Pj(my)

We assume in this analysis that an individual's utility or satisfaction with an option is a linear function of his preference rating. In reality this is not necessarily the case since he might prefer an option that he has incomplete knowledge of and find out later that it contained much less or greater utility than he had initially imagined. To simplify our analysis we assume that complete knowledge of the consequences of choosing a particular alternative is available to the individual so that his utility has a direct correlation with his preference rating. We speak of a utility rating as a complete specification of utilities over all options. We may also speak of the utility of a particular option. We have


Uj(mi) = cPj(mi) + d for all i and j and c,d integers


Therefore, there is a unique one-to-one mapping between the rational integers, x, (0≤x≤1), and the utility function so that


Uj(mi) = ex + f for 0≤x≤1 and e,f integers


It can be seen that the most complete specification of both the preference ordering and the preference intensity of two options can be made from the comparison of the two rational numbers assigned to each option. If mx is preferred to my, then Pj(mx) > Pj(my) and the intensity is Pj(mx) - Pj(my).


Let us assume that we have two options which are very close to being equally preferred, m1 and m2. Then as long as Pj(m1) > Pj(m2), m1 is preferred to m2, and the slightness of preference can be expressed to any level of perception by increasing the number of places to the right of the

decimal point. For example, let


Pj(m1) = 100.49742

Pj(m2) = 100.49741


If a person is capable of a higher level of discrimination, this might also be expressed, for example, as


Pj(m1) = 100.497419254

Pj(m2) = 100.497414859


In the extreme case, Pj(m1) - Pj(m2) approaches zero and we say that the jth individual is indifferent between m1 and m2. If m1 is greatly preferred to m2, then Pj(m1) - Pj(m2) is very large and can be expressed by a very large rational number. In the extreme case, Pj(m1) - Pj(m2) approaches infinity. In actual practise, the real axis could be divided into equal segments each of which represents the minimum level of discrimination which a person is capable of expressing.

Therefore, we conclude that an individual can completely (and without loss of generality) express both the amount of preference and the intensity of preference relative to any other option over a set of options by assigning a rational number to each option. This represents a complete and unique specification of preferences. If an individual's preference ratings are to be used as inputs in a democratic voting system, then all preference ratings must be given equal weight and power. This constraint is sufficient to allow us to normalize the preference rating function by choosing a = 1, b = 0, so that Pj(mi) = x for all i and j and x is a rational integer, 0≤x≤1.

Let us address the issue of interpersonal comparison of utility or satisfaction about which there has been much controversy in the literature. Some individuals may derive more satisfaction out of a certain result than others even though their preference rating for the options are exactly the same. For example,


Pi(mx) = Pj(mx)

but Ui(mx) ≠ Uj(mx)


This seems due to the fact that some people are capable of taking greater satisfaction in the same set of circumstances than others. It does not seem that this should be taken into account when prescribing a rule for the combination of individual preferences in order to arrive at a democratic social result. For instance, in the selection of a candidate, the fact that individual i will be happier if candidate A wins than individual j will be if candidate B wins although


Pi(A) = Pj(B),


should not influence the results. To say otherwise would be

tantamount to giving i a more powerful vote or more consideration in

arriving at a result than j, which would be a transgression of the democratic ideal of equality in that each individual should have the same voting power and the same consideration regardless of whether he will enjoy the results more or less than another individual.

Therefore, it seems only fair to treat not only everyone's preferences the same in arriving at a social decision but also to treat everyone's utility the same in assessing the effects of the social decision. We do this by normalizing everyone's preference and utility ratings to the same standard. This is a generalization of "one man-one vote." Therefore, we assume the following:


e = f = 0 and

0≤Uj(mi)≤1 for all i and j


Furthermore, without loss of generality, we can choose c = d = 1 so that preference ratings become identical with utility ratings for any particular individual. Therefore,


0 ≤ Pj(mi) = Uj(mi) ≤ 1 for all i and j.




A social decision making function is a rule that transforms the data represented by individual preferences in such a way as to choose an option or options that represent the decision of society as a whole. For example, in a democratic election governed by majority rule and assuming two candidates, the input data are the individual votes, the social decision function is the rule for determining the outcome by choosing the candidate with the most votes and the social choice is the candidate with the most votes.

A general model for a social decision making process includes the preference ratings of each individual as the input data, the social decision making function and the utilities of the outcome for each individual as the output data resulting from the social choice that was made. In terms of systems engineering, the social decision making function transforms the input into the output by narrowing down the input option set, O, into an output option set, O'. The output option set may include one option from the input set or a combination of options from the input set.

Let us denote the social decision making function by D. Then


D{P(O)} = U(O')


In this model there are many inputs and many outputs as shown in Figure 2.



Figure 2

Social Decision Making Process



The social utility is defined as the sum of the individual utilities:


U(O') = Uj(O')



We can assess both the input and the output of the social decision making function with respect to the concepts of freedom and equality. At the input freedom is embodied in the quantity and quality of options in the option set, O. Equality is embodied in the principle of giving equal power and weight to each individual's preference rating which is done by normalizing each rating. At the output individual utility is a function of the outcomes which apply to that particular individual. The social decision making function assigns a certain outcome to each individual. How well that outcome "fits" the individual (in accordance with his preference rating) determines the individual's utility. The social utility is the sum of the individual utilities. The greater this is, the greater, it can be assumed, is the satisfaction inherent in the society. Equality at the output must be assessed in terms of the differences in individual utilities. The individual inequality can be expressed as the difference between an individual's utility and the average utility where

Obviously, the most desirable result obtains when each individual's ideal is represented in the option set and the decision is such that each individual receives his top choice. In this situation, the outcome consists of many options rather than one. Then


Ui = 1, 1<i<N

Total Utility, TU = N

Average Utility = AU = 1

Individual inequality = II = 0

Average inequality = AI = 0


This represents the ideal situation which in some instances may be approached in practice. However, in many practical situations the output option set, O', is constrained so that not everyone may receive his first preference. For example, in a political election for President, there is only one outcome which applies to everyone.





The history of social decision functions starts with the theory of elections studied by Jean-Charles Borda (1733-1799) who wrote a paper in 1781 in which he recognized what has been called the voting paradox. For three candidates A, B, and C and 21 voters, Borda showed, for example, that, if 1 voter preferred A to B and B to C, 7 voters preferred A to C and C to B, 7 preferred B to C and C to A, and 6 preferred C to B and B to A, then by the method of casting 1 vote for one's favorite candidate, A would win with 8 votes although the voters would prefer B or C to A by a majority of 13 to 8i.e.if just A and B or A and C were put to a vote, A would lose. Borda proposed to remedy this defect by a voting system or social decision function which assigned points to each candidate dependent on their place in the ranking. Thus, we could assign a value of 0 to last place, 1 to second place and 2 to first place. The winner of the election then would be the one with the most points. In the above example, A would have 16 points, B would have 21 and C would have 26.

Alternatively, the points can be determined by having a series of elections taking two candidates at a time. For instance, if the voters vote on A and B, the result would be 8 votes for A and 13 for B. Voting on A and C would produce 8 more votes for A and 13 for C. Voting between B and C would result in 8 votes for B and 13 votes for C. The final tally then is 16 for A, 21 for B and 26 for C which is exactly the same result as in the first method. Duncan Black, whose history of the theory of elections we are synopsizing, had this to say about the value of Borda's work: "Of this paper as a whole we can say that it is short, clearly expressed and suggests the possible lines of development that a theory of elections might take. A specific defect is that it fails to distinguish between two separate lines of thought-the criterion which it advocates explicitly and another criterion (that of Condorcet) of which, earlier, it had made implicit use. A more important deficiency is that, to be satisfactory, a theory of elections must give a real insight into the nature of group decisions; and Borda's paper cannot be said to do this. It is a good first step towards a more significant type of theory, but no more than that."2 It should be noted that Borda's method gives the same results either by pair-wise comparison (the candidates are compared two at a time) or by overall comparison (all the candidates are evaluated at one time).

The next important historical figure is the Marquis de Condorcet, (1743-1794), a French mathematician, philosopher, economicst and social scientist who played a part in the French Revolution and ended up dying in jail. The Condorcet criterion for determining an election was for voters to vote on each pair of candidates. If one candidate had a majority in each case, then he is declared the winner. The problem is that this is not always the case, and this gives rise again to the voting paradox. The voting paradox comes into play when there are more than two voters and more than three alternatives. For example, let us assume there are three voters and three alternatives. Voter 1 prefers A to B and B to C. Voter 2 prefers B to C and C to A. Voter 3 prefers C to A and A to B. Now if we put A and B to the vote, A will get 2 votes and B, 1, with the result that A is the winner. If B and C are put to the vote, B will get 2 votes and C, 1, and B will be the winner. At this point, we can say that society prefers A to B and B to C. Hence it is logical that society should prefer A to C and A should be the overall winner. However, when A and C are put to the vote, C gets 2 votes and A, 1, with the result that C is the winner. So society prefers A to B, B to C and C to A. Therefore, we conclude that there is no rational solution to the problem of picking a winner using this method. There are ad hoc methods for resolving this difficulty. For instance, if there are N voters and 3 alternatives and each wins once in the three pair-wise comparisons, we could pick as the winner the one with the largest pair-wise majority.

The next contributor to the field was also a prominent Frenchman, the Marquis de Laplace (1749-1827) who by a very elegant argument ended up with the same method of marks as Borda. "Starting from slightly different premises he arrives at the same set of marks as Borda."3 Laplace did foresee a difficulty with his system. "The only obstacle which Laplace foresees to the use of the scheme is that its working might be frustrated by electors placing the strongest opponents to their favourite candidates at the foot of their list. This would give a great advantage to candidates of mediocre merit, for while getting few top places they would also get few lowest places. The same difficulty had been foreseen by Borda, who, when confronted with it, had replied, 'My scheme is only intended for honest men.'"4

In his own work Black seems to have followed a line of development which owes more to Condorcet than to Borda and Laplace and seems to think Condorcet's work has more value. In the present book, our line of development owes more to Borda and Laplace. Black's criticism of the Borda method is based on two main points. "The approach of these two authors, however, seems to require assumptions which are quite untenable. It supposes, for instance, that a voter considers the relative worths of two candidates to stand in some definite ratio to each other, such as 5:4 or 6:1 or some other. But we may feel quite sure that the human mind does not work this way. There are definite ratios of exchange for goods in the market, and a purchaser may value equally two units of one and three units of another; but the same is not true of the elector's valuation of candidates."5 Brown's objection boils down to an argument over the mathematics by which a voter expresses his preference rating. According to Brown it is acceptable to have preferences regarding candidates but not intensities of preference. Black also does not think the voters are capable of expressing a more sophisticated preference rating than simple ranking of candidates. When the expression of preferences is visualized graphically, then it is not difficult to see that preference intensity can be expressed as naturally as preference ranking without undue strain on the voter. The mathematics associated with this graphical representation then can be derived later with or without professional help.

Brown's other objection has to do with interpersonal comparisons. "Next the scheme assumes the merit attributed by one elector to be exactly the same in kind as that attributed by another. The difficulties of making interpersonal comparisons of merit have already been examined by economists under a slightly different guise, in connexions with making interpersonal comparasons of utility. Just as we would be reluctant to adopt an economic theory of this type, so we would be unwilling to adopt a political theory which assumed the merit attributed by voters to be measurable and the same in kind as between one voter and another."6 The point here is that in a democratic system the merit attributed by one elector should be considered the same in kind as that attributed by another since each vote should be given equal weight and consideration. To assert that some voters' votes are worthy of more consideration because of their superior knowledge or intelligence is to subscribe to an aristocratic system in which some votes count for more than others. The same holds true for economics. Why should one person's utility be given more consideration than another's? Why, indeed, unless some people are deemed to be more important than others.

Black goes on to write about the intellectual fervor that existed in the Age of Enlightenment when the techniques of mathematics were enthusiastically brought to bear on political and economic problems. "The second half of the eighteenth century in France was one of the outstanding epochs of scientific thought. Science had felt its strength and its impulse and did not know what barriers it might not cross.

"The hope had sprung up to carry the methods of rigorous and mathematical thought beyond the physical and into the realms of the human sciences. But after a brilliant start, its fate was to be misunderstood and forgotten."7 The interesting thing about these early researchers is that they all had achieved distinction in other fields and were attempting to apply mathematical techniques to societal problems.

Next come a series of English researchers the most important of which is the Rev. C.L. Dodgson who under his pen name, Lewis Carroll, wrote "Alice in Wonderland." Dodgson, it should be pointed out, was a mathematics lecturer at an English college, Christ Church, and so was in the tradition of mathematicians who had done research on the theory of elections or voting systems. Dodgson considered the method of marks of Borda and Laplace to be perfect except for the flaw that it allowed voters to vote insincerelyi.e. to misrepresent their true preferences in order to stand a better chance of getting their favorite elected. However, it should be pointed out, that the successful use of this misrepresentation was predicated on a knowledge of how the other voters would be voting. To quote Dodgson:


"This method would, I think, be absolutely perfect, if only each elector wished to do all in his power to secure the election of that candidate who should be the most generally acceptable, even if that candidate should not be the one of his own choice: in this case he would be careful to make the marks exactly represent his estimate of the relative eligibility of all the candidates, even of those he least desired to see elected; and the desired result would be secured.

But we are not sufficiently unselfish and public-spirited to give any hope of this result being attained. Each elector would feel that it was possible for each other elector to assign the entire number of marks to his favorite candidate, giving to all other candidates zero: and he would conclude that, in order to give his own favorite candidate any chance of success, he must do the same for him."8

Dodgson proposed a method of voting which was a modified version of the method of marks. He then went on to propose a method which was a modified version of the Condorcet method of pairwise comparison. In the event of a cyclical majority this was to be resolved by considering how many vote changes each alternative would require in order to have a majority when paired with every other alternative. The one with the least change required would then be declared the winner.

Basically, the history of voting systems boils down to a competition between the method of Borda and Laplace and the method of Condorcet. The advantages of the Borda method are that it always produces a winner and a winner who is a justified winner. The disadvantages are that in some cases the winner is not the same as would be produced by a simple majority rule even though majority rule itself seems less than justified in certain cases, and the system is unstable in that an insincere representation of one's true preferences can sometimes be advantageous. The advantage of the Condorcet method is that it produces the same winner as majority rule if there is a winner and the system is stablei.e.there is no advantage in voting insincerely. The disadvantage to the Condorcet method is that sometimes the results are intransitivei.e.there are cyclical majorities-A is preferred to B which is preferred to C which is preferred to A etc.

Additionally, it has been thought that the Borda criterion, what Murakami calls a 'finite ranking rule' was arbitrary in that different winners might be selected from the same input data depending on how the system was set numbers were assigned to the various ranks.


"If a social decision-making rule is to be evaluated according to an 'amount of information' it gathers, a complete finite ranking rule is unquestionably the best rule. However, as we have noted before, a finite ranking rule is marred by arbitrariness.

...The point is that, whether we collect information fully or not, we have no absolute principle for utilizing the information concerning individual decisions. For we have no intrinsic principle for expressing numerically the ranks in each individual's preference ranking. Neutrality demands that we treat all alternatives in an equal manner. Symmetry demands that we treat all individuals in an equal manner. However, no principle within the democratic framework seems capable of telling us how to treat the ranks in individual orderings-of telling us what rule is better in this respect..."9

Part of the arbitrariness stems from the fact that the finite ranking rules only take preference ordering and not preference intensity also into account. This is remedied by the geometrical expression of preference ratings presented in this book. A further difficulty is presented by the fact that there aren't two non-arbitrary reference points with respect to which preferences are measured. This situation will be further discussed and remedied a little later.





First let us consider an option set of 2 alternatives, A and B, such as is the case in an American Presidential election, and a voting population of N voters. The social decision function that is traditionally used is majority rule. That is each voter casts one vote for one or the other candidate and the candidate with the most votes, which necessarily constitutes a majority, wins. Let us consider an example which casts doubt on the value of this system. Let us say that we have three voters and the following set of preference ratings:


P1(A) = P2(A) = .6, P3(A) = .1

P1(B) = P2(B) = .5, P3(B) = .9


In the single vote, majority rule system, A would get 2 votes, and B would get one with the result that A would be the winner. When we consider the total utility of each candidate,


U(A) = .6 + .6 + .1 = 1.3

U(B) = .5 + .5 + .9 = 1.9


Therefore, the selection of candidate B would result in a higher level of satisfaction among the population as a whole. In general, whenever the majority candidate is not greatly preferred to the other candidate by the majority and the minority candidate is passionately preferred by the minority, it seems that the public interest is not best served by the utilization of majority rule.

An example given by the Rev. C.L. Dodgson (Lewis Carroll) is the following:


"Let us suppose that there are eleven electors, and four candidates, a,b,c,d; and that each elector has arranged in a column the names of the candidates, in the order of his preference; and that the eleven columns stand thus:



a a a b b b b c c c d

c c c a a a a a a a a

d d d c c c c d d d c

b b b d d d d b b b b


Here a is considered best by three of the electors, and second by all the rest. It seems clear that he ought to be elected; and yet, by the above method, b would be the clear winner-a candidate who is considered worst by seven of the electors!"10


In this example each voter votes for his first choice and the one with the most votes wins. Therefore, a gets 3 votes, b gets 4, c gets 3 and d, 1. In our terminology, the above example can be represented as follows:













When we consider more than two alternatives, we run into the same problem as pointed out by Rev. Dodgson. In the following example the winner has an absolute majority (over half the voters rank him first) although it is doubtful if he should win.


b b b b b b a a a a a

a a a a a a c c c d d

c c c d d d d d d c c

d d d c c c b b b b b


"Here a is considered best by nearly half the electors (one more vote would give him an absolute majority), and never put lower than second by any; while b is put last by five of the electors, and c and d by three each. There seems to be no doubt that a ought to be elected; and yet, by the above method, b would win."11

It can seen from these examples that, when there are more than two alternatives, it is doubtful that the use of majority rule or an extension of majority rule results in justice being served. "Moreover, we may conceive a society where the minority prefers an alternative much more ardently

than the majority prefers the contrary alternative. We may well doubt that the majority principle still makes sense. It is worth while trying to include preference intensity as an admissible element of individual decisions, and also as a factor in social decision-making."12 When there are more than two alternatives, preference intensity is automatically expressed by virtue of preference ordering. However, it is not the most complete specification of preference intensity. If majority rule is suspect as a democratic principle, what then should replace it? It seems obvious that the Borda criterion or some variation of the finite ranking rule or the related concept of maximization of utilities. These criteria, in general, seem to promote the general welfare better and result in more social justice while the Condorcet criterion and the extensions of majority rule result more in the use of power by a decisive group whether a majority or a minority in order to promote that group's interests at the expense of the other segments of the population. "The proposed rule of social decision-making is simply that all alternatives should be ordered according to the magnitudes of 'social utility index' defined as a summation of all individuals' cardinal utility indices."13

Returning to the voting paradox, let us consider a few examples that would produce an intransitive social decision. The first is as follows:


P1(A)=.9 P1(B)=.2 P1(C)=.1

P2(B)=.5 P2(C)=.4 P2(A)=.3

P3(C)=.9 P3(A)=.8 P3(B)=.1


By the extension of majority voting, this social decision is intransitive. But, if we compute the utilities, we have






Therefore, if we pick the alternative with the highest social utility, A would be the winner. We can see that an intransitivity will never occur when we maximize utilities. Not only that but the decision makes sense since the second voter is relatively indifferent among all candidates, the first voter strongly prefers A to both B and C and the third voter is relatively indifferent between A and C but strongly prefers both of them to B. So it makes sense for A to win.

Let us consider another example:


P1(A)=.8 P1(B)=.5 P1(C)=.1

P2(A)=.1 P2(B)=.8 P2(C)=.2

P3(A)=.7 P3(B)=.6 P3(C)=.8


By the principle of maximization of utilities, we have






Again by the Condorcet criterion of pairwise comparison, this example is intransitive. By the maximization of social utility criterion B is the winner. Note that the first place preferences for all three voters have the same ratings. However, the second voter strongly prefers B to either A or C, the first voter does not strongly prefer A to B and the third voter is relatively indifferent among all three candidates. Again the outcome makes sense.

Another intransitive example:


P1(A)=.3 P1(B)=.2 P1(C)=.1

P2(A)=.1 P2(B)=.9 P2(C)=.8

P3(A)=.5 P3(B)=.4 P3(C)=.9






Here, by maximization of social utility, the winner is C since the first voter is relatively indifferent among all three, the second voter is relatively indifferent between B and C but prefers both strongly to A, and the third voter prefers C quite a bit more than either A or B.

We see that, all other things being equal, a strong preference intensity by one voter between two candidates will tend to elevate the candidate more strongly preferred to the winning position. In each case more complete information about the preference ratings has led to a logical solution using the rule of maximization of social utility.

Kenneth J. Arrow14 argues that a social decision function, or to use his terminology a social welfare function, which satisfies five basic and reasonable criteria does not exist. If in fact one does not exist, then there is no reason to pursue the search for finding an optimal social decision function and this whole area of inquiry has become a blind alley. We submit here that Arrow made an error in one of his basic assumptions so that his results concerning the non-existence of a social welfare function are invalid. We will prove that there does exist a social welfare function which does satisfy Arrow's criteria, and that that is the maximization of social utility social decision function. This social decision function also solves the impasse of the voting paradox.

Arrow's Condition 1 that there are three alternatives such that there are no restrictions on their ordering by the voting public is certainly satisfied. In fact there are no restrictions whatsoever on the ordering of any of the alternatives so that a much stronger condition is satisfied. Arrow's Condition 2 is the positive association of social and individual values. This means that, if an individual elevates a certain candidate in his preference rating, all else being equal, then the social decision should reflect this. The social decision will at least not change adversely against the candidate who has been so elevated. It may be indifferent to this change or it may change so as to elevate that candidate in the social decision. The maximization of social utility decision function clearly satisfies this criterion.

For the moment, let us skip Arrow's Condition 3. We will come back to it later.

Arrow's Condition 4 is called the condition of citizen's sovereignty. This means that the social welfare function shall reflect the individual preference ratings and only them and shall not have some built-in bias which favors one alternative over another. This is clearly the case with the maximization of social utility social decision function.

Condition 5 is the condition of non-dictatorship. There shall not be an individual who determines the outcome of the social decision regardless of the preferences of the other voters. Clearly, the maximization of social utility social decision function satisfies this criterion. In fact we have complete equality among all voters in the determination of the outcome; neutrality, in that there is no bias in favor of any particular alternative; monotonicity, in that if an individual's preference regarding an alternative changes, the social decision changes in the same direction; and symmetry, in that all voters are treated in exactly the same manner.

According to Y. Murakami, the following strongest versions of three desiderata may be regarded as a sufficient condition for democracy: "A social decision is a democracy if the function is neutral, strongly monotonic and symmetric. The condition should be regarded as the maximum requirement."15 Therefore, the maximization of social utility social decision function satisfies the strongest conditions required for democracy.

The only other consideration is Arrow's Condition 3, the independence of irrelevant alternatives. What this condition demands is that the decision function not be arbitrary in the sense that if candidates are added to or deleted from the option set, this does not influence which of the candidates is selected as the winner provided that the winner is not one of the candidates who is added or deleted. "Suppose that an election is held, with a certain number of candidates in the field, each individual filing his list of preferences, and then one of the candidates dies. Surely the social choice should be made by taking each of the individuals' preference lists, blotting out completely the dead candidate's name, and considering only the orderings of the remaining names in going through the procedure of determining the winner. That is, the choice to be made among the set S of surviving candidates should be independent of the preferences of individuals for candidates not in S. To assume otherwise would be to make the result of the election dependent on the obvious accidental circumstances of whether a candidate died before or after the day of polling."16

Arrow seems to think that a maximization of social utility approach violates his Condition 3 for two reasons:

1) That there are no grounds for the interpersonal comparison of utilities;

2) That the assignment of utility values necessarily is a function of the size of the option set and hence is arbitrary.


Objection (1) can be dismissed on the grounds that a democratic social decision function should treat all the inputs and outputs equally and this requires that one person's preferences and utilities not be given more consideration than anothers. Objection (2) depends on the way that utilities are assigned. Quoting Arrow again: "Assume that for each individual there is always one alternative which is preferred or indifferent to all other conceivable alternatives and one to which all other alternatives are preferred or indifferent. Then, for each individual, the utility indicator can be defined uniquely among the previously defined class, which is unique up to a linear transformation by assigning the utility 1 to the best conceivable alternative and 0 to the worst conceivable alternative. The assignment of values is designed to make individual utilities interpersonally comparable."17

What Arrow is assuming here is that the ideal best and worst alternatives are contained in the actual option set, but then he goes on to cite an example in which this is not true and uses the example to argue that there is arbitrariness involved in the assignment of utilities. He continues: "It is not hard to see that the suggested assignment of utilities is extremely unsatisfactory. Suppose there are altogether three alternatives and three individuals. Let two of the individuals have the utility 1 for the alternative x, .9 for y, and 0 for z; and let the third individual have the utility 1 for y, .5 for x and 0 for z. According to the above criterion, y is preferred to x. [by adding up the utilities, we get 2.5 for x, 2.8 for y and 0 for z.] Clearly, z is a very undesirable alternative since each regards it as worst. If z were blotted out of existence, it should not make any difference to the final outcome; yet, under the proposed rule for assigning utilities to alternatives, doing so would cause the first two individuals to have utility 1 for x and 0 for y, while the third individual has utility 0 for x and 1 for y, so that the ordering by sum of utilities would cause x to be preferred to y."18

Arrow's problem here is that he doesn't follow the assignment rule he, himself, suggests which is to give utility 1 to the best conceivable alternative and 0 to the worst conceivable alternative. If 1 is assigned to the best conceivable alternative and 0 to the worst conceivable alternative, then all comparisons of actual alternatives are with these two poles which remain fixed regardless of the size of the option set and hence all ratings are non-arbitrary. Clearly, if we assume the worst ideal, z, is in the set even though it may not be a viable alternative, then the ratings for x and y do not change and the ratings lose their arbitrariness. Therefore, if each individual assigns the values 0 and 1 to his worst and best conceivable ideals and adds them to the set for the purpose of determining the ratings of the other options, even though these ideals may not be actual options in the sense that they can be chosen, the ratings assigned will not be arbitrary regardless of whether or not certain alternatives are added to or deleted from the set.

If the positive and negative ideals are not included in the set, there is a certain amount of arbitrariness which is introduced as to where to position the most preferred and least preferred candidates. If these candidates are positioned at 1 and 0, respectively, and then one of them drops out of the race, it would be natural to reposition the other candidates so that the second most or least preferred is now positioned at 1 or 0 respectively. Also in the case in which a new candidate enters the race, for example, who is more preferred to the candidates already in it, if the former favorite was already positioned at 1, then all the candidates would have to be readjusted. In order to make the preference ratings non-arbitrary, each individual must include in the option set his positive and negative ideal candidates and assign them values of 1 and 0 respectively. These candidates may or may not be included in the set in actuality. They may or may not even exist. However, their inclusion in the option set guarantees that the preference ratings for the other candidates will be non-arbitrary and not subject to change should a candidate or candidates drop out of or drop into the race. The preference ratings hence become what is called in the literature "independent of irrelevant alternatives."

As Murakami states: "...the origin and unit of measurement of preference intensity is, as we noted, arbitrary. As we choose the Fahrenheit system or the Centigrade system in measuring temperature, so we have to determine the origin and unit of measurement here. In other words, we have to select two 'base' alternatives, such as freezing point and boiling point in the Centigrade system, for which the magnitudes of utility are artificially fixed."19 Choosing the two fixed points as 'the best conceivable alternative' and 'the worst conceivable alternative' is similar to picking the two fixed points in the measurement of temperature as the boiling and freezing points of water.

Arrow's confusion about this issue causes him to come to the conclusion that arbitrariness is involved in the assignment of utilities. From this he concludes that no social welfare function exists which satisfies his basic conditions. :"The point is, in general, that the choice of two particular alternatives to produce given utilities (say 0 and 1) is an arbitrary act, and this arbitrariness is ultimately reflected in the failure of the implied social welfare function to satisfy one of the conditions laid down."20 This is true if the values 0 and 1 are assigned to non-ideal alternatives contained within the set instead of introducing within the set the best and worst ideal alternatives and assigning them the values 1 and 0, repectively. There are important implications to this question of the existence or non-existence of a theoretically sound democratic voting system or social welfare function (which amount on an abstract level to the same thing) since, if they are thought not to exist, then systems which are unjust or discriminatory or imperfect in practice seem to be more justified. However, if it can be shown that a theoretically optimal system does exist, it makes continued justification of existing systems more difficult, and it makes the implementation of the more advanced and principled system only a matter of time.

It is important, therefore, to lay down the rule for assigning utilities or preference ratings very precisely as follows. Assign the value 1 to the best conceivable alternative, the value 0 to the worst conceivable. Geometrically visualize that part of the real axis extending from 0 to 1 and, for each available option, place it on the axis according to its relative position betwen the best and worst alternatives. Then assign to it a rational number according to its placement as in Figure 3.


Figure 3



Now a problem might arise if each individual chose their best and worst ideals according to different criteria. For example, one voter's ideals might be so far removed from the actual candidates as to cluster all preference ratings for the actual candidates close together as in Figure 4.


Figure 4



Another voter might identify his ideals more with the best and worst actual candidates so that his preference rating might look like this:


Figure 5



Actual candidates would be placed at 0 and 1. The problem here is that Figure 5's voter's preferences would carry more weight than Figure 4's and hence Figure 4's voter's input would tend to be discounted. This can be remedied by giving the voters instructions as to how to select their ideals. If each voter selects his ideals according to the same criteria, then there is no problem. If we are selecting a candidate for a representative, legislative body, our best ideal should be one who would vote exactly as we would always, the worst ideal would vote as we would never. For each actual candidate, we would have to judge what percentage of the issues he would vote as we would. Then the preference ratings would all have equal voting power and be non-arbitrary as long as voters voted sincerely, that is, they did not misrepresent themselves.

If voters voted insincerely, that is they wished to maximize the power of their vote by misrepresenting their true preferences, they would apply a linear transformation to their preference rating to "blow it up" so that their highest preference corresponded to a 1 and their lowest to a 0. This would maximize the power of their vote without changing the relative distance between alternatives. For example:


Figure 6




We obtain the linear transformation by solving for a and b in the following equation:

P(mi) = ax + b


We do this by plugging in the new and old values of most preferred and least preferred alternatives into the above equation.


1 = a(.9) + b

0 = a(.1) + b

a=1.25, b=-.125


We get the new values for P(mi) by solving for x in the above equation where x represents the old values. The results are shown in Figure 7:


Figure 7



There would be no risk involved in this type of misrepresentation as opposed to misrepresentation involving changing the relative distances between options.If the voter changes the relative distances between options putting, for instance, his favorite candidate near 1 and all others near 0, he would be taking the chance that one of the other candidates for whom he actually had a higher preference than what he represented and who would have won if he had not misrepresented himself would not be elected, and the actual elected candidate would be someone farther down his list. It would make sense to advise all voters to maximize and hence equalize the power of their vote by applying the above procedure. The final preference rating or utility assigned to each candidate would not be arbitrary since the application of the linear transformation is a non-arbitrary, rational procedure. If, for a particular voter, the candidate added or deleted was his top or bottom -ranked candidate, then the ratings assigned to his other candidates would change after the linear transformation is applied. Therefore, the overall rankings of the election may be shifted and the winner could even change. However, this change of outcome is rational and non-arbitrary when seen in the light that each voter is trying to maximize the effectiveness of his vote and the addition or deletion of his top or bottom-ranked candidate gives him additional opportunities for doing so.

If we assume that each voter votes sincerely, and that each voter picks his positive and negative ideals according to the above procedure, the situation is much simpler. Then the ratings don't change if candidates are added to or deleted from the list and the overall ratings which are the result of the voting process do not change. Then, clearly, Arrow's Condition 3 is satisfied.

If we assume insincere voting with each voter maximizing the power of his vote according to the above procedure, then we have to modify Arrow's Condition 3 to the following: If for some individuals the alternative added or deleted represents their top or bottom-ranked alternative, then the values assigned in the two environments will not be the same and the social rankings and decision may not be the same. However, if the alternative added or deleted does not represent a bottom or top-ranked alternative for any individual voter, then the social decision will be the same in each case. We can see that any change in outcome due to a change in the option set is non-arbitrary and in fact derives from an individual voter's wanting to make the most of his vote in any particular environment.

Going back to Arrow's example again where


P1(x) = P2(x) = 1, P3(x) = .5

P1(y) = P2(y) = .9, P3(y) = 1

P1(z) = P2(z) = P3(z) = 0


Under sincere voting the elimination of z would not change the other ratings since z would still represent the negative ideal for each voter. Under insincere voting it makes sense that, if z were eliminated, the ratings should change in order that each voter maximize the power of his vote and also that the outcome should change.

We, therefore, conclude that the assignment of preference ratings by each individual voter is non-arbitrary and that the social decision when the maximization of utilities rule is utilized, is non-arbitrary. Generalizing to M alternatives and N voters, it can seen that the maximization of utilities social decision function fulfills Arrow's and Murakami's criteria and is strongly democratic.

The case of insincere voting involving a block of voters acting in concert to effect a change in the outcome by distorting not only the intensity of their preference ratings but also the ordering of their preferences will not be considered here except to note that for this to be effective complete a priori knowledge of the actual preferences of the voters is required and also there is a risk to the voters who engage in such a procedure that they will end up worse off than if they would have voted sincerely if their knowledge and hence their calculations are imprecise.

This conclusion overturns Arrow's proof that no social decision function is possible which fulfills his five conditions. As Murakami says: "Arrow tried to establish that these five conditions are not consistent, or that a social decision function satisfying these five conditions is not 'generally possible.' In the first edition of his book, "Social Choice and Individual Values," he presented a 'general possibility theorem' which shows, in his own words, 'that, if no prior assumptions are made about the nature of individual orderings, there is no method of voting which will remove the paradox of voting.'"21 The voting paradox has been solved by the following means:

1) requirement of full information from each voter including preference intensities which can best be represented geometrically;

2) requirement that in a democratic system all preference ratings be treated similarly which means that all geometrical expressions of preference rating be give the same point of origin and length;

3) that the two reference points of best ideal and worst ideal be included in the option set.


In the final analysis, the voting paradox has been solved by a more complete specification of information from the voters.

Number 1 guarantees that the social decision will be based on complete information about individual preferences. Number 2 resolves the controversy over interpersonal comparison of utilities by pointing out that a democracy requires that the inputs be treated equally. Number 3 makes individual preference ratings and hence the social decision independent of irrelevant alternatives. Not only is the complete specification of information important but also a social decision function which utilizes all the information specified in arriving at a decision.

Murakami indicated that the voting paradox could be dissolved in this way. "As a matter of fact, C. Hildreth presented a fairly plausible example of such a rule on the assumption that every individual behaves according to an expected utility hypothesis, and he proved that his example satisfies Condition 1', 2, 3', 4 and 5. His rule includes, among other things, a trick for determining base alternatives for each individual utility index. This implies that an arbitrary principle of numerical interpersonal comparison is being adopted. As in the case of finite ranking rule, such arbitrariness is inevitable in this latter-day version of Bentham-Edgeworth's utility calculus."22 As we have shown arbitrariness is not inevitable in such decision functions. Murakami goes on to state that such a social decision function as Hildreth's conceivably can avoid the voting paradox. He continues, "Hildreth's example established that formally the paradox of social decision can be dissolved if every individual's decision includes not only preference orders but also preference intensities. However, we have to question, as before, the workability of such a social decision-making rule or, in other words, how a society could be informed of every individual's cardinal utility index. We may now pose two questions. In the first place, we may wonder whether every individual has the introspective ability to measure his own preference intensity in numerical terms. Secondly, we may wonder-even if every individual were capable of so measuring introspectively-how a society could obtain access to the numerical values of these individual cardinal utility indices."23 We have shown that by portraying the preference rating graphically, simplification of the process of assigning preference ratings is assured. As far as society obtaining this information, it could obtain it in the same way it obtains it now, by having people go to a voting booth and filling out a ballot, or people could input the information using their home computer and telecommunicate it to a central data bank.


Finally, we consider an example involving a voting on four candidates by three voters. Figure 8 shows the preference ratings for the individual voters both in graphic and in matrix form as well as the social utilities for the four candidates.


Figure 8






In the election of a legislative body, we want to elect a number of candidates not just one. From a theoretical point of view, we can view the process as a selection of m' options out of an option set of M options where m'<M. In the US House of Representatives, for example, the process is carried out by choosing one representative from each geographical area where the districts are defined according to population so that the population as a whole does not vote on the candidates as a whole. The population is segmented into districts and each district elects one representative to represent that district. Does this make sense? To the extent that that representative represents the interests of his particular district, it does make sense. But to the extent that the legislative body is in the business of deciding national policy which represents the population as a whole, it doesn't.

The same holds true for the US Senate where each state is represented by two Senators. It makes sense if the function of the Senators is to act purely in the interests of the state they represent. If their purpose is to represent and decide policy for the nation as a whole, the election process does not make sense. The individual would be better served by a process in which he had a say in the election of all the representatives in the body than he is in having a say just in the election of one representative whose vote is then combined with many others in arriving at a social conclusion or result.

What makes more sense, empowers each individual to a greater extent, and results in a greater consensus among voters in the election of a national legislative body representing all citizens is a process whereby all citizens vote on all candidates and the social decision function selects m' out of the M total candidates to fill the seats. In this way each representative represents all the people although certain representatives will tend to be more closely aligned with certain segments of the population (i.e.those segments that strongly prefer that representative as expressed by their preference rating) and other representatives will be more closely aligned with other segments of the population. Presumably, geographical interests will be represented by candidates most closely aligned with certain geographical areasi.e.candidates from certain areas presumably will be more highly preferred by people living in those areas. Even now certain Representatives or Senators specialize in certain areas and represent national constituencies in those areas even though most of the people they represent will never get a chance to vote for them. For instance, Claude Pepper represents senior citizens on a national basis and not just those in his Florida constituency. The kind of a election process we are proposing allows the population as a whole to have a voice in the selection of the legislative body as a whole and not just in the selection of a small number of representatives to that body and hence is more democratic.

Since it would be a prodigious task to ask each voter to familiarize himself with the credentials of several hundred candidates, sampling techniques similar to polling could be used so that each voter might just be asked to express his preferences over a moderate number of candidates, and then these results could be integrated to get the overall results as if each voter had expressed preferences over each candidate. The results could be made accurate to a predetermined small probability of error using sampling techniques. The result would be that much more information will have been used in determining the results than is used in present-day systems in which tremendous quantities of information are not even collected or considered in determining the results. Using statistical techniques, it could be determined just how much information would be needed to reach a predetermined probability of error, and this information could be collected in the most expeditious manner. Even if the effort made on the part of the voting populace were exactly the same or even quite a bit less than that exerted today, the results would be more greatly accurate than in present-day systems.

Let us consider the general problem. Each voter expresses a preference rating over the total of M candidates. m' are to be selected to represent the population. We assume that the utility of m' candidates for each individual voter is equal to the sum of the utilities of each candidate in the set. For example, the utility of the candidate subset (A,B,C) for the jth voter is


Uj(A,B,C) = Uj(A) + Uj(B) + Uj(C)


We can normalize this by dividing by the number of candidates in the subset.

Similarly, the social utility of any candidate subset would be equal to the sum of the social utilities of each candidate in the subset. Therefore, the social decision function for selecting m' out of M candidates would be: Compute the social utility of each subset of m' candidates. The subset that has the highest social utility is elected. Because of the nature of the social decision function we are considering (the maximization of social utility), we can simplify the above rule to the following: select the m' candidates with the top m' social utility ratings.

The problem from a statistical point of view is this: how many individuals have to rate a particular candidate before that candidate's overall utility rating is known within a predetermined error margin as if the candidate had been rated by the entire population. Assuming the independence of irrelevant alternatives, the social utility information on each candidate can be merged to get a composite preference rating on all the candidates by the population as a whole. A legislative body so constituted would represent the population as a whole and hence there would be no need for more than one legislative body unless there was a desire for a second body to provide a check on the first as is the case in the present becameral US system. Certain representatives would be more closely aligned with certain voters and these voters would represent that legislator's constituency although they might be dispersed geographically over the nation as a whole. Also there would be some alignment as expressed by some degree of preference for each legislator spread out over a large segment of the population, and hence there would be a feeling of responsibility on the part of each legislator to a broad constituency distributed over the entire population and not just to narrow, sectarian interests.

How does this compare with proportional representation? Proportional representation is a system in which representatives are selected by political parties in accordance with the percentage of votes cast for that party in a general election. For example, if 30% of the votes are cast for party A, party A gets 30% of the seats in the legislative body. In this way even fairly small minorities can be represented in the legislature. This system goes beyond the geographically based systems and applies to the voting population as a whole. The problem here is that there is no expression of preferences over the various parties or various candidates, and it is assumed that each representative selected by each party totally represents the voters that voted for that party. This is assuming a lot of intra-party cohesiveness. The advantage of this system is that even a small minority has some representation. In the system represented by the maximization of social utility, the case may arise that a certain minority may derive very little or no utility from the elected representatives, and thus might be effectively disenfranchised. This can be corrected by adopting a variation of the maximum social utility social decision function. We could decide in advance that every voter should at least derive a minimum amount of utility out of the election process so that the amended rule would be: maximize social utility subject to the restriction that the resulting individual utility for any voter may not be less than a certain predetermined amount. Let us call this the minimum individual utility. This introduces a certain measure of equality in the outcome that will also result in a diminution of overall social utility. It will also serve to deter any voting bloc that might use its collective power to secure an outcome more advantageous to itself that would also decrease the utility of some minority.The higher we raise the minimum individual utility, the more equality will exist in the distribution of outcomes and the less will be the overall social utility.

We pose the question: Is it worth a diminution of overall social utility so that there can be a minimum guaranteed individual utility for each individual? The answer is probably yes if the overall social utility does not have to decline by too much in order to raise the utility levels of the worst-off segment of the population. Therefore, we might adopt an even more sophisticated criterion: maximize social utility subject to the constraint that the minimum individual utility is a preset amount, ∂, provided that the total utility added to those individuals supported at this level not be times more than the total utility subtracted from the other individuals. If the limit is exceeded, then systematically reduce the minimum individual utility until the limit is met.

This rule imposes a reduction in welfare on the better off individuals in order to help the less fortunate while at the same time recognizing that there is a trade-off which in some cases may not be worth it if the better-off individuals' utility is decreased by a very large amount in order to secure a very low level of utility for a small minority in each particular voting process. In such cases it probably is better to take care of the minority whose utility is below minimal standards outside of the voting process itself. This can be done by guaranteeing certain rights to all which in effect represents certain minimal standards. These rights cannot be voted away in a voting process by the majority. They precede and transcend any voting process and guarantee a certain safety net below which no one is allowed to fall. These rights may be both political and economic.





The method of maximizing utility which has been applied to political systems can just as well be applied to economic systems. The concept of democracy has meaning not only in a political sense but in an economic sense as well. We can speak of economic democracy as a system in which all citizens are free to enter the marketplace and the workplace and all have an equal share of the power to determine what goes on there. If political democracy is represented by the phrase "one man-one vote," then economic democracy can be represented by the phrase, "one man-one share of the decision making power regarding work and the distribution of goods and services." We visualize a system in which each person's input is considered equally as in a direct political democracy. We do not consider either a collectivist system in which the economic decisions are made by an elite group on behalf of the entire citizenry or even a representative system in which elected officials make the economic decisions but rather the economic equivalent of direct democracy, that is direct participation by each individual in determining both his work choices and his consumption choices. In such a system there would not only be more equality than in the capitalist system but more freedom as well. There would be more equality since economic power would be shared equally and the ownership of wealth and power would not put one in the position to set wages and determine what was to be produced and consumed. It would be more free in the sense that the decisions for each individual about his work and his consumption patterns would be made by him and not for him by market forces beyond his control or by a political body. There would be a jobs market not a labor market. The worker would be empowered to choose his work patterns by making demands on the jobs market rather than being at the behest of employers. Jobs would be available on demand just as, in the consumption market, goods and services are available on demand. It would be a demand economy in which the individual worker would be empowered finally to make the demands. These choices and decisions made by each individual would be treated equally in arriving at an overall social decision rather than unequally based upon economic power which is a function of wealth (which is very unequally distributed) as is presently the case in capitalist systems.

We envision a system in which the outcome for each individual in terms of his work and his consumption is individually tailored to that individual's needs by that individual himself. It is not decided for him by forces beyond him whether those forces are bureaucratic or market forces. Let us be clear that in the so-called free market economy, the choices that one has are limited and shaped by the people who have economic power so that the individual is presented with a set of choices whether they be in terms of job opportunities or in terms of consumable items that he has no direct control over. He has the illusion of freedom simply because he has a range of choices; however, he has not been consulted in determining the set of choices with which he is presented. If there are no job opportunities in the field in which he is most interested, then that individual is not free. If the things he wishes to consume are not available although the market is glutted with items in which he is not interested, then that individual is not free.

This situation of unfreedom is evident in the US today which is supposedly the epitome of freedom. Essentially the only jobs available are technical positions in the military-industrial complex. There are very few jobs available for the graduates in liberal arts. Majors in art, philosophy, history, literature, music go begging in the job market. They are a dime a dozen and end up in many cases not working in their fields of interest. Likewise in the consumption arena. We are not so much given a choice as we are molded by advertising to have our tastes conform to what is profitable for the producer. Rather than being encouraged to develop individual tastes, we are encouraged to conform, to have the same tastes as our neighbors, so as to expedite and maximize the sales of the same products to as many people as possible.

Thus market forces conspire to eliminate true excellence and true freedom of choice from the marketplace and give us instead trivia-trivia which is readily marketed and trivia for which a mass market can be created through mass advertising. In Norman Corwin's book, "Trivializing America," which had to be special ordered by the way because the book store's shelves were filled with trivia, there is a chapter, "Songs Unheard, Films Unseen," which lists some fifty films, all of them excellent, all of them virtually unknown according to the results of a survey of approximately half a million people. According to Corwin, "...most of these films gather dust in the vaults, and have been followed into those dim chambers by at least as many more fine documentaries produced since the 50 were made. A few of them were shown on television, but only eight...were shown in theaters, and then very sparsely. And that is particularly sad in these cruel times, because more often than any other vehicle, the documentary film is inspired by compassion, or energized by a crusading sense of justice. In a time as callous and cynical as the 80's, it is heartening to realize that at least one medium cares-about the handicapped, about the rights of minorities, about underprivileged children, endangered animals, drug addiction, pollution, the environment, victims of all kinds of predation. One can only watch with awed admiration the performance of documentarians who lavish time, energy and funds, sometimes cashing in their insurance policies or borrowing money to complete their films, sometimes risking health and even life, to do work which they hope will accomplish some good through disclosure, interpretation, argument or just plain truth-seeking."24

Meanwhile, while these films go unseen, we are offered the likes of "Rambo" and "Commando," filled to the brim with violence, for our edification. While films of compassion and quality are passed over, films of violence and terror are ubiquitous. As Corwin points out, this does not mean that there is a conspiracy involved in all of this, but there certainly is a willingness to purvey profitable trash on the part of film producers as opposed to creating films which have some integrity and redeeming social value. Let's face it-critics from other societies who talk about American decadence are not all wet. And it is a decadence created for profit and purveyed to mass audiences for their emulation.

The subject of applying social decision functions to economics is known in the literature as welfare economics. According to Murakami, "Welfare economics may be defined as an analysis of social decision in a society where each individual is a consumer. By a consumer we mean an individual who makes his decision concerning only his own expenditure plans. As an expenditure plan is composed of the quantities of commodities to be purchased and consumed, each plan can be expressed as a point in a finite-dimensional Euclidean space, where each dimension represents a quantity of each commodity."25 Not only consumption but work must be included in an individual's preference rating. The preference rating should be over all work-consumption states relevant to the individual.

The totality of all individual consumption-work states comprises the possible social states.Whereas we assume that it's only relevant for an individual to express preferences over his own individual state, Arrow assumes each individual expresses preferences over all other individuals' states as well. "In the present study the objects of choice are social states. The most precise definition of a social state would be a complete description of the amount of each type of commodity in the hands of each individual, the amount of labor to be supplied by each individual, the amount of each productive resource invested in each type of productive activity, and the amounts of various types of collective activity, such as municipal services, diplomacy and its continuation by other means, and the erection of statues to famous men. It is assumed that each individual in the community has a definite ordering of all conceivable social states, in terms of their desirability to him. It is not assumed here that an individual's attitiude toward different social states is determined exclusively by the commodity bundles which accrue to his lot under each. It is simply assumed that the individual orders all social states by whatever standards he deems relevant."26

In the economic democracy we envision, the input from each person is considered to be a preference rating, Pj(w,g), over a number of hypothetical economic states where w represents possible work states and g represents possible consumption states. For the most part these states are personal in that they apply only to that specific individual. However, allocation for and consumption of public resources such as parks, schools, libraries, etc. can also be included. Each economic state consists of a two-fold quantity: the amount, type and quality of work to be done by that individual and a list of goods and services to be obtained in return. Note that each individual's preference rating applies only to himself, and the social decision function must integrate all this data and come up with a solution which assigns a specific economic state to each individual as in Figure 9.


Figure 9

Pj(wj,gj) represents the preference rating of the jth individual over all possible work-consumption states available to him. Uj(wj',gj') represents the utility or satisfaction of the jth individual with the specific state, (wj',gj') selected by the social decision function.


There are many possible social decision functions and many possible ways the efforts of each individual in the workforce can be utilized to provide a variety of goods and services. There are also a variety of ways that these goods and services can be distributed. The social decision function must decide qualitatively and quantitatively how much and what kind of labor each individual is to perform and how much of each good and service is to be provided to each individual.

We consider the application of the system of the maximization of social utility that we used in the political sphere to the economic sphere. Let us assume that each individual has a preference rating over each of his personal economic states. Let us assume that each individual is indifferent to the economic states of other individuals that do not affect him. For example, the consumption pattern and work pattern of another individual should not concern him so long as that individual's preferences are integrated into the overall solution by the social decision function in the same way as everyone else'si.e.everyone's "vote" has equal power and is considered equally by the social decision function. Also, since the individual is presumed to have both private and public needs, preference ratings can include items that are consumed collectively. And there may be a trade-off between needs which could be provided for either privately or publicly. For instance, it might be much more economical to provide for a need publicly rather than privately and vice versa. It might be better and more economical to build one nice community pool, say, than to build a number of smaller private pools. Ultimately, this would be decided by the mechanism itself, that is which solution would maximize social utility.

Each individual rates his personal economic states from highest to lowest in accordance with the same techniques set out earlier. To simplify the problem, let's set some limits. We assume there is a maximum number of hours each individual may work in a given time frame, wmax, and a minimum number, wmin. Additionally, we assume that there is a minimum increment of time which might, for example, be one hour. Therefore, there are wmax-wmin possible work quantities. Note that an individual could get very specific about his work time in terms of how many hours he works in a given day or week, starting and stopping times and type of work. This individual tailoring of work programs is possible because of the ability of computers to integrate large quantities of data and integrate individually tailored programs into a consistent overall solution. The organizational capabilities of computers when programmed correctly mean increased freedom of choice, liberation and individual tailoring of work programs. There can be many degrees of freedom and flexibility. It is up to the social decision function to integrate the information rationally, effectively, taking everyone's interests into account equally and in such a way as to maximize social utility and promote happiness.

Let us assume that there are M different types of goods and services. Let us further assume that each good or service is divided into units. Let the number of units be denoted by a variable, g. Then each individual must make out a list: gj = (g1,g2,...,gM)j which represents the amount of each type of good and service he desires. Therefore, a personal economic state can be represented by sji = (wj,gj)i where wj = (w1,w2,...,wN)j represents the work to be performed in its various dimensions. In general there are a number of possible states for each individual (let us say R states), and it is assumed that each individual has a preference rating over all the possible states. The totality of possible personal economic states can be represented by sj = (s1,s2,...,sR)j. The preference rating for the jth individual, then, would be denoted by Pj(s). Figure 10 illustrates a possible preference rating over personal economic states.


Figure 10


Preference Rating, Pj(s), Over Personal Economic States


0 and 1 represent the worst and best ideal states as perceived by each individual. It might be assumed that an individual's most preferred states would be ones in which he worked relatively less hours and consumed relatively more goods and services. However, these desires will vary over a population. Some people may desire to work longer hours because they love their work and yet are frugal in their consumption. Some may wish to devote their time and energy to some extent to the betterment of society and so might choose to work more and consume less.

Once an individual's preferences have been defined, the problem of integrating all this data and coming up with a social decision is exactly analogous to the political problem. We wish to maximize the utility over society as a whole. There is a constraint involved that wasn't present in the political case and that is that the total work required be sufficient to produce the total amount of goods and services needed by the society as a whole-no more and no less. But the criterion that determined the total amount of work required as well as the total amount of goods and services produced is based on the maximization of social utility. The social decision function would consider all the possible social economic statesi.e.all the possible work states,


and all the possible combinations of goods and services states,

The overall social work-consumption state, Si = (Wi,Gi). For each possible state, the social decision function would compute the social utility,


and select that state with the highest utility. If there are many states that achieve the highest utilityi.e.the state of highest utility is not unique, then the rule might be amended to be to choose that state of highest utility with the least inequality. Arrow's five conditions apply to the economic situation as well as to the political since the situations are analogous. However, one more constraint is needed in addition to Arrow's five. A situation of slavery may arise if the social decision function decides on a state in which one or more individuals are doing more work and receiving less goods and services than some other individuals. Let's just consider two individuals, A and B.


wA > wB

gA < gB


This could come about if B placed high value on receiving a lot and working very little and individual A placed some value on working a lot and receiving very little.Then the social decision function might decide on a state in which A worked more than B and B received more than A which would be unfair although not all such states would be unfair. To prevent this situation we add Condition 6 to Arrow's original 5.


Condition 6: If there is a solution, S*, such that

wA > wB and gA < gB

and A's utility can be increased by decreasing wA and/or increasing gA, then S* is not an acceptable solution.


If wA > wB and gA < gB, then there is a transfer of wealth from A to B. This is not acceptable even if it results in maximizing the overall utility if A would be better off working less and/or consuming more. If, on the other hand, A would be worse off working less and/or consuming more according to his own preference rating, then the solution is acceptable. In this case A prefers to transfer wealth and would be less happy if he could not do this. It represents a non-coercive transfer. Instead of "from each according to his abilities," we have "from each according to his preferences." We have made this kind of transfer of wealth from the more able to the more needy possible, but we have not made it mandatory. Similarly, it is up to the individual to define his needs rather than have them defined for him, with the possibility but not the certainty that all those needs will be met. There is no intermediary such as "the state" to either decide how much an individual should contribute or to define what a person's needs are for them. However, each person is dealt with according to the same fairness criterion. The more able cannot be made to be slaves to the less able, and the less able cannot be made to be slaves to the more able.

In the case that there is a voluntary transfer of wealth possible but no recipient available, that is there are more willing producers than there are willing consumers, it is not fair to decrease someone's happiness by forcing a transfer of wealth upon him, by forcing him to consume. Put differently, a person's opportunity to produce should not be denied to him. Therefore, we have Condition 6'.


Condition 6': If there is a solution, S*, such that wA > wB and

gA < gB and B's utility can be increased by increasing wB and/or

decreasing gB, then S* is not an acceptable solution.


No individual has the option of working less and consuming more thereby forcing another individual to work more and consume less. Likewise, no individual has the option of working more and consuming less thereby forcing another individual to work less and consume more. However, symbiotic relationships in which one individual works more and consumes less, either out of the goodness of his heart or out of a fondness for work, and another individual works less and consumes more, either out of disability or special consumption needs or sheer preference, are admissible. Abuses are possible in terms of voluntary underwork (laziness) and overconsumption. Presumably, these individuals who would only be abusing themselves and not others, it is to be noted, would be candidates for counselling to help them achieve a better and healthier balance between work and consumption.

This system opens up possibilities for charitable giving by connecting up donors with receivers, and, therefore, expedites the desire of those who wish to help those less fortunate than themselves. It, therefore, is a social system which encourages rather than frustrates or impedes individual morality. It allows for those who wish to live according to Jesus' admonition, "Love your neighbor as yourself," but it does not force people to live this way. Also these decisions as to transfers of wealth between individuals are not arbitrary, are not made by some bureaucrat or by someone in authority, but are made according to the social decision function of maximization of social utility. There is no room for abuse due to the arbitrariness of individuals in positions of power because there are no individuals in positions of power. Instead of the "invisible guiding hand" of Adam Smith, we have a rational guiding hand which represents a market system in which the market operates in accordance with ethically defined principles. Decisions are made in accordance with the algorithm which embodies the principles of freedom, equality, fairness, satisfaction and charity.

It should be noted that the first recipients of voluntary giving should not be perfectly capable people who tend to be somewhat lazy and overconsuming, but rather people with bonafide special needs, handicapped people, people who need to consume more and need to work less. In some cases, individuals are incapable of working at all and have consumption needs far above the average person. In the voting procedure, these people with bonafide special needs could be identified so that all voluntary transfers of wealth would go to them first. Then if there were additional wealth to be transferred from workoholics to lazoholics, that would take place after all special needs were met. If all special needs were not met by voluntary transfers, then there should be a "tax" that would allow for the satisfaction of special and communal needs. What this means is that everyone would have to work a little harder and/or consume a little less in order to provide for those members of society who are poor, handicapped, diseased, elderly, children or incapacitated. This could be done in such a way as to reduce the total utility by the least amount or by reducing each individual's utility by an equal amount or by coming down somewhere in between these criteria of equality and maximization of social utility.

The idea of guaranteeing everyone a certain minimal level of utility which then represents a tax on the rest of the population is similar to the ideas discussed in conjunction with proportional representation. Guaranteeing a minimum utility for everyone and then maximizing the overall social utility with respect to that constraint is analogous to proportional representation. Of course this will reduce the overall social utility. So a similar criterion to the one discussed under political systems might be invoked such as the following:

Let i be the minimal individual utility. Let S* be the maximal social utility if there were no minimal individual guarantees. Let n be the number of individuals falling below the minimal level. Let S** be the maximal social utility when minimal individual guarantees are in effect. Then, if

qni < S*-S**

where q is a number decided upon in advance, the corresponding tax and transfer would be put into effect. If the above condition does not hold, then i would be incrementally reduced until the condition does hold. In other words if the utility added to bring everyone up to a minimal level (ni) results in more than a certain number (q) times this utility being subtracted from the overall utility of the population, then it is not considered to be worth it.

Alternatively, the guarantee of certain minimum economic rights could be handled outside the system altogether. A person would have a right to a basic level of food, clothing, shelter and medical care guaranteed by the society. He would also have a right to a job and an obligation to take a job if he could work. These minimum guarantees would have to be provided for by a transfer of wealth or a tax on the rest of the population. These minimum levels would be somewhat arbitrary as would be the value of the quantity, q, above and, therefore, would be subject to political debate and, ultimately, political decision.

Such a system as this would work in industries with well-defined jobs and well-defined products, mature industries so to speak, where the technology is well-known and readily available. What about the creation of new industries, new products and new inventions? People can be employed to pursue research into new and experimental areas. And also individuals might pursue these activities in their spare time. There might be incentives and subsidies provided to people who came up with worthwhile inventions and innovative ideas. And task forces could be delegated to solve problems which society had earmarked as needing to be solved. Invention and innovation might also be encouraged by the provision of "social venture capital"i.e.the provision of capital by society to individuals with ideas for innovation and development deemed worthy by society with the return on the investment jointly split between the inventor-whose work results in a change for the better for society-and society who provided the "capital" for the venture.

Just as in political democracy there is "one man-one vote," in an economic democracy there is "one man-one vote." The vote in the latter case is a preference rating over a list of alternatives just as in the political situation. The differences are that

1) the solution is an individual solution that is tailored to each individual's preferences;

2) the alternatives voted upon represent different allocations of production and consumption that apply only to the individual voter.


The productivity of the entire society has to do with how well the social decision function integrates the labor force so as to minimize the total work involved to achieve any specified level of production. In the economic case the votes as well as the content of the votes and the outcome or solution are all individually based and apply to that particular individual only. In the political case the votes are individually based but the alternatives as well as the outcome may be collective and apply to a large number of individuals or even the entire society. In the economic system we have been describing, everyone would have the right to participate, the right to work, the right to be compensated for that work. Everyone would be equal in terms of their input (preference ratings). Not everyone would be equal in terms of their individual wealth. Some people would work harder to create wealth for themselves. They would be free to pursue that option, and there would be a direct correlation between individual effort and the fruits accruing to the individual from that effort. Some people might not be that interested in material wealth. They would also have the option of not pursuing it and devoting their time and energy to other pursuits, but the system would be open to their working just enough to meet their needs and also to changes in lifestyle if and when they occur. There would be a distribution of individual wealth that would span the economic spectrum. But, the guarantee of minimal economic rights would put an end to poverty. In fact the guarantee of minimal economic rights should be designed so as to bring everyone at least up to the level of non-poverty. Also there would be no unemployment since the necessary work would be distributed over the entire population. This work might not be distributed equally, but that would be in accordance with individual desires and preferences and not in accordance with capitalistic market forces or political dictates. The market here is not a market created for individuals by corporations in the interests of corporations, but a market created for individuals by individuals in the interests of individuals.

Everyone would have equal access to the market place which would consist of the right to choose over a wide variety of patterns of work and consumption. This, ultimately, is the market-place, and represents the epitome of free choice in the market-place. When a person can specify his preferences for work patterns of his own design and desire rather than choosing from the jobs and work-styles that happen to be available, then he is truly free. Similarly, when a person can specify his consumption patterns out of his own mind rather than choosing from the products made available by corporations and the items of consumption can be individually tailored to his needs instead of standardized, then he is truly free in his consumption-style. In present-day capitalism and communism, the jobs available as well as the goods which are produced are decided upon by the owners of the means of production and by the bureaucratic elite, respectively, not by the people at large. The people are then brainwashed through the propaganda of advertising to believe that the choices available to them represent their ultimate desires rather than a limited assortment based on what is advantageous to the people who do have the power to set up and decide on the scope and content of available choices, the people who have economic power, the people who own or manage the means of production.

The proposed system is really a synthesis of free enterprise involving the law of supply and demand and a centrally planned economy involving rational organization of the labor force instead of the "invisible guiding hand" of Adam Smith. This system is centrally organized in the sense that the decisions are made centrally, and, therefore, economies of scale are realized, but the decision making capability does not reside in a group of planners who make their decisions without consulting the people in general but according to an algorithm-a rational guiding hand-which fairly and impartially integrates the individual decisions of the entire population. The decisions, themselves, are individually based and tailored which represents the ultimate in decentralization. Decision making power is evenly distributed over the entire population.

Demand is created by individual consumers not with relation to desires which are artificially molded by corporations through advertising, but with reference to their own innate desires and needs. Supply is created based upon knowledge of the cumulative demand and cumulative work force and is not based upon speculation. Supply equals demand. There is no overproduction or underproduction. There is no scarcity and no waste providing that natural resources are sufficient for the demand. There are no failing businesses due to misjudgment of the market and no lay-offs. Workers are transferred from one area of production to another without discontinuity and as needs shift. There is no generation of excessive surpluses again due to misjudgment of demand or speculation. There would be nothing to speculate about as production would be geared to a demand specified a priori. We would not have a situation as exists in the US in which farmers are putting themselves out of business due to their own productivity by producing bumper crops and thereby driving prices down to the point at which they're losing money. In the proposed system only sufficient effort and resources would be allocated to ensure production for current needs and reserves; there would be no overproduction or waste, no one would suffer from being overproductive and labor and resources that were not needed in farming would go into some other area without loss of income on the part of the farmer.

The law of supply and demand is at work in the satisfaction of individual needs, but scarcity and waste due to misjudgment of those needs is not a problem since those needs are specified directly by each individual in advance. Social needs are then determined by integrating or summing up individual needs. The allocation of resources, labor, and production is determined by consumer demand. However, prices do not fluctuate according to whether there is an under or over supply since supply equals demand and this knowledge is available a priori from the individual votes and not a posteriori from dollars spent in the marketplace.

Each individual has an economic situation that he individually tailors to suit his own needs including the total amount of time he works, how that time is distributed and the type and quantity of non-standardized goods and services he consumes. Non-standardized or individually-tailored consumer goods are possible because assembly lines can be set up which are pre-programmed by computers to individually tailor each item. The central planning is really an organization of the work force in the most rational generate the production of goods and services demanded by consumers with the minimum amount of labor and in the most efficient and equitable manner. Individual needs are taken into account. There is no dictation as to work or consumption by a central planning group.

The negatives of the law of supply and demand-miscalculation of demand resulting in over or under supply which translates to huge surpluses and lost remuneration on the part of the producers on the one hand or scarce supplies and over-priced goods on the other-does not take place. The hardships caused by the free enterprise system both to producers and consumers alike are eliminated. The benefits are retained. Individual initiative is increased over the free enterprise system due to the fact that there are no dead end jobs, no economic exploitation of workers which dampens their initiative and people reap benefits according to the efforts they are willing to put in. The rational integration of work inputs guarantees that workers are involved in a cooperative effort in which the sum is greater than its parts rather than competitive efforts which result in the sum being less than its constituent parts due to the grinding together of individual efforts which competition produces. Therefore, on a strictly social level people are more in harmony with each other. People are rewarded according to their work and according to their contribution to society and not according to their economic position, not according to the amount of capital and wealth they possess. The possession of wealth does not confer economic power as no increase of decision making power is conferred thereby. Wealth can only be used to acquire a higher standard of living, can only be expended in the marketplace. It cannot be used to control the marketplace. There is room in this system for individual creativity to be rewarded when some new invention or discovery which is beneficial to society is generated by an individual. In fact a program of incentives should be set up in order to unleash individual creativity, and since we anticipate a reduction in the average work week due to the rational organization of labor and increased automation, there will be ample time for the creative expression of individuals in all spheres of life from scientific to technological to craftsmanly to artistic to poetic to spiritual. This would be the creative flowering and expression, the unleashing of creative energy by the masses-relieved of drudgery and exploitation and assured of basic social security-which Marx, Marcuse, Fromm and other great philosophers have spoken of. And, yes, there should be societally sponsored incentives and encouragement given to people who come up with socially beneficial creations whether in the form of ideas or things or art. These incentives might be in the form of economic inducements and/or recognition.

This is economic democracy. There is freedom and equality. Freedom to tailor one's individual economic life to one's own needs. Freedom to have a job. Freedom to live a decent life in the case of one who is incapacitated. Economic rights oriented to the basics of life rather than the luxuries-to the needs of the poor rather than the needs of the rich or the strong who would be rich. Economic rights rather than President Reagan's economic Bill of Rights which is nothing more than a manifesto for the powerful and a denial of the validity of the claims of the weak. In the proposed system, there is equality in that everyone's input, needs and preferences are treated equally. No one can take advantage of or exploit another either by means of relative advantage or power or by means of superior intelligence. There is rationality in the sense that the amount of work performed is exactly what is necessary (insofar as can be determined) to produce the quantity and quality of goods and services required. There are no failing businesses or unemployment because supply has exceeded demand. There is no scarcity because demand has exceeded supply. Individual responsibility and initiative is increased because everyone has an equal stake in the system, everyone has the same economic power, no one stands over anyone, the system is very responsive to individually-determined needs both for work and consumption, and individual incentives for outstanding performance and originality and creativity can be built in. The stifling of individual creativity and talents will be eliminated and the unleashing of individual energy and productivity will be facilitated.

With the addition of the guarantee of certain minimal economic rights-the rights of at least decent subsistence level food, clothing, shelter, medical care and education, we have an economic system which is the equivalent of the political system of proportional representation. It is somewhere between socialism (to each according to his work) and communism (to each according to his needs). However, the system contains the possibility of a voluntary transfer from the more able to the less able, from the less needy to the more needy, which represents both pure Christian charity ("Thou shalt love thy neighbor as thyself") and pure communism ("from each according to his abilities, to each according to his needs"). It is also conservative in the sense that to the extent it promotes voluntarism, it reduces taxes. It does show compassion in providing for at least basic human needs at a minimal level for those who can't provide for themselves. It is conservative in that there are no budget deficits, no surpluses (except those thought rationally necessary), no programs that transfer wealth to people who don't need it. There is also no exploitation of labor which is essentially a transfer of wealth from people who work more to people who work less, nor is there exploitation or manipulation of the consumer through advertising and the creation of products not in the interest of the health and welfare of the consumer.

It is important that there are economic rights just as it is important that there are political rights. They represent society's safety net, and they represent security in the sense that they are guaranteed by the society and represent a committment by society to its citizens. A society without rights but with an ill-defined safety net is toying with the peace of mind of its citizens in that it may be capricious and insincere in its lack of commitment to caring for its citizens. In the US the people's welfare is kicked around like a political football with the buck being passed back and forth from local to state to federal government none of which want to assume responsibility for it. It is time that economic rights as well as civil rights are guaranteed on a national, if not a global, level.

Our system contains opportunities and it contains guarantees and it contains possibilities. It is possible to become rich through hard work and ingenuity, but it is not possible to use wealth to control production, investment, the marketplace or other people. It is possible to choose to be decently poor without the fear of poverty, without being marginalized by virtue of being forced into poverty. In present day capitalist society, there is a centrifugal force which pushes people at the margins of society over the brink into extreme states. To be marginal is to be vulnerable to those with power. In the society we are proposing people at the margins would be supported and held inside the bounds of decency not forced into poverty. Our system contains the possibility of helping one's neighbor both as an individual and as a society. It contains the possibility of symbiotic relationships between workoholics and lazoholics. It contains provisions for the sick, handicapped and people with special needs. It allows cooperation. It prevents the exploitation which is inevitable in a competitive system while offering incentives and rewards for innovation and accomplishment. It allows people to determine and pursue their self-interest and their dreams while minimizing the risks involved in a competitive system. It is a system that allows us to love our neighbor instead of predestining us to endless competition with him.

The provision of social insurance makes it unnecessary for each individual to have to pursue wealth in order to provide for his own personal, private security program. In the event of an emergency or catastrophe, he will be provided for so that each individual does not have to amass a large nest-egg just for such a low probability contingency. In this way resources are released for more productive purposes and are utilized by society as a whole more effectively.

This system represents a synthesis of capitalism and communism, individually-based decision making power and central planning, Christian ethics and the pursuit of self-interest, individualism and communalism, political and economic rights, East and West. It embodies the principles of freedom, equality, fairness, rationality and love. It is proposed in the spirit of resolving the conflict between capitalism and communism, the US and the USSR, before the conflict destroys us all in a nuclear holocaust. It is proposed in order to show that there is something better, something beyond both of these systems as they are presently constituted. Seen in this perspective, it would be insane to destroy the world in order to save a system or systems that our vision has already allowed us to transcend. It is not a question of which system is right. Both contain elements of truth. Both exist in history and are adequate for a time but inadequate for all time. A system which synthesizes the best aspects of each should make the preservation of each as a separate entity frozen for all time not worth fighting over. It is the quest for peace, the quest for transcendance, the quest for synthesis which motivates the presentation of this system. If people on both sides could realize that they might all be better off under a new system, a system that does not even exist today except in the human mind, then they will not destroy the world or allow for its destruction with nuclear weapons in order to preserve something which, in the light of new vision, is destined to become a historical relic. Neither will they have an unreasonable fear of being forced to live under another system which itself is not the end-all and be-all, which itself is not permanent and final.