A
Quiet Revolution in Welfare Economics- by Michael Albert and Robin Hahnel
According to the traditional paradigm, markets economize on unneeded information. Our new paradigm sees them as deleting information critical for self-managed, solidaritous decision making. The traditional paradigm celebrates the motivational miracle of competition that supposedly harnesses egotism and greed for socially beneficial tasks. Our new paradigm helps elucidate how market roles undermine cooperative behavior and relationships built on mutual respect and solidarity.
Markets simultaneously require competitive behavior and prohibit cooperation (not to be confused with collusion) as irrational by creating a direct opposition of interests between seller and buyer. The interest of the seller is to sell a-good or service that is least troublesome to provide at the highest possible price. The interest of the buyer is to buy the most beneficial good or service at the lowest possible price. Since "least troublesome" seldom coincides with "most beneficial" the opposition of interests is complete over both the good or service exchanged and what is paid for it. In market exchange, neither party can afford to be concerned with the needs and desires of the other, except in a mercenary sense. To be concerned with the well-being of one's "partner" in exchange undermines the functioning of the market mechanism. To put it starkly, market allocation establishes a war of each against all.
It is interesting that markets' greatest critics and admirers have agreed on this point. While the critic writes:
Every product is a bait with which to seduce away the other's very being, his money; every real and possible need is a weakness which will lead the fly to the gluepot ... every need is an opportunity to approach one's neighbor under the guise of the utmost amiability and say to him: "Dear friend, I give you what you need, but you know the conditio sine qua non; you know the ink in which you have to sign yourself over to me; in providing for your pleasure, I fleece you." 38
The admirer concurs:
It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard for their own interest. We address ourselves, not to their humanity but to their self-love, and never talk to them of our own necessities but of their advantage. 39
Market competition motivates economic effort based on self-interest, not concern for others who are similarly concerned for you. Market competition forces people to think of ways to take advantage of others they buy from and sell to and with whom they compete for suppliers and buyers. In sum, market roles require participants to practice antisolidaritous behavior on a daily basis, which cognitive dissonance then translates into consciousness. And while solidarity is expressed by the maxim "do unto others as you would have them do unto you," market competition expresses the maxim "do in others before they do you in."
Karl Marx summarized the problem succinctly over 130 years ago when he wrote that those who wanted markets in socialism want competition without the pernicious effects of competition. 40 More recently Gar Alperowitz; pointed out that "as long as the social and economic security of any economic unit is not guaranteed, it is likely to function to protect (and out of insecurity, to extend) its own special status-quo interests--even when they run counter to the broader interests of society." 41 But this is precisely the situation a market system creates. Under market competition managers of plants that pollute-whether they be managing in the interests of stockholders or workers-dissimulate with good reason since pollution taxes and more ecological technology would lower profits. Managers of plants that create shoddy and dangerous products have an incentive to engage in deceptive advertising to preserve demand for their products. And those whose livelihoods depend on the automobile industry will continue to oppose changes they know to be desirable in transportation systems as long as market competition means they would bear the burden of the social costs of transition in the loss of income and dignity that comes from unemployment.
To generalize, markets systematically establish divergences between individual and societal well-being. They establish an incentive to pursue individual interest at the expense of social interest because they guarantee that the rest of society cannot be relied on to safeguard one's individual welfare. Markets create a kind of lowest common denominator consciousness. "Under market conditions a minority of the workers in an industryperhaps even one enterprise--can impose its preferences on all the rest." If one firm chooses to use deceptive advertising or lower the quality of the product in imperceptible ways, "all the other firms must follow suit-or find themselves driven out of business." 42
In sum, the information markets delete, the roles they define, and the incentives they establish all combine to re ward pursuit of narrow self-interest at others' expense. It is not that self-interest and competition against others are the only forms of behavior of which people are capable. It is that these are the only forms of behavior rewarded by markets! Just as markets systematically delete information necessary for self-managed, solidaritous decision making, markets establish roles and incentives that penalize solidaritous behavior and reward its opposite.
At a minimum, it is important to recognize these properties of markets irrespective of whether or not better allocative mechanisms exist. That is, even should all other feasible allocative mechanisms have worse failings, that is no reason to deny weaknesses inherent in market institutions. The cybernetic and incentive properties of markets not only do not promote solidarity, they are systematically destructive of any solidarity that may arise from other spheres of social intercourse. This is true regardless of whether or not they are the least of the allocative evils. And this is true irrespective of whether or not the prices generated by competitive markets are accurate reflections of the relative social costs and benefits associated with the production and use of different commodities. But, as we will see, markets are far less deserving of this reputation for accuracy than commonly supposed.
As we discussed in chapter 3 there has long been a caveat to the thesis that competitive market prices are efficiency prices: namely, that no externalities or public goods are involved. At the risk of belaboring the commonplace, the problem with market prices in the case of externalities and public goods is that the balancing of costs and benefits in market transactions is done only by the buyer and seller. If others beside the buyer and seller are affected, there is every reason to believe that market competition excludes their interests from consideration. In this case, effects on others beside the buyer and seller are not captured by market prices.
Buyers evaluate the worth of a commodity in terms of how its consumption affects them alone, paying the price if the benefit is deemed more than the loss of other commodities the purchase out of a limited income implies. Sellers sell a commodity if their loss in well-being is deemed less than their gain from other commodities the sale enables them to purchase. For either to do otherwise-for either side to take account of benefits or costs to others-would be to engage in acts of charity with no reason to expect reciprocation and, indeed, with every reason to expect a "fleecing."
"With no reason to expect reciprocation" is crucial here. We are not saying people are unable to make decisions that take others' interests into accountthough as discussed above they have little relevant information that would enable them to do so in market economies But we agree with conventional wisdom that people are unlikely to be mutually supportive when they have every reason to expect that others are not reciprocating. To extend oneself, at cost to one's self-interest, in full knowledge that reciprocation will not be forthcoming and that, indeed, one will simply be taken advantage of, is to engage in the "saintly" behavior of continuously turning the other cheek. But this is not to say that people are incapable of taking others' interest to heart when they are free from fears of being taken advantage of, or better, when they have every reason to expect others to take their interest into account, More to the point, it is not to say that if people find themselves in a situation in which it is unlikely they can succeed in advancing their own interests except by taking the interest of others as seriously as their own, that they will be unable to do so.
In any case, if markets misestimate the social worth of commodities whose production or consumption by one agent generates effects for other agents because they do not provide means for joint, or social expression of desires, why can't a number of affected agents band together and become joint buyers of goods whose consumption has external effects? The answer is that they can and will whenever the result of failure to do so is so grossly inadequate as to drive them to forge a "makeshift social structure" 43 outside the institutional structures provided by the market. Which brings us to the questions posed in chapter 3:
1. How rare, or how prevalent are externalities and public goods?
2. Do market mechanisms create incentives for people to generate negative externalities?
3. To what extent can previously known makeshift social structures be expected to resolve these problems?
4. To what extent could these problems be ameliorated by any of the "incentive-compatible mechanisms" that have been recently proposed?
We now discuss what our new welfare paradigm has to contribute in these regards.
In chapter 5 we elaborated a theory of human sociality emphasizing that most of our needs and potentials were developed and satisfied only through social interaction, and that human consciousness contributes special features to our sociality beyond what is enjoyed by other social animals. We explained how even Robinson Crusoe--of great fame in the traditional welfare paradigm-was unintelligible apart from the society from which he had become separated.The fact is ... most of the millions of acts of production and consumption in which we daily engage involve externalities .... Externalities are totally pervasive. 44
Moreover, our paradigm makes clear that it is not the nature of goods that is at issue. It is not that some goods are public and others are private, that some goods have externalities and others do not. To think in terms of the supposed properties of goods is to misconceive the issue. The question is to what extent various kinds of human activities have more or less pervasive effects on others that go unconsidered in the decision-making, system established by markets in conjunction with a particular definition of property rights. To what extent do the choices and actions of economic actors cast a wider or narrower net of effects on others than we can reasonably expect to be evaluated by a market-organized, decision-making process?
The issue is best conceived not as one of the characteristics of goods, but as one of the characteristics of markets. It is markets in the context of a system of property rights that establish who gets to make what decisions in market economies. If the markets and property rights confer on an economic actor the right to make a decision that affects the well-being of other economic actors, we have an external effect. And this will be the case whenever the market-property-rights system grants economic actors the right to make a decision that changes the state of the world in some way that matters to someone other than themselves. 45
Moreover, in assessing "breadth of effects," our paradigm reminds us that economic actions have human as well as material inputs and outputs and transform the states of both individual and group human characteristics as well as physical implements. This conception of human activity broadens the possibilities for external effects considerably. One actor's choice may affect flow or state variables of relevance to other actors, and those flow or state variables may be physical or human in form. Any time actors affect the state of the world we all live in they affect the possibilities we all face. It is not so much whether external effects do or do not exist, but how strong the external effects of any particular action may be.
Our paradigm makes this point symbolically by defining individual well-being functions, U(t), on the activity space of all human action, Ak(i,t) for i = 1.... H, rather than only on the activity space of actions in which the individual participates directly, that is where i ranges only over those groups in which i directly participates. 46 But the emphasis on human sociality and the implication that much of what determines our individual well-being hinges on the situations of others with whom we will/can interact implies a subtle problem that few traditional theorists have confronted. As is frequently the case, Arrow is an exception:
There is one deep problem in the interpretation of externalities which can only be signalled here. What aspects of others' behavior do we consider as affecting a utility function? If we take a hard-boiled revealed preference attitude, then if an individual expends resources in supporting legislation regulating another's behavior, it must be assumed that that behavior affects his utility. Yet in the cases that students of criminal law call "crimes without victims," such as homosexuality or drug-taking, there is no direct relation between the parties. Do we have to extend the concept of externality to all matters that an individual cares about? Or, in the spirit of John Stuart Mill, is there a second-order value judgement which excludes some of these preferences from the formation of social policy as being illegitimate infringements of indiviudal freedom. 47
While this dilemma is more often than not ignored by traditional theorists, it is a real dilemma. And as Arrow implies, the traditional paradigm offers very little help in resolving the dilemma other than posing an arbitrary choice of whether or not to impose a "second-order value judgement." Certainly there is nothing in the traditional paradigm that helps us decide when and where to impose such a value judgment, even should we decide it were justifiable to do so.
While paradigms neither create nor resolve real dilemmas, they can be more or less helpful in clarifying their origins and what is at stake in their resolution. Our paradigm implies that the dilemma occurs far more frequently than the traditional paradigm would lead one to suspect. Whereas the traditional paradigm disguises conflicts between different economic actors' "individual freedom," ours highlights the prevalence of the problem. Our paradigm also suggests that the degree of conflict varies widely. We do not expect that an individual's potential well-being will be affected to an equal extent by all other human activity irrespective of who is directly involved in performing the activity and how distant the activity is in time and space. We would expect, ceteris paribus, individuals would be more affected by activities that they and their closer acquaintances are involved in and by activities taking place closer to them in space and time. Our paradigm casts doubt on the plausibility of the impermeable membranes the traditional paradigm envisions surrounding individual members of our species and suggests that the problem is twofold.
First, a universal interdependence is denied by the traditional paradigm, which is inadequately accounted for by market/property rights, decisionmaking systems, thereby leading to demonstrable inefficiencies. Second, the universal interdependence is highly unequal. Any particular decision and economic activity has greater effects on some than others. For this reason, we defined the valuative concept, self-management, to serve as a more useful guide 48 to this ever-present dilemma than the traditional notion of individual freedom. Whereas the traditional paradigm establishes no clearcut goal in the context of external effects, 49 the valuative imperative of our welfare paradigm is clear: try and approximate, as best possible, a selfmanaged, decision-making system, that is, a system in which all actors have decision-making input to the extent they are affected by the outcome. Whether or not this is easily done is another matter.
Although we have deliberately chosen to conduct our welfare analysis, for the most part, under the assumption of perfect knowledge, we emphasize an important implication should that assumption be relaxed. If everyone knew exactly how development would translate into later fulfillment and what kind of derived characteristics and activities best suited them, there would be nothing beyond self-management to aim for. But while we would expect people possessed of perfect knowledge to frequently "exchange proxies" created by the reality of external effects so as to allow variety in others' development for all the reasons discussed in chapter 5; in the world of doubt, there is even more reason to weight decision-making authority in the direction of those most directly involved. In general, there is good reason to believe the more directly one is involved in an activity, the less imperfect one's knowledge about the consequences is likely to be. So the only dilemma implied by our welfare theory is how to apply the principle of self-management in a context of unequal knowledge about consequences-which amounts to determining the extent to which those most immediately involved should be granted extra authority due to their greater knowledge as well as the greater effects. 50
Finally, we will discover that it is the unevenness of external effects, more than their existence, that makes it difficult to construct desirable economic decision-making mechanisms. Our paradigm emphasizes not only the high likelihood of the prevalence of external effects, but the unevenness of those effects as well. The breadth of the net of effects cast by any particular decision and action can fall any place on a continuum ranging from zero effects on others to equally powerful effects on all. As we will see, it is unfortunate for simple economic decision-making mechanisms (like markets and central planning) that there is every reason to believe that the degree of external effects of most economic decisions falls somewhere between the two extremes.
What is of interest here is that taking an endogenous rather than exogenous view of preferences turns what was already a bad dream for Hunt into a fullblown nightmare. Not only will different preferences over the state of the world create opportunities for self-advancement by choosing so as to maximize others misery, ceteris paribus, the market "solution" to external effects creates incentives for individuals to magnify preference differences over social outcomes to enhance their ability to extract greater reward in the bargaining process" in the future.Each man will soon discover that through contrivance he can impose external diseconomies on other men, knowing that the bargaining within the new market that will be established will surely make him better off. The more significant the social cost imposed upon his neighbor, the greater will be his reward in the bargaining process. 51
If my preferences over the state of the world are the same as my neighbors" there is little opportunity to "contrive" to impose external diseconomies on them (that they will have to bargain with me to cease) without cutting off my own nose to spite my face. In the tradition of using trivial examples, if we both strongly prefer azaleas to forsythia, planting forsythia in the hopes that my neighbor will bribe me to replace them with azaleas is a dubious strategy. Of course, not everyone has the same preferences over the state of the world. What interested Hunt was how inevitable differences between individuals' preferences over the state of the world give rise to the logic of maximizing one's own well-being by maximizing others' misery. If I am truly indifferent between forsythia and azaleas, I can "contrive" to receive a bribe from my forsythia-hating neighbor by threatening to plant forsythias rather than azaleas even though I would be just as happy with azaleas. Any divergence of preferences over flora between me and my neighbor creates an opportunity that an individual "maximizer" is "honor bound" to pursue.
Our new paradigm makes very clear that people's preferences are not exogenous, but amenable to conscious manipulation. As soon as we recognize that preferences are endogenous, and that all choices have preference development as well as preference fulfillment effects, the (socially negative) possibilities multiply. As an individual maximizer I must not only take into account to what extent a choice will satisfy my present desires (traditional theory) and thwart the desires of my neighbor (Hunt's insight), I must also weigh the extent to which it will mold my future preferences to be as contrary to those of my neighbor as possible in order to enhance my opportunity for "conniving" greater bribes out of my neighbor in the future. In other words, even if my neighbor and I were fortunate enough to begin with similar preferences over flora, so that Hunt's invisible foot did not initially conspire to maximize our joint misery, we would each have an incentive to undo our good fate and develop preferences in the most contradictory directions possible!
By viewing preferences as endogenous, our new paradigm deepens another of Hunt's insights. Hunt complained that traditional welfare theory "simply takes the externalities, for which property rights and markets are to be established, as somehow metaphysically given and fixed. In ignoring the relational aspects of social life their theory ignores the fact that individuals can create externalities almost at Will." 52 Hunt meant that individuals with different preferences over the state of the world they must share can create negative externalities for one another (or withold positive externalities) for bargaining purposes "almost at will." In this sense those externalities are not "given and fixed." But in a world of exogenous preferences the opportunity costs (to me) of the negative externalities I can choose to cause others is fixed. However, in a world of endogenous preferences they are not. People can expand on their effective ability to do each other harm by "rationally" developing their preferences in contradictory directions. Of course, in this case the "invisible foot" comes down with an even heavier thud as people who "seek only their private gain, and neither intend to promote the public misery nor know how much they promote it" nonetheless make it increasingly impossible for everyone to be satisfied in the one world in which they must all live.
For theorists such as Richard Musgrave, the key to the public good "problem" was nonexclusion.
In the satisfaction of private wants,
the market functions as an auction place, forcing individual consumers to reveal
their true preferences: Unless I bid, I shall not be able to get what I want.
In the satisfaction of social wants, this reasoning does not hold. The services
supplied are not subject to the exclusion principle. Consumers will be able
to obtain the benefits whether they contribute or not. Since the exclusion principle
cannot be applied, there is no reason why consumers should reveal true
53 preferences.
Let us try to summarize the traditional assessment of the source of the external effects problem in a way that unifies the treatment of "public goods" and "externalities" in market economies whether they occur in "consumption" or "production" and whether their effect is "direct" or "indirect." Whenever a decision is taken in any economy without regard to some of the effects of that decision, we have external effects and the high probability of social inefficiency. In market economies, economic decisions are taken by individual actors who have limited information about the effects their decisions may have on others and certainly no incentive to advance others' interests at their own expense. When this occurs, an obvious incentive exists for those whose interests are being disregarded in the decision-making process to seek to negotiate with the actor whose activity affects them. Not only does this imply an incentive for the actor to contrive to thwart the desires of others in order to maximize the side payment he or she receives; even if we ignore Hunt's "invisible foot," there is little reason to believe the results of negotiations would generate Pareto optimal outcomes. There are at least two different reasons for this pessimism.
1. If many actors are affected, while they may attempt to band together to express their views jointly, the coalition of affected partners will be plagued by the problem of nonexcludability. The coalition cannot effectively challenge individual members' deliberate misrepresentations of the degree to which they are affected in efforts to minimize their individual assessments. 57 For the only way to challenge the veracity of coalition members' suspicious estimate of the degree to which they are affected is to exclude them individually from the benefits of the negotiations. And the only way to do this is to break off negotiations with the actor whose decision generates external effects for the coalition. But this implies that all other members of the coalition will be excluded from any benefits that could have been obtained as well. In more colloquial terms, the coalition cannot effectively challenge members' dissimulation without cutting off its own nose to spite its face. 58
2. Even if the external effect is confined to one other actor in the economy (thereby eliminating any problems of dissimulation within the coalition of affected parties), it is not obvious that Pareto optimal outcomes should be expected from negotiations between two parties. Without the "discipline" of competitive markets to structure their negotiations (more exactly, to do away with negotiations entirely), tactics and strategies of "gamesmanship" are perfectly rational. While some have argued that unrestricted bargaining may result in efficient outcomes, the consensus seems to be that this need not be the case.59
It is worth noting that the problem of market thinness applies to negotiations between coalitions of affected parties and a single actor as well-beyond whatever internal problems the coalition may face in truthfully assessing the extent to which members are differentially affected. It is also worth noting that if such unrestricted bargaining could be relied upon to generate socially efficient outcomes, one of the principal virtues of competitive markets would disappear. For while competitive markets might still be convenient-as a way of saving the trees that would be consumed by greatly lengthened classified sections in daily newspapersthey would no longer be necessary for generating socially efficient outcomes. After all, the chief virtue of competitive markets that led Adam Smith to ascribe to them the properties of an omniscient, beneficent invisible hand was that by eliminating the indeterminacy of unrestricted, individual negotiations they drove negotiations to socially efficient outcomes that might not otherwise have arrived there.
For flood control projects, sanitary campaigns, national defense, or pollution, the greater problem is clearly that the group cannot effectively challenge individuals' underestimation of benefits without being willing to do without the benefit themselves. In other words, the fact that all will end up "consuming" the same amount of a public good (bad) creates an incentive for individuals to underestimate (overestimate) the benefit (cost) they enjoy (suffer) from a public good (bad) in hopes of minimizing (maximizing) the payment (subsidy) they will be assessed as a member of the coalition of affected partners.
For parks, bridges, or lighthouses 60 "exclusion is perfectly possible ... but there would be only one buyer [each individual user] and one seller [the supplier of the public good]" and no reason to believe the two would arrive at a Pareto optimum since there are "no competitive forces to drive the two into a competitive equilibrium." 61
In Meade's classic example of an apple farmer and a beekeeper both problems pertain. The apple farmer could not exclude the beekeeper from benefiting from any expansion of his orchard without giving up the benefits of that expansion himself. But the fact that there were only two individuals negotiating over the sharing of payments also contribute to the unlikelihood that the two would agree to an optimal expansion of the orchard.
In conclusion:
1. There is every reason to expect that markets, even should they be accompanied by voluntary associations, will misallocate resources in the case of public goods and externalities.
2. There is no qualitative difference between public goods and externalities and no reason to treat them differently at a theoretical level. Public goods or bads are simply externalities whose effects are more pervasive than what convention has chosen to call an externality. Externalities are simply public goods or bads whose effects are too circumscribed to merit the label public.
3. The primary intentions of the economic actor whose decision generates external effects for others and the private or public status of that actor, are of no theoretical consequence as far as social efficiency is concerned.
4. The issue is not the characteristics of goods but the characteristics of markets, since it is the market/property right system that establishes who gets to make what decisions.
How is it that "traditional solutions" that involve a government with coercive powers might improve matters? And why have economists concluded that while "traditional solutions" might improve upon market allocation and voluntary association, they are all, to some extent, inevitably flawed? Here we will uncover a very peculiar lack of rigor in the analysis by traditional welfare theorists.
Suppose we substitute a government for the market-property-right system of allocating resources in all cases where external effects exist and instruct the government to do so in a socially efficient manner. The goverment must decide two things: how much of each activity that generates effects external to the market decision-making process should occur, and how much to assess each citizen to defray whatever costs may arise from implementing its first decision. In language more familiar to students of public finance, the government must make (1) a decision about expenditures-how much to spend on the provision of different public goods, and (2) a decision about taxationhow to collect the revenues necessary to cover the expenditures.
The common approach of actual governments is to deter-mine how much of different public goods to provide (and how much to subsidize or tax externalities) through some political process, paying the cost out of general tax revenues. Here we can permit our theoretical government to contemplate what might be more time-consuming and less practical options. On the expenditure side, the government could ask individuals to report the benefits they feel they would receive from different amounts of each public good and provide the amount of each that equalizes the reported marginal social benefit of another unit with the marginal social cost of providing it. 62 On the revenue side, the critical question is whether the government should correlate the size of individuals' tax assessments with the size of the benefits they report they would receive.
What Musgrave and others describe as the "benefit approach" holds that individuals' assessments should correlate with benefits. The motivation for this approach is obvious. Why should an individual benefited little by a package of public goods pay as much as an individual benefited greatly? But Musgrave points out the equally obvious difficulty of implementing a "benefit approach": the government is in no better position to elicit truthful information about the degree to which individuals are benefited than a voluntary coalition of affected agents once citizens become aware that tax assessments are positive functions of individuals' reported benefits. For if tax assessments are positive functions of reported benefits, while citizens could not decline membership in the coalition of affected partners, we cannot avoid the conclusion they would have every incentive to lie about their true preferences. And if people underrepresented their true "marginal willingness to pay," the government would end up doing just what markets in conjunction with voluntary associations do; it would undersupply goods with positive external effects compared with a socially efficient allocation. So we agree with Musgrave and the traditional consensus regarding the apparent dilemma of the "benefit approach."However, when Musgrave and others analyze what they term "the ability to pay approach" and "voting systems together with a compulsory application of the budget plan thus decided upon," they display what we believe is a remarkable lack of rigor. Remarkable because they apparently confuse "Pareto efficient" with "Pareto improvemen" --two fundamental concepts of traditional welfare theory-in a way that implies they allowed their instincts about equity to blind them about issues of efficiency!
According to Musgrave, the "very necessity for compulsory application of a general tax formula means that the, resulting solution will not be optimal". 63 We take this to mean that if tax assessments are based on any formula that ignores differences between individuals' benefits the package of public goods and assessments cannot be Pareto optimal . But this is wrong. If tax assessments are independent of reported benefits, there is every reason to believe that the outcome will be Pareto optimal , for there is no incentive for individuals to misrepresent their true preferences when their tax assessments are independent of their reported marginal willingness to pay. And there is (at least a slight) incentive for them to report truthfully since that is the best way for them to influence the package of public goods they (along with everyone else) will have to live with, no matter how slight that influence may be.
The fact that separating assessments from benefits preserves the efficiency properties of a government allocative procedure is so obvious that it begs an explanation for why so many traditional theorists would have thought otherwise. Such speculation is problematical, but we can only imagine that what Musgrave and others mean is that compulsory application of a general tax formula cannot be a Pareto improvement over the market (mis)allocation. For this is certainly true as well. Undoubtedly, citizens who care little for public goods are made worse off if we move from a market allocation with no public goods and no assessments to a situation with public goods and assessments that are as great for them as for others who are greatly benefited by the public good package. 64
But this is most certainly not the point. It is well known that a move from a particular non-Pareto optimal situation to a particular Pareto optimal situation may not be a Pareto improvement, 65 just as it is well known that the number of Pareto optimal situations is infinite. But this is the case whether or not external effects exist. So, Musgrave appears to be off the mark as well when he concludes:
It is true there is no unique Pareto optimum when the same amount of services must be consumed by all. But if this is all Musgrave means it hardly bears saying because there is no unique Pareto optimum even if all goods are private. If, instead, he means no Pareto optimum exists, he is simply wrong. The fact that we all must live in the same public good world does not imply there are not packages of public goods and individual assessments in which any change would make at least some individual worse off. In other words, public goods have no bearing on the existence of Pareto optima, which was the sum of what Samuelson demonstrated in his famous article. 67The basic problem in the theory of public economy, therefore, is not that social wants are generated in some different and mysterious fashion; rather it is that the same amounts of services are consumed by all, so that (1) true individual preferences for such wants are not revealed at the market, and (2) there is no single solution that is optimal in the Pareto sense. 66
It seems Musgrave simply confused Pareto improvement and Pareto optimum, a confusion that plagued the literature on "compensation effects" as well. Moreover, this confusion is related to the traditional welfare paradigm. If one's paradigm fosters a vision of individual actors immersed in free-market exchange who notice instances of social inefficiency associated with external effects and then try to negotiate a government that would intervene on their behalf-, and if one's paradigm poses the question of whether or not there is any government that imposes a "compulsory general tax formula" they would unanimously agree to; then the answer is no." The answer is "no" because if every citizen must believe that he or she could only become better off, and never worse off, under the rules a government would apply, then presumably some citizen who expected to be little benefited by the public goods that would be provided would veto any government that applied a "compulsory general tax formula."
To state it differently, a path from free-market anarchy to any government that taxes according to a formula not based on individual benefit that consists only of Pareto improvement is highly improbable. Yet any government that makes a reasonable attempt to determine and abide by citizens' preferences regarding public goods and taxes on other than a benefit basis can reasonably expect to achieve Pareto optimality .On the other hand, if citizens immersed in free-market anarchy became aware of the "free-rider problem" and tried to negotiate a government that taxed on a benefit basis, they might well be able to establish a government by unanimous agreement. For at worst the government would replicate their present predicament. And any difference in outcome the government made could be arranged so as to benefit each individual at least as much by the public goods provided as it inconvenienced him or her through tax assessment. But this is not to say such a government would achieve Pareto optimality. Prior to the revolution in incentive-compatible mechanisms it was believed governments that taxed on a benefit basis would inevitably lead people to misrepresent their preferences and therefore be unable to generate Pareto optimal allocations even if some such government were unanimously agreed to by people who sought to improve upon their situation in freemarket anarchy.
The last variety of "traditional" solution was outlined by C. M. Tiebout in his famous 1956 article in which he proposed that citizens could express their preferences regarding different packages of public goods by "voting with their feet." 68 In other words, there would be a competitive market in local finance. Besides the fact that this solution begs the question concerning problems of central as opposed to local finance, which Musgrave among others duly noted, the conditions under which "voting with ones' feet" would generate Pareto optimal outcomes in local finance are highly bizarre. There would have to be an infinity of different local communities within costless walking distance exhibiting an infinite variety of choices on every single issue of public good provision and payment. That is, there would have to be a different community for every possible choice on any single issue and for every other possible set of choices on every other issue of local finance. While we economists have been undaunted by no less far-fetched assumptions in our theorizing, the real problem in this case is the need for a law to prevent members of communities from voting by ballot to change the mix for their own community to keep Tiebout's "local finance market" from collapsing. More than anything else, the necessity of outlawing democratic collective action within communities made Tiebout's "solution" unattractive for many, even for problems of local finance. In retrospect it is apparent that rather than solve the problem of social choice, Tiebout merely suggested using our feet to run around it.
And the truth is this was all that was needed in order for governments to be capable, in theory, of supplying Pareto optimal allocations in the presence of external effects. But it is obvious traditional theorists desired more. Although many traditional theorists may have been confused about why they found taxation unrelated to benefits less than satisfying, it is clear they did find it so. And even though we have demonstrated that the conviction that such governments cannot generate Pareto optimal situations is misfounded, there may be other reasons to object.
Clearly, what traditional theorists wanted was a system of taxation that would permit the government to discover people's true preferences in order to provide public goods in socially efficient ways while at the same time collecting the necessary revenues from people according to the degree to which they were benefited. In other words, they wanted not just an efficient solution to the public goods problem, but an efficient benefits approach. While traditional theorists may have erroneously believed the problem with nonbenefit approaches was lack of efficiency, their real objection was that they did just what they were designed to do: they divorced assessments from benefits. While this is an objection on equity rather than efficiency grounds, it has an obvious appeal nonetheless. Why should someone little benefited pay as much as someone greatly benefited? What many theorists wanted was an efficient benefits approach, and this was what the profession had come to believe was impossible. When Richard Musgrave concluded that neither markets and voluntary associations, nor any alternative procedures were fully adequate to resolving the problem of public goods, this is what he meant. He could not foresee an efficient benefits approach. And he was by no means alone. This was the overwhelming consensus among experts in the theory of public finance prior to the advent of "incentive-compatible mechanisms."
What those who unleashed the revolution in incentive-compatible mechanisms did was conceive efficient "benefits" approaches. They accomplished what their peers had come to believe was impossible. However, the opinions of a majority of traditional theorists notwithstanding, "incentive-compatible mechanisms" were not the first procedures for providing and funding public goods with reasonable claims to efficiency but they were the first that taxed according to benefit and still could lay reasonable claims to efficiency. There is no need to treat the various innovations in greater detail here than in chapter 3 because we do not propose to offer new procedures, nor engage in technical critiques of others' mechanisms. In our opinion the innovations are far more than clever. They recognize that efficient provision of public goods requires truthful reporting of preferences. And they accept the conclusion that if people are placed in a position in which they will be penalized by reporting truthfully, they may lie. But where others concluded this implied an irreconcilable conflict between efficiency and taxation according to benefits, contributors to the incentivecompatible mechanisms literature saw a flaw in what had become an implicit impossibility theorem.
Individuals must be rendered unable to influence their tax assessments by the preferences they report to elicit truthful reporting of preference&. But the same truthful preferences that permit the government to calculate efficient supplies of public goods can be used to adjust individuals' assessments so that those benefited more will pay more and those benefited less will pay less provided the preferences used for adjustment purposes are the preferences of others. ""pivot mechanism"s" use the reported preferences of others to calculate how much others are inconvenienced by the degree to which I influence the provision of public goods, and then charge me for the inconvenience I cause others. "demand-revealing mechanisms" use the reported preferences of others to calculate, the net benefit others will receive from the public goods provided, if they pay their proportional share, and deduct their net benefits from my assessment.
In either case, incentive for me to manipulate the reporting of my preferences is negligible since what I report does not enter into the calculation of my tax assessment. But in both cases, individuals benefit more (or cause others greater inconvenience) will be asses higher taxes, and individuals who benefit less (or cause others lesser inconvenience) will pay less. If pivot mechanisms are used, those cause others greater inconvenience will be charged more. If demand revealing mechanisms are used, those whose net benefit is less will receive larger tax abatements, so to speak, because their abatements are based the net benefits of others who benefited more. With the benefit of hindsight the idea appears perfectly simple. But this could be said about brilliant ideas. The obstacle to correlating taxes with benefits with inducing an incentive to lie had been removed; compensations must be based on the effects others' report.
The clarification of what is accomplished by incentive-compatible mechanisms and how they achieve it has little to do with our new welfare paradigm and theory. But the new welfare paradigm contributes to the interpretation of incentive-compatible mechanisms a better understanding of the importance of the issue. What is at stake is whether or not individuals with "minority" preferences for public goods should be compensated or penalized in some way. The issue is important because:1. If such compensations should be deemed warranted, the calculations necessary for carrying them out efficiently must be applied to the entire allocative system
2. Beyond its distributive effect, compensation, penalization, or tax "neutrality" affects the degree of variety or uniformity that will evolve in people's preferences with respect to external effects
Since the reader may find these interpretations novel, we elaborate.
Everybody "consumes" the same public goods, but not everyone is equally benefited by doing so. People's preferences can vary with regard to how much public goods they would like provided and what they would like those public goods to be. An individual who enjoys the good fortune of having preferences that coincide closely to the package finally provided will enjoy greater benefits than individuals who would have preferred more (or less) and different kinds of public goods. Since what will be provided corresponds to average preferences, those whose preferences are closer to the average will benefit more than those whose preferences regarding public goods are farther from the norm. 70
In one view, the question is whether individuals who are less benefited by the package of public goods all receive should be compensated for their "misfortune." In other words, should "odd balls" receive positive compensations? Should those whose preferences deviate farther from the social norm be compensated for the fact that they got less of what they wanted than those whose preferences were closer to average? Should the "preference minority" be compensated for the fact that their "candidates" lost in the public goods elections? 71
At one level this is an equity issue-having verified that efficiency could be preserved in either case. If we implement proportional taxation our implicit answer is "no." If our answer is "yes" we should tax on a benefit basis, which could be accomplished without loss of efficiency by using a "demandrevealing mechanism," such as Groves and Ledyard developed.
Similarly, everybody influences the package of public goods provided ' 72 but the extent to which everyone's influence inconveniences others, by skewing the package from what it would have been without their influence, is not the same. Those whose preferences are farther from the average will exercise a greater influence on the outcome than those whose preferences more closely approximate the norm. Moreover, if preferences are distributed normally, outliers' influence will inconvenience more others, causing a greater inconvenience. Seen in this light the question is whether individuals who inconvenience others to a greater extent should be penalized for doing so. Should "odd balls" be penalized? Should those whose preferences deviate farther from the social norm pay for the fact that they imposed their will on others to a greater extent than others imposed their will on them with regard to public goods? Should the "preference minority" be penalized for the fact that their "vote" counted for more in the public goods elections despite the fact that their "candidates" lost by larger margins? Again, at one level this is simply an equity issue. The implicit answer of proportional taxation is "no," while the implicit answer of a ""pivot mechanism"" is "yes." But regardless of which view one takes and how one decides to answer the question in either case-that is, regardless of one's conception of what constitutes an "equitable" procedure for financing the provision of public goods-our paradigm implies the issue is of much greater importance than the traditional paradigm would lead one to suspect. For our paradigm emphasizes the reasons for believing that external effects are prevalent rather than exceptional. If one believes that compensations or penalties are warranted on equity grounds with regard to public goods, then our paradigm suggests that an appropriate "demand-revealing" or "pivot" mechanism must be implemented as the general mechanism for resource allocation. Since not even their most fervent supporters claim demandrevealing mechanisms are highly practical and easy to implement, this is a highly significant conclusion.
Our paradigm also suggests important consequences beyond equity are involved in the choice of a principle of taxation. If people's preferences are to some extent endogenous, and any advantage can be obtained by molding them in one direction rather than another, our welfare theory highlights why it is rational for people to do so. What preferences over "public good commodity space" would it benefit me to have? From the perspective of maximizing fulfillment from the package of public goods provided, it is advantageous to have preferences as close as possible to the social average. Here the incentive is always toward a greater conformity of preferences regarding public goods. From the point of view of minimizing tax assessments, it is advantageous to have whatever kind of preferences are compensated and not penalized, but that will depend on what principle of taxation is adopted. A "pivot mechanism" penalizes minority preferences, pushing preference development toward greater conformity. Any mechanism unrelated to benefit or inconvenience to others provides no incentive to develop either conformist or deviant preferences for public goods. A demand-revealing mechanism rewards minority preferences and establishes an incentive for individuals to develop diverse preferences.
But presumably it is the overall package of the public good "consumption" effect and the "payment" effect that is relevant. In this light, taxation unrelated to benefit or inconvenience to others exhibits a single conforming influence on preference development. A pivot mechanism contains a double conforming influence. And a demand-revealing mechanism "compensates" the conforming influence of the "consumption" effect with the diversifying influence of the "payment" effect. If compensation were complete, there would be no net influence toward greater diversity or conformity. Presumably overcompensation could be arranged to provide a net diversifying impact.
In any case, the preference development effects of taxing principles and the implication for greater conformity or variety of human preferences with regard to public goods appear to be as interesting as the implications for equity. And just as one might conclude there is no clear-cut case for one version of equity over another regarding payment for public goods, the case for greater variety or conformity of preferences for public goods appears ambiguous as well. For all the reasons we examined in chapter 6 -uncertainty, vicarious enjoyment, etc.-greater variety is better, ceteris paribus. But in the case of public goods greater conformity of preferences has an advantage in efficiency. Since we all have to consume the same public goods, the advantage in having similar preferences is that more people get closer to what they want. The greater the variety of preferences for public goods, the more disappointed more people must be. We hasten to point out that this "tradeoff' pertains only in the case of public goods. To the extent that goods are private, or can be suitably handled via local finance, there is no disadvantage to greater diversity of preferences. But for truly public goods greater variety of preferences implies a loss of well-being. All our new welfare paradigm can do is make us better aware of this trade-off. It does not imply any particular solution.
So, where does all this leave us regarding the efficiency properties of competitive markets?-the question we began with. At one level we only wish to borrow a conclusion that, until recently, no serious economic theorist would have denied: market allocations are demonstrably inefficient for public goods and externalities. While this opinion is no longer unanimous, we attribute recent dissenting opinions to the overexuberant atmosphere of the free-market jubilee. We are fully confident that a return to normalcy will restore the time-honored consensus among reputable economists that voluntary associations appended to markets must still fail to yield efficient allocations when external effects are present.
At a second level, we wish to point out that this very traditional conclusion is of far greater consequence than usually supposed. While the traditional paradigm confines the problem to a small number of exceptional "goods," our new paradigm projects the problem to be the general case. In general, human activities have external effects on others' possibilities-to a greater or lesser degree. The vision of economies as millions of isolated Robinson Crusoes connected only through the material goods they interchange is precisely the view we have criticized as grossly misrepresenting the human condition. While in market economies maybe only material "outputs" are exchanged, in fact what is shared is the physical and human environment in which we all must live, which is the joint outcome of all economic activities engaged in. Instead of seeing only the physical objects of market exchange we should see them as proxies for all the human activities that stand behind them. And while the effects of any particular activity may be much greater for some than others, the effects seldom are entirely confined to a single economic actor.
Moreover, since resource allocation in the context of external effects is the general resource allocation problem rather than a special case, the subject matter of the recent literature in incentive-compatible mechanisms becomes far more important. Some alternative to the market mechanism is required, in general, if there is to be any hope of allocative efficiency. The most important point is that this already takes us beyond the subject matter of this chapter. Market allocations will be, in general, inefficient. The efficiency, equity, and preference developmental properties of various alternative allocative procedures are another set of questions made interesting precisely because of markets' failures.
Finally, we have sought to throw some light on what is at stake in considering different proposals. Efficiency in the traditional sense becomes a rather trivial concern that can be satisfied by perfectly common approaches to financing public goods. And while one's notion of "equity" is of great importance in defining desirable allocative mechanisms, the effects of different procedures on preference development and the relative advantages and disadvantages of variety and conformity may be of equal or greater interest. We will comment briefly on criteria for an alternative desirable allocative mechanism in our conclusion, but return now to the allocative mechanism under consideration in this chapter, markets.
Here, we formalize the only conclusion we require from the previous section and reformulate it as a theorem in a slightly unusual form that will be useful in the analysis that follows.
Market institutions have an inherent tendency to provide a relative overabundance of private goods compared to public goods. More exactly, markets can be expected to allocate too much of society's scarce productive resources to the production of private goods, goods with negative external effects, and public "bads," and too little of society's resources to the production of public goods and goods with positive external effects, resulting in a relative oversupply of the former kinds of goods and relative undersupply of the latter.
The usual way of seeing this is to note that social efficiency requires that output of every good be such that the marginal social cost of production be equal to its marginal social benefit. In private goods the marginal social benefit is simply the additional benefit enjoyed by a single purchaser. But for public goods the marginal social benefit incorporates the benefits enjoyed by everyone when the supply is expanded by one unit. Because free markets, even in conjunction with voluntary associations, are unable to overcome the free-rider problem, the effective demand expressed for public goods underestimates the true marginal social benefits.
The inefficiency can be seen from the perspective of the aggregate market as follows. The actual demand expressed in the marketplace for a public good, a, lies everywhere below the marginal-social benefit curve, yielding the wellknown result that the actual amount of the public good, aA that would be supplied under a market allocative mechanism would be less than the socially optimal amount, a0. From the suppliers' perspective, the actual market price, pA, they will receive is less than the price that would elicit an optimal supply, no, so they supply less than the socially optimal amount of the public good as depicted in figure 7.4.
An individual, i, should only be charged an amount equal to the net marginal social cost of supplying him or her with another unit of the public good, a. That is, individual consumers should be charged all additional costs to society of supplying another unit minus the marginal social benefits to all others. To charge individuals otherwise necessarily generates social inefficiency. To see this consider that
is the usual condition for social
efficiency for public goods. 73
As long as the right side is
at least as large as the left side in this expression, it is socially inefficient
not to supply the marginal unit. This can be rewritten as
where we separate the marginal benefits of all others from the marginal benefit to individual i. Suppose i is charged an amount in excess of
Then she or he may not purchase an additional unit of the good, even though the marginal social benefits of another unit are as great as the marginal social costs of providing it. On the other hand, if i is charged less than
then she or he may purchase units even when the marginal social benefits are smaller than the marginal social costs. This implies that the socially optimal price to charge individual i for another unit of a, pi(a)O, is exactly
Charging i
either more or less than pi(a)O
can lead to socially inefficient outcomes.
But since suppliers of goods with positive external effects (or public good
a in our example)
cannot collect all of
from others-for the reasons explained above-they make each individual, i,
pay some actual supply price, pi(a)A,
in excess of the optimal individual supply price, pi(a)O
in their efforts to cover costs. In other words, the inability to force others
to pay for the benefits they actually obtain means private market suppliers
must charge purchasers prices that are higher than would be socially optimal
as pictured in figure 7.5.
In figure 7.5 we formulate the situation for period (t+k) to apply our theorems from chapter 6 directly. As can be seen, the actual supply price to individualsthe price that will be charged buyers under market allocations of goods with positive external effects-yields individual demands for the good, ai(t+k)A, that are less than the optimal individual demands would be, ai(t+k)O. If we compare this situation with the situation with a good displaying no (or less) positive external effects we can formulate the following useful theorem:
In chapter 5 we argued preferences are endogenous to economic systems, and in chapter 6 we explained that people can be expected to orient their future needs toward goods and activities that will be relatively easy to obtain and away from goods and activities that will be hard to obtain. We proved Theorem 6.6, establishing that if there is a bias in the conditions of supply of a good or activity, the resulting degree of nonoptimality in the economy will be greater than indicated by a traditional welfare theory that treats preferences as exogenous, and that the divergence will "snowball," or grow, over time. In Theorem 6.7 we demonstrated that "rational" people can be expected to influence their own development in a direction that mirrors any bias in the conditions of supply, or that individual rationality dictates a kind of "self-warping" in response to biases in the functioning of economic systems. And in Theorem 6.8 we demonstrated that any such biases will tend to be disguised to the economy's inhabitants to the extent that their "rational self-warping" is unconscious or forgotten after the fact. We have just reformulated a well-known theorem regarding public goods and externalities as our Theorem 7.1 that markets generate a bias in conditions of supply for goods with external effects.
But the sum of these theorems implies that if people have reason to believe that markets will be used as allocative institutions in the future, they also have reason to expect a bias in the conditions of future supply of goods with external effects. Combining yields the following result:
The proof of Theorem 7.2 is immediate, applying Theorem 6.6 to Theorem 7.1 written for period (t+k). What the theory and theorems of this and chapter 6 permit us to see that was not apparent to traditional theorists is not only the prevalence of the problem of effects external to market decision making, but the full consequences of market failure. Not only are misallocations due to external effects pervasive in market economies, the severity of the misallocations worsens over time as people pursue the "individually rational" strategy of molding their characteristics and preferences to better conform to institutional biases, thereby producing the "socially irrational" outcome of allocations that diverge even farther from efficient allocations than would have occurred had people not adjusted. 74
Beyond theorems concerning social efficiency, we have tried to pinpoint the mechanism whereby people's perceived needs and desires come to correspond with the individualizing characteristics of market institutions. In our terms, we clarified the mechanism of "rational" self-manipulation or warping of personality structures, skill development, and consciousness that reproduce the core characteristics of market institutions in the human center of people's personalities as they evolve in market economies. While the effects of institutions on the characteristics of those whose activities they organize has been apparent to many, there has been a lack of clarity about the mechanisms through which those effects are exerted. Moreover, it has not been apparent how one could draw conclusions about the welfare consequences of such effects since the institutional influence affected people's preference structures as well as their characteristics. We hope our new welfare theory clarifies both how the characteristics of institutions affect the characteristics of inhabitants and why this need not subvert prospects for drawing welfare conclusions.
We hope our first application of these new tools contributes to a fuller evaluation of markets-an evaluation that highlights liabilities of markets under the most generous assumptions of how they may function. And we hope our analysis also clarifies what has frequently been only a vague "intuition" of those critical of markets. More than a few social critics have voiced the sentiment that markets have an undesirable effect on the people who use them-that markets tend to make us into people we would rather not become. Critics have sometimes implied that besides whatever else markets do or do not do, they have a destructive effect on the quality of social life. Traditional welfare theory ridicules such sentiments as unscientific -the sentiments of fuzzy-minded romantics-that does not withstand the rigors of clear-thinking analysis. Our theory suggests that such intuitions about markets are perfectly well founded, and that the criticisms of "romantic visionaries" of certain effects of modern economic institutions can be expressed and supported by rigorous "scientific" analysis.This chapter has attempted to identify the cybernetic, incentive, and allocative characteristics of market institutions. We claim to have uncovered a pervasive bias in market institutions against discovering, expressing, and developing needs that require social rather than individual activity for their fulfillment. Markets do not provide concrete information about how my decisions affect the life prospects of others. They do not even provide accurate summaries of the social benefits and costs associated with what I decide to do. Markets establish an incentive structure that rewards competitive behavior and penalizes cooperative behavior. And actual market allocations undersupply social goods and activities and oversupply individual goods and activities. They thereby establish the kind of individually rational incentives described in Theorem 7.2 to wean myself of i needs that require socially coordinated intercourse for their fulfillment and accentuate needs that can be met through individualized activities.
In any case, we have demonstrated the logic that would lead "rational" people to mold their own characteristics to better conform to market biases. And we have demonstrated that this "individually rational process" is "socially irrational." By doing the best they can within the limiting environment of market institutions, individuals will adjust in ways that imply even greater losses of potential fulfillment than if they did not adjust to markets' limitations. As time goes on, the preferences that are "individually rational" for people to develop combine with the biases inherent in market allocations to yield allocational outcomes increasingly farther from those that would have maximized fulfillment. In the end, the worst fears of "romantic" critics of the "socially alienating" affects of markets prove more to the point than the endless volumes of "scientific equations" produced by academic economists claiming to "prove" that markets are ideal allocative institutions.