In this chapter I will explore the idea that an important realization of the All-Affected Principle is the idea that persons ought to have capacities for participating in voluntary exchange. This realization occurs in the context where decision making is decentralized. Normally, the All-Affected Principle is meant to give people power over decisions in which their legitimate interests are at stake. And this is usually thought to mean that persons ought to have a say in a centralized collective decision procedure. But centralized decision making is not the only context in which this principle can be realized. It can also be realized in situations in which decisions are more properly made in a decentralized way, allowing different persons to pursue their own interests in their own way, as John Stuart Mill thought.

There are several contexts in which decentralized decision making seems especially appropriate. One, reasonably open markets are desirable ways of organizing production in society because they tend to put resources to their best use. Two, states make decisions by negotiating agreements with one another. Three, relations between persons in clubs, associations, friendships and other groups are created through voluntary agreements. And finally, the relations between various associations and clubs are also often determined by voluntary agreement.

Centralized and decentralized decision making do not come in all-or-nothing forms. A main component of markets are firms, which are centrally organized; and markets are usually properly heavily regulated in a centralized way so as to overcome the ill effects of unequal endowments and market imperfections such as externalities, asymmetries of information, high transaction costs, natural monopoly or monopsony as well as the forms of unfairness typical of imperfect and incomplete markets. International negotiation among states is often – though not often enough – constrained by procedures and methods that are meant to overcome the problems of imperfect information and unfairness through multilateral conferences. And many states are federally organized, which makes room for decentralized agreement making among the federal units. And democratic processes also depend in large degree on decentralized decision making in the form of citizens creating and joining associations such as political parties and interest group associations for democratic purposes or for the purpose of expressing ideas.

I will argue that in some circumstances the capacity for voluntary agreement making can be an adequate realization of the All-Affected Principle. The basic idea is that one can, with this capacity, attempt to advance one’s interests by entering into voluntary agreements with others. The All-Affected Principle can be satisfied if persons are able to enter into agreements with those whose actions affect them or with those who can advance their interests. Persons should have an equal say or a say proportionate to their legitimate interests and this can be realized in voluntary agreement making, or so I shall argue.

In what follows, I attempt to explain and vindicate this proposition. I start by drawing an analogy between democratic decision making traditionally conceived and voluntary agreement making. This helps us see how we can define appropriate procedural norms for the evaluation of processes of voluntary agreement making in both market and international contexts. I argue that fair voluntary agreement in markets and international decision making is a realization of the same principle as fair collective decision making in democracy, only one is for decentralized decision making and the other is for centralized decision making.

I then lay out the Proportionality Principle, which I take to be the fundamental principle underlying both fair collective decision making and fair voluntary agreement making. I give some examples of when the Proportionality Principle departs from equality. And I defend the principle. We can see, however, that there is a deep puzzle about voluntary exchange that makes it especially prone to unfairness. I then say something about institutions that can realize the principles above. The democratic conception of fair voluntary exchange actually can help us see when voluntary exchange needs to be regulated or even replaced in order to achieve the realization of the Proportionality Principle.

Voluntary Agreements and Democratic Participation

The immediate object of agreements are arrangements of rights (understood broadly to include liberties, claims, powers, and immunities) and duties among those who make the agreement. The participants can set up a system of rules of interaction, which establish a complex of rights and duties and to which they are committed. Or they can merely exchange rights (and the associated duties) to particular things. Further purposes of the agreement consist in a change in the division of labor and the production of benefits and burdens among the parties. The parties shape the social world they live in in terms of basic rights and duties, the division of labor, and they alter it so as to bring about benefits and burdens that the social world can achieve. We can think abstractly about the benefits minus the burdens as the surplus the agreement brings about above the level of benefits the status quo realizes. The transaction takes place in the context of a prior set of conditions, which might be thought of as conditions of the process of transaction. Among these are the absence of force and fraud and other background conditions that enable the participants to treat each other as equals. When a person engages in making agreements, contracts, and other arrangements with other people, they are in effect attempting to shape the social world they live in.

If a person buys a house, they rearrange the social world in that others are now excluded from this house unless the owner consents to their coming in and is able to conduct activities in this particular space as they wish, with or without others. The owner has rearranged the social world in that now there is a space in the world that others can only enter with their consent, whereas before there was none. The owner has revised the relationships they have with others. When a person takes a job, they agree to work for a certain reward. They have rearranged the social world in that certain others now work with them under certain conditions and resources are transferred to them while they employ their skill in producing with others certain kinds of goods.

These different transactions have various effects in shaping the social world the person lives in by altering the rights, duties, and powers that different people have in relation to that person. And if we think of the whole series of agreements a person enters into in the course of a life, we can see that they shape the whole social world they live in through this activity. A person constructs the conditions of their life through this activity of rearranging the rules in their social world. They do this, to be sure, with other people who are also engaged in the activity of shaping their lives with others. And the activity of determining a person’s social world is, of course, constrained by the activities of others, both because they have to make agreements with others and because they must respect the social arrangements that others have created. The social world is an aggregate product of all these different activities of social shaping. It is the product of coordination and it is the product of separate activities that are not coordinated but that mesh together, because the activities are engaged in side by side and within the context of a common legal system.

There is an important similarity between the activity of citizenship and the activities of persons in the processes of agreement making with others. Just as a citizen participates in shaping the overall character of the society they live in by participating in collective decision making about the overall collective features of the society, so the ordinary person in everyday life shapes parts of the social world in which they live by engaging in agreement making with others – the difference being that in the case of citizenship each has a small voice in a very large activity, while in private life each has a large voice in relatively small issues.

The justification of these different powers of shaping the social world is grounded in common liberal concerns. Persons have different interests that conflict and they disagree on how best to shape their social worlds; indeed, they are often uncertain about their own interests. We think that people ought to be able to make the basic decisions about how their society is organized, and how their lives with others are organized, by exercising the power to advance their legitimate aims in accordance with their own judgments. They need this power both because they do not want to be imposed upon by people with less understanding and concern for their interests and because they need to explore their interests.

The analogy of democratic citizenship with the activities of persons in decentralized settings is not meant to imply that these decentralized settings ought to be centralized and democratized in the traditional way. My intention is to show that there is an analogy between democratic citizenship and the activities of persons in decentralized settings, not to suggest that we should eliminate the decentralized settings. There are many important values that are realized in decentralized settings that make them unsuitable for at least complete collectivization. The values involved in personal relationships and development and the distinctive values that arise from people cultivating their particular talents and ideas must be given some significant protection from collectivization. And I think, with the tradition of economic theory of Adam Smith and John Maynard Keynes, that some kind of open market system is important for putting resources to productive uses. Furthermore, any effort to replace decentralization in international politics with serious centralization in the near-term future is likely to lead to very unhappy and tyrannical government. The integrity of states, so essential to the realization of the basic interests of persons, must be preserved for a significant time into the future.

Nevertheless, that the democratic idea has a realization in decentralized settings as well as centralized settings will shed light on how we ought to think normatively about voluntary exchange. It will help us develop some basic principles for how to evaluate a system of exchange and provide us with a way of thinking about when and how to reform or change institutions. I elaborate next a picture of the basic idea of fairness in exchange that builds on the democratic analogy.

Fairness in Individual Transactions

The conception of fairness in individual exchange here articulated attempts to avoid the classical natural law approach of equal exchange in value¹ and attempts to develop a procedural conception of fair exchange that goes beyond the standards of absence of coercion and fraud.² The reason for the procedural approach is that the benefits of transactions can be quite heterogeneous and hard to compare outside the points of view of each of the participants. So, it may be very unclear in many circumstances whether the goods exchanged are equal in value in some more objective sense.

Let us first think of an individual exchange as if there were only one exchange for each person’s whole life. The appropriate background fairness conditions for such an exchange consist in the realization of equal capacities for that exchange. This breaks down into two components: equal cognitive conditions, including equal access to information relevant to discerning one’s interests and concerns and abilities to negotiate desirable arrangements, and equal opportunity for exiting or refusing entry into the arrangement.

When we extend the principle to the usual case in which each person engages in a series of many exchanges, the persons must have equal capacities globally in the sense that they start from background conditions that ensure equal capacities for all. This equal background condition need not be fully maintained throughout the series, because earlier agreements persons enter into may curtail opportunities they have in later agreement making. If this is done knowingly, the later agreement making in which there may be some inequality of opportunity is not unfair. Furthermore, individuals may choose to focus on some agreements in which they think of themselves as having much at stake and focus less on other exchanges in which they think of themselves as having a lot less at stake. This stake sensitivity in agreement making will have some importance later.

This realizes a kind of democratic value in everyday life, because the two conditions in the one-shot case specify circumstances in which persons have an equal say in the structuring of their relations with each other. And the global principle of equal capacity gives persons a kind of equal say in the formation of their social lives together with others regarding the contents of the series of agreements they enter into.³ It specifies a kind of condition of global equal bargaining power between parties such that each person has an equal say overall. It does not imply equal bargaining power in each agreement-making context but only over the sum of agreements a person enters into.

This principle of equal capacity is meant to give persons equal power in the process of the creation of the informal social world they live in. Their interests are, roughly speaking, equally at stake in the system of agreement-making overall. This is like the idea of equality in democratic collective decision making in which persons are to have equal power because we think that, for the most part, their interests are equally in play in the political system.

We achieve this condition of equal power in agreement making by making sure that people have the education and resources that enable them to discern their interests and exit or refuse transactions and enter others that advance their interests. Education, basic needs provision, and other goods give people opportunities to choose among transactions by enhancing their bargaining power.

The principle of equal opportunity for exit is meant to be a special principle for societies like the modern nation-state in which there is a rough equality of stakes for persons overall in the process of agreement making, at least for the great majority of people. There are a number of qualifications that need to be explained concerning this principle. First, we need to say what kinds of things can produce inequality in the outcomes of the processes of agreement making. One, those who knowingly exert themselves and make use of their opportunities for the sake of a particular good are more likely to achieve that good, other things being equal, than those who knowingly do not exert themselves for the sake of that good. Two, more controversially, it makes sense for differences in natural talent to bring about some inequalities in social power that arise from occupying different positions in the division of labor. With Rawls, we should not accept that differences in talent should bring about significant differences in income or wealth. But two people who are competing for a particular social position may justly end up in different positions in the division of labor if differences in their natural talents imply that they should occupy those different positions, even if one of the positions is more desirable and interesting than the other.

The question is, what can justify this? It can be justified by the principle that we ought to think inequality can be justified over equality when all are better off and by the idea that it is important that people be able to realize their talents. The realization of natural talent implies that persons are benefited when they exercise those talents. To require that people not be able to exercise their talents so that they have no more bargaining power than others would be to make others worse off as well as the person who is deprived of the opportunity to exercise talent. There is injustice here, but it is desirable in order to avoid leveling down.⁴

The Proportionality Principle

The second observation is that the principle of equal opportunity is really a special case of a more general principle, which also governs collective decision making. The more general principle is that persons ought to have a say in a transaction in proportion to the legitimate stakes they have in the transaction.⁵ This principle is narrower than Mark Warren’s principle of proportionality in this volume. His is a more general principle of equity and includes distribution in proportion to need or desert. My principle only regulates the distribution of power. The principle of equal opportunity is a principle that respects this more general principle for the special case of overall equal legitimate stakes.

The stakes a person has in an agreement or set of agreements consists of the range of potential legitimate interest affecting outcomes of that agreement. This includes the effect on interests if there is no agreement and the effects on interests of the various agreements available in the circumstances. We must distinguish between, on the one hand, ex ante stakes a person has in a transaction or system of transactions and, on the other hand, the actual stakes in a particular transaction. Ex ante stakes are the extent to which the transaction or the system can advance the legitimate interests of each party independent of the distribution of resources among the parties. This varies according to context in the sense that it depends on the real possibilities of the system or the transaction. Actual stakes will depend on the interests in the transaction as well as the distribution of resources among the parties, which will determine what each party brings to the transactions. In the context of a particular transaction, a person with very few resources will have higher actual stakes in the agreement because they have fewer resources to fall back on or to offer others in alternative transactions. The transaction matters all the more. A person with a lot of resources has less need of the particular transaction and so has a lesser stake.

This principle applies to democratic decision making as well as voluntary exchange. In the case of democratic decision making, the stakes are the range of interests affected by plausible outcomes that can be achieved by the collective decision process. The equal distribution of power is a special case of its proportional distribution.⁶ In the case of democratic decision making we see the application of proportionality in two ways. First, some decisions in a federalist system are made in a way that gives all persons an equal say; other decisions give greater weight to persons in local jurisdictions. A natural interpretation is that when people’s interests are affected in a roughly equal way overall, they have an equal say, and when some persons’ interests are much more affected than others (as in local political decisions), they have significantly more say. This is institutionalized in the system of federal institutions.

Second, proportionality operates at the level of individual democratic decisions as well, but only informally and according to the judgment of the individual. That is, when people with equal political power think that some issue is of great importance, they devote many more resources to that issue and fewer resources to things that matter less. The democratic process overall, on the assumption that the differences cancel each other out when all issues are taken into account, gives people broadly egalitarian opportunities to trade with people through logrolling. It is equality of stakes in the collective decision making overall that justifies the equal vote.

Voluntary agreement making is different from collective decision making. There are usually two distinct dimensions: one is whether to make an agreement at all, and the other is what the content of the agreement is. It is for this reason that the definition of stakes is different from that of collective decision making. The basic determinant of relative stakes in the case of particular actual voluntary exchange is the relative significance of the non-agreement point. This means that how much a person benefits from an agreement need not determine stakes. How the surplus of the agreement is divided up will depend on what the parties agree to. But the initial condition of the agreement, which are the values or disvalues of no agreement for each of the parties, is external to the agreement and thus can provide an external reference point for determining the stakes.

The principle of equal capacity is based on the idea that persons have equal ex ante stakes in a system of transactions. In society virtually all the fundamental interests we have (with respect to voluntary exchange) are in play in the whole system of voluntary exchanges. The non-agreement point is essentially life outside society. It then requires that the distribution of resources be adjusted so as to achieve equal capacity. This is a special case of the more general principle that persons ought to have capacities that are proportionate to their ex ante stakes. Someone who has a lot less stake in a transaction ought to have less power over it than one who has more stake.

The principle of proportionality is qualified in the two ways of the equal capacities principle above. A further qualification results from a worry, pointed out by Nozick, that this gives people power over the intimate choices of others. If one person loves another deeply and the other chooses to marry a third person, then the first person has great stakes in the second person’s choice. Does this mean that the first person ought to have a say in the choice? This seems really implausible. We should allay this worry with the idea that some interests are protected interests in the sense that a person may choose to fulfill these interests without the leave of others. This might be the case, for example, in choices of friends or life partners and perhaps choices about basic questions of conscience.⁷

A clarification is in order here. I take stakes to be defined in terms of well-being and I think of well-being as being essentially an objective state. But the stakes involved in the Proportionality Principle in a pluralistic society must proceed with a fairly generic conception of the interests of persons, since there is likely to be significant disagreement and uncertainty regarding what particular interests persons have. Persons have interests in health, friendship, resources, education, and other generic goods and the principle of proportionality allows them to explore those interests with the power that it assigns them.

Some Cases in Which the Proportionality Principle Applies

I have already pointed out that sometimes proportionality applies to politics in a non- egalitarian way. But there are important contexts in which persons or groups come together to make important agreements where they are not really part of an extensive and unified system of agreement making and in which they have different stakes.

First, states with different populations make negotiated agreements. Larger societies have a greater stake than smaller societies in their negotiations. The democratic principle in voluntary exchange, as in the case of collective decision making, is an individualistic principle. Hence the theoretical challenge is to aggregate the interests of the members of the society in a way that is relevant to defining the stakes each society has.

Second, agreement making among states in the making of international law is fairly modest compared to the domestic legislation states engage in and relatively modest compared to the amount of agreements individuals enter into in modern societies. In climate change negotiations, we can see that certain societies have much more at stake in these negotiations than other societies. The average Bangladeshi has a great deal more at stake than the United States citizen in the medium short run, which is a function of the extent to which the society is affected by non-agreement. Proportionality would suggest that Bangladesh has more of a say in the content of its agreement with the United States on climate change (abstracting from difference in size of population).

Third, people enter into contractual arrangements across different societies, such as when a firm from a wealthy society sets up shop in a poor country to employ relatively cheap labor. The poor person has a significant amount at stake in this activity, presumably more than the members of the company and its consumers. The members of the company do have an important stake in setting up desirable arrangements because they must make sure that the company’s costs do not go too high. The consumers have an interest in paying lower prices and this is what is partly driving the move to sweatshop labor. But perhaps here the consumer’s interest in lower prices is a far lesser stake than the poor person’s interest in having a decent wage. Of course, even here there is a lot more complexity: some poor consumers may have a pretty big stake in low-priced goods such as shoes.

Argument for the Proportionality Principle

Intuitively, when we are setting up an arrangement with another person whom we know well and we know that the other person has a lot at stake while we have less, we often give more say to that person over the arrangement: “This matters more to you [or for you], so you should decide.”

There are really two main connected theoretical arguments for this principle that start from the All-Affected Principle, suitably modified. The first argument is essentially a welfarist argument. The idea is that people have an important interest in taking care of their own interests. The basic reasons are that they know their interests better than others do, they have incentives to concern themselves with their own interests, and they have the capacity to learn from mistakes about their own interests. To the extent that more is at stake in an agreement for a person than for others, they have more of their interests at stake and so there is reason for them to take greater care. But this also gives reason to others to give them a greater say in the agreement. The welfare of persons is better advanced when they have a greater say in those issues in which they have a greater interest.

There is an egalitarian argument for this as well, at least under normal circumstances with normal people. The idea is that a principle that extends a greater say to those who have more interests at stake in agreements will, over the long run, equalize the say of persons over the matters that affect their interests. This is based on the idea that over the long run normal people have roughly equal stakes in how the external world is organized. The principle gives people an equal say over the external world they live in when we think of its global application.

One question that we cannot address here is whether the principle we are defending is essentially an instrumentally defended principle or whether there may be some intrinsic value to implementing the principle. The arguments above suggest an instrumental cast, but the democratic interpretation of equality of opportunity offered above suggests that there may be intrinsic value in implementing the idea.

Before moving on, I want to address three important concerns. First, the principle I have defended is an individualistic principle in a limited sense. It focuses on the distribution of power to individuals in the process of exchange and whole systems of exchange. In this respect it is much like the democratic principle for collective decision making. And it is committed to a kind of deeper individualism, which asserts that the ultimate values that we are dealing with are the well-being of the individual and justice among individual persons. Even the notion of the common good will have an individualistic interpretation on this account. But the principles of justice are not individualistic; they specify a collective order with the proper distribution of benefits and burdens among persons. I also reject individualism regarding the content of well-being. I want to argue that many, or perhaps most, of the important interests people have consist in their social relations with others. Friendship is an obvious case of this, but so is the interest in being recognized and affirmed as an equal and the interest in making the world a home for oneself, or the interest in seeing that one is making a contribution to valuable cooperative activities. The social nature of our interests is compatible with the democratic conception of voluntary exchange.

A second clarification has to do with the nature of the duties people have with regard to this principle of the proportional distribution of power in the system of exchange. The principle applies in the first instance to institutions just as much as conventionally democratic collective decision-making principles. Another departure from individualism obtains here because individuals have jointly held duties to bring about, maintain, and improve these institutions. Since the duties are jointly held, each person has a duty to do their fair share in bringing about, maintaining, reforming, and complying with the institutions that distribute power. In reasonably well-functioning institutions, individuals may go about pursuing their partial interests as they see fit.

One important question is: What kind of duties do people have to take up the slack when others fail to do their share? It is not the case that each person must do everything possible to make up for every inequality of power they have not produced, but they do have duties to do their share – but what is that when others fail to do theirs? I do not have a complete answer for this, and it will depend on the specific problem confronting a person. If a particular person is in dire straits, an agent may have the duty to do quite a bit. If a particular person unjustly has less power than other persons, it is not clear what it is fair to ask a better-situated person to do. But it is not normally everything that can be done. The duties are jointly held and so fairness in the distribution of duties plays a role even in the case of taking up the slack.

A final clarification repeats what I have already said. The principle I have defended is a principle of equal distribution of power among individuals. But much of decentralized decision making involves interactions between groups such as firms or states and sometimes between these and individuals. I do think the individualistic principle still applies here, but it is not between the firm and the person as if these are two individuals. The firm is a group of persons and so the question of power ultimately comes down to the distribution of power between the persons in the firm and the persons the firm is negotiating with. For a complete analysis of this kind of interaction we will need a method of aggregating the persons’ stakes to arrive at an account of the proper distribution of power between the firm and the person. And this method will vary depending on the institution in question. For example, something close to an additive method may be acceptable in the case of the relations between states with different populations. But additivity will not work in the case of firms as they are constituted in the United States, where there is significant conflict between the different elements. An account of the proper methods of aggregation of stakes is absolutely essential to the full articulation of the principle I am defending in this paper, but I am not in a position to offer one now.⁸

The Puzzle

One of the most fundamental puzzles in a system of free transactions is that the principle of power proportionate to stakes is prone to violation in many normal circumstances. For example, if you have two persons who depend on making an agreement to advance certain interests, the one who has the least stake will often have more power. This is because they can more easily afford no agreement. But this means that power is inversely proportioned to stakes in a scheme of free transactions, while the normative principle tells us that power ought to be proportioned to stakes.⁹ This is the fundamental puzzle that Marx pointed to in the relations between capital and labor. The laborer has a great deal at stake in a transaction since their life depends on it, while the capitalist has a lot less at stake. The consequence is that the capitalist has more bargaining power.¹⁰

A scheme of equal opportunity in a context of equal ex ante stakes provides a unique resolution to this problem. Power cannot be inversely proportioned to stakes in this context overall. Proportionality can still continue to function in this context but it is informally applied by the parties. For example, if two parties have equal capacity overall and one party has more at stake in a particular transaction than another because of difference of interest, the usual consequence is that the one with less at stake will devote less attention to the issue (in order to devote more time to other issues) while the one with more at stake will devote more time. To the extent that each has overall equal stakes and capacities, the disproportions will cancel out. In this way, the overall proportion between capacity and interest is maintained.

What Marx was concerned with was that the capitalists, who have less actual stake in each particular transaction with a particular worker, also have many more resources to devote to the making of the transaction. This compounds the initial disproportion.

Institutions

It looks like this problem does occur in perfectly competitive and complete markets but only in a highly attenuated way. In markets where there are lots of buyers and sellers and no impediments to exchange and there is no incompleteness so that credit and insurance are available to everyone, there is no difference in bargaining power. And a genuine inequality of stakes only obtains as a result of unequal initial endowments, but this difference may not normally count for much when there is unlimited availability of credit and insurance. Only when there is extreme difference of endowments, including natural endowments, will complete markets be very unfair to participants.

The puzzle is nevertheless a deep one, because in fact we always observe significant divergence from complete and perfect markets. And many of the principal institutions in what we call the market are specifically designed to overcome imperfections in actual markets. Indeed, one of the central institutions of modern capitalism, the firm, is essentially a response to market imperfections.¹¹ At the same time, some form of a system of voluntary transactions is a highly effective tool in figuring out how to allocate resources to their most productive uses. We do not have a clear idea of an alternative system for the allocation of scarce resources that does not involve the use of voluntary market exchange playing a central role.

The question is, how can we make use of this tool for allocating resources in an efficient way but without sacrificing fairness? There are really four basic kinds of mechanisms that we can introduce to try to ensure a kind of fairness in the process of making agreements. The first is redistribution so as to achieve a kind of background distribution of resources that enables people to meet each other on reasonably equal terms. The second is direct regulation of the employment relation. The third is direct regulation of the bargaining process so as to help achieve a kind of balance between worker and employer. The fourth involves some kind of collectivization in the decision making in economic institutions. Here we replace bargaining with collective decision making among the parties and try to achieve equality in the collective decision process. We can think of a number of institutions that might help to alleviate the disproportionality involved in voluntary transactions.

One basic mechanism for overcoming the problem is redistribution of resources from those who have low stakes to those who have high stakes so as to equalize the stakes as much as possible, or so as to provide some kind of basic floor of resources to protect those who have very few resources and thus lessen the stakes for the less well positioned. The welfare state comes to mind as a complex of institutions that are designed to lessen the stakes for workers when they choose where and how to work.

The provision of public education is a way to ensure that persons are able to develop their abilities so as to participate as equals in economic life. Unemployment insurance and income maintenance programs imply that the alternative to work at a particular job is not starvation or penury but some kind of minimally decent life. It is important to notice one distinctive function of the welfare state is that it enhances the power of persons to shape their social world. It enhances people’s power to shape their work conditions and wages. The dismantling of the welfare state is in effect a kind of authoritarian strategy to disempower workers.

A second mechanism is direct regulation of the employment relation. Minimum wage legislation, legislation concerning working conditions, legislation concerning hours of work, and legislation constraining how workers can be fired are all direct regulations imposed by the political society. They do this because workers tend to be at a bargaining disadvantage in negotiating the work agreement with employers. These constraints on work conditions increase the bargaining power of workers by decreasing the bargaining power of the employer since they lessen the alternatives available to employers.

The third kind of mechanism is direct regulation of the bargaining process. Here the idea is to make use of facts that are already present without having to redistribute resources such as the great number of workers and their potential for organization. One institution for altering the conditions under which agreement to work is made to avoid great unfairness is union organization. The effect of union organization is not to lessen the stakes of the worker in work, but to increase the stakes of capital in refusing terms of work by requiring capital to deal with a lot of workers at the same time. Unions do this by themselves, but the state can help by setting up rules for negotiation that are favorable to unions. Notice that this mechanism introduces elements of voluntary agreement and elements of collective decision making. Workers participate in unions as participants in a collective decision-making process, but that collective agent then bargains with employers.¹²

A fourth kind of mechanism is local collectivization. I do not mean here to say that the whole structure of society is collectivized. I am assuming some kind of decentralized market system for allocating resources to their most efficient uses. The idea here is to replace the process of bargaining with all its inequalities with a collective decision-making method that includes unions, laborers, managers, and capitalists.¹³ This seems to be the mechanism at work in the German codetermination system. The particular value of the approach I have taken makes egalitarian voluntary exchange and egalitarian collective decision making derive from a single basic principle. They are two distinct types of realizations of the principle of proportionality. What this approach suggests is that one kind of decision making can conceivably be a remedy for a deficiency in the other. I have in mind, for example, various kinds of workplace democracy. Workplace democracy can be a partial remedy for imperfections in the market that put workers at a serious disadvantage in the bargaining process. Consider the familiar case of partial monopsony in which there are lot of workers but it is not easy for them to move from job to job, either because they have to make asset-specific investments or because they have imperfect information or because of high transaction costs in moving from one job to another. And let us further suppose that the circumstance workers find themselves in is not the result of a set of fully understood choices made earlier. The workers are at an unjust disadvantage. The consequence of this is that they tend to have relatively little say in the process of agreeing to a work contract. They have little say over the conditions of work. We can possibly rectify this deficiency by giving them more direct and locally collective power over the conditions of work, as in workplace democracy. The approach I am suggesting has this as a result, and I think it is a good result.

But the approach I am suggesting is not a general argument for workplace democracy, because the conditions under which the argument above works are not general conditions. In fact, I think that there is continuity here. We can imagine partial deficiencies remedied by partial requirements for workers’ participation. In cases where the deficiencies are great, the requirement of fairness suggests a greater say. In case where they are present but not so great, they suggest a lesser say. Union organization is in some ways a lesser form of participation and it may be defensible as a response to lesser problems.

Conclusion

I have attempted to present here a vindication of the idea that the possession of a capacity to enter voluntary exchange is a realization of the All-Affected Principle. It is a way in which one can try to advance one’s interests in the context of other persons. And I have tried to show that the fundamental principle for evaluating voluntary exchange is the principle of proportionality and that this is the same principle for evaluating collective decision making. I have tried also to show what some of the benefits of thinking of exchange in this way are.