Property, Power, and Public Choice



In this chapter some of the concepts in Part II will be combined as needed and further developed in the context of several broad areas of public choice including the form of business and political organization. Thus will be tested the utility of the concepts outlined above in producing insights and verifiable hypotheses as to the impact of alternative property institutions.


It is important to distinguish use rights and exchange rights. In a self-sufficient economy, use rights are all that matters. Other than physical trespass, there is little that one person can do to another. Use rights remain the prime ingredient of some rather complex nonmarket organized economies.1 In feudal Europe, people had a traditional right to utilize certain land for cultivation and had common access to certain pasture land. If a person could not make use of his entitlement, the land was simply absorbed by others. There was no way for a person to transfer rights explicitly to some specific other person. If one did not make use of it directly, there was no way to realize any personal value from the resource. One could not even make a gift of it to another specific person. One had use rights but not the right to exchange or grant.

In time, exchange rights were added and markets developed for consumer goods, while land often remained subject to use rights. The transformation from use rights to markets in most resources was one of the most wrenching and misery-producing periods in history. Many words have been written trying to prove that this was all worth it and that the resulting market system went on to produce the best of all possible worlds (North and Thomas 1973). Many words have also been written criticizing this development and holding it responsible for many of our current human ills (Polanyi 1944). A discussion of any possible normative conclusions will be left to Chapter 11. Here some of the substantive performance implications of the right to transfer will be explored.

The right to transfer means that it becomes possible to gain assets by withholding what you own from others even when you cannot make use of it yourself. This is a proprietary scarcity. The right to withhold from others what they need but do not own is the potential substance for market income.2 Coupled with the right to transfer, it is the engine by which it is possible for some people to accumulate great wealth and in effect have other people work for them. In a non-transfer economy, some people are more energetic and skilled than others and many have access to use rights in natural resources that are more productive than the resources of others. Differential access to resources means that some people will have more wealth than others, but the differences are limited (unless concentrated administrative control of labor is practiced). In an exchange economy, the differences can be and are extreme.

In a nontransfer economy with given use rights, the only cleverness that counts is against nature. But, in an exchange economy, cleverness against other people redistributes assets even where total wealth remains unchanged. If the other party can be fooled or kept in ignorance, you may make a more advantageous bargain. In a modern market economy, a good deal of talent and energy goes into manipulation of paper asset values and tax avoidance that has little to do with production of wealth. Also a lot of effort goes into creating scarcity to increase price. In a stable non-transfer economy, no one can increase one's assets without increasing total wealth (except of course by plunder).


A much advertised advantage of the market is that both parties' satisfaction can be increased by exchanging what you have for that of another, which you prefer. Stated somewhat more abstractly, exchange allows the internalization of certain external effects. If A exercises her use rights in X in ignorance of the fact that B would be willing to give A a good that A would prefer in exchange for letting B utilize the good X, both have missed an opportunity to be better off.

The widely used concept of internalization can be confusing since an external effect that is internalized does not disappear. Recall that externalities are an inevitable aspect of interdependence and can be shifted and transformed but never eliminated (Chapter 1). By definition, external effects are internalized when trade is allowed. But the interdependence-externality does not go away. The nonowner substitutes the impact of the old externality (unavailable opportunity) for the cost of giving up the item exchanged. Market trade may enable A's interests translated into a bid to become an opportunity cost for B.3 This is the definition of internalization. If A's bid is rejected, because of inadequate purchasing power or high transaction costs, the former externality continues although it has been internalized. As long as A's pitiful bid is received, it has been internalized. In either case, market internalization does not cause interdependence externalities to go away but can change their form. When B's use of resources causes A to have reduced opportunities, rights allowing internalization can ameliorate this reduction.

In tracing the effects of alternative systems of rules, care must be taken to avoid an extreme partial equilibrium analysis. (What follows is still partial equilibrium in the sense that feedback of institutional performance on attitudes and rule making is acknowledged here but is not the focus in the time frame chosen.) While it is true that with a even distribution of property rights between two parties, the right to trade can increase the satisfaction of both, we must inquire of the effects on third parties. For example, in most of the states of the eastern United States, utilization of water is controlled by use rights. These so-called riparian rights are not usable and not salable separate from the riparian lands. Thus, even though a riparian has no use for the water, he cannot sell it to a non-riparian who might put it to a valuable use. (In some societies, where labor is not exchangeable, the owner of land and water use rights could not obtain additional output by letting wage earners utilize a portion of the resources.) With a partial analysis, this situation appears inefficient since resources cannot move to their most valuable use. The potential superior opportunity of a non-riparian is not internalized.

But we must examine how third parties are affected by others' potential choices. It can be observed that every time the riparian doctrine is challenged, it is defended by fishermen. Increased sales and diversions of water would affect the interests of fishermen. The fact that sale is prohibited (or that there are high transaction costs in obtaining consent from all of the owners in common on the stream) is to the benefit of fishermen. Water cannot now be made salable without affecting their interests. Prohibition of sale can prevent A and B from increasing their satisfaction and apparent value of output from the water. But the enhancement of their satisfaction is at the expense of C, who is not represented in that market. The use right of fisherman C is protected by the fact that A only has an incompatible use right that cannot be sold. One person's limitation (rights attenuation) is another person's bread. The right to transfer is the right to create costs for the parties not represented in the market (that is, it is the right to create costs for others, which are protected in no other way than by prevention of marketability). Where the uses of people are interdependent, the right to sell or otherwise transfer has external effects.

Any society must have some system for disaggregating scarcity. In a nontransfer economy, scarce resources are rationed or allocated when use rights are assigned or claimed by custom. In this context, changed users (and sometimes uses) are difficult to create. For example, in certain communal tribal land tenures in Africa, each family has a traditional use right (Parsons 1965, pt. 8). However, there is no provision by which the tribe could alienate part of the land to an outsider to make new uses. When the University of Nigeria was being built, it had difficulty obtaining land in the traditional tribal area. No one (either ordinary individual or chief) had the power to transfer land to someone else. This has important consequences. On the one hand, no tribal member can be dispossessed. There is never a destitute, landless poor. No matter how scarce the available resources, all have some share in its output. There is no need for a separate governmental welfare program or even private grants. On the other hand, quick reallocations of resources are difficult.

Use rights that cannot be sold provide great security to their holders if the rights conform to customary behavior. Where group norms change slowly, however, they provide little individual liberty to innovate. In addition, the liberty to try new things also creates exposure and insecurity. A mortgage on alienable land means the individual risks loss of the land. Where bargaining power and information availability are unequal, the move from use to exchange rights often has resulted in a larger class of landless people (Parsons 1974). Where traditional use rights still exist in agriculture, it is a great challenge to institutional design to create a system of rules that allows for mobility, creativity, and innovation and that draws forth extra energy without creating a very unequal income distribution.

The use rights of American Indians to the resources of their reservations is a good example. There is no way for a tribal member to leave the reservation and realize any benefit of past labor or birthright. This can be a significant barrier to labor mobility. But, among Indians, it means that it is harder for income distribution to be uneven, and no one can be completely denied access to resources or be subject to a bad bargain.

There are various other examples of nontransferable rights. For instance, some pension plans are only available if the worker continues to work for the company; again mobility is discouraged. In summary, internalization (addition of exchange rights to use rights) affects different parties in different ways.


There is one function that only the right to exchange seems capable of serving. This is the process of capitalization, which is a major part of what is meant by the capitalist system of organization. Capitalization is the process by which future values are converted into present values. Or, more concretely, it is the process by which A can enjoy present consumption, which otherwise would have been used by B, who in turn uses the future output created by A. Investment is possible in all economic systems whereby an action taken in the present has the possibility of increasing the flow of future output. In a system of nontransferable use rights, one must stick around if one is going to capture any of this future. But, in a system of exchange rights, one can cash in one's claim, extract a present value of the future output, and leave for other activities. This basic situation can be explored in many contexts,4 and only a few will be noted here.

Suppose that a person develops a new product or a cost-saving technology. Choice of rights affects the reward to the innovator. In a competitive market, the innovator earns a short-term profit until other firms enter the market and copy the product or process. While the short-run disequilibrium in supply and demand may produce sufficient incentive to innovate, these profits cannot be a source of large fortunes and great inequality in income. In other words, market competition distributes the benefits of innovation widely to consumers over time.

But where firms have market power and can prevent copying of new products and processes, the capital market plays a unique role in creating great rewards to innovators, instantaneous fortunes, and great income inequality. In the U. S. economy, large differences in rates of return persist over time (Thurow 1975, p. 143). If the right to a future above-average yearly rate of return can be sold in a stock market, the owner capitalizes the future flow into an instantaneous fortune. Buyers of stock seeking above-average return bid up the price of the stock of firms with high returns until the returns on the stock investment are again at the average. The buyer at this point earns only the average rate of return, but the purchasers have made the original innovator or stockholder wealthy without having to wait for the future output. Some buyers of stock try to anticipate which stock will earn higher-than-average return on present prices. Lester Thurow (1975, pp. 142-54) argues that in fact great success is determined by luck (he calls it a random walk) rather than superior intelligence. He notes that professional financial managers do not do better than the stock market average over time even though they can invest in market information.

To summarize, when noncompetitive industry is tied to a competitive stock market, it can result in greater income inequalities than are inherent in administrative or status-grant transactions. The rights of bargained exchange mean that successful innovators with market power and lucky betters in the casino that is called the stock market will achieve fortunes many times greater than that available to patient savings and reinvestment following the principles of compound interest.

This inequality-producing aspect of the stock market is often justified by arguing that the prospective high payoff to some in the stock market is functional in providing a source of new funds for creation of new capital goods. As people seek higher-than-average returns, they make available new capital to innovators. In fact the issuance of new stock is seldom a major source of new investment. In 1972, of all private industrial and commercial investment (excluding residential structures) in the United States, 99 percent was generated by retained earnings and depreciation allowances (U. S. Department of Commerce 1973, pp. 10 and 38 and Berle 1959, Chapter 1). While this percentage varies with the business cycle, it suggests a plausible hypothesis that a prime effect of the stock market is to enhance inequality and not to obtain capital for new investment. Some insight into how rights affect savings and investment can be gained by examining some of the socialist economies of eastern Europe.

In Yugoslavia, workers in an industrial plant help decide how much of the surplus of their work will be paid in wages and how much will be reinvested in the firm. If money is reinvested, it should increase future surpluses and wages. But, if the worker should move to another firm, he would not be around to enjoy the future product that his reduction in current consumption helped create. The worker cannot sell his share to future output and take the money and invest in still another enterprise where he might expect future returns to be even higher. Some writers criticize this property system and argue that investment and production will be lower under such a system than it otherwise would be under the right to sell (Furubotn 1971).

But back to the main question of whose interests are to count. Some individual Yugoslav worker may think that there is money to be made in certain consumer goods, say, a pizza parlor. But others prefer to expand heavy industry and education. Productivity is always ambiguous. The real question is who gets to decide what outputs are the most important. There is no doubt that individual saleability is related to mobility, but anyone who has seen a town that a major industry has moved from knows that certain costs were ignored by market choosers as they sought the least-cost (to them) location. Some person who has better opportunities elsewhere wants to sell out and take her capital elsewhere, but this move may create costs for the person who has many immobile assets in the current location and few other opportunities. One person's right to sell can create costs for others. Whose interests are labeled as the most productive depends on your point of view.

Even setting aside the issue of the composition of output, the empirical evidence is not supportive of the theoretical prediction that the lack of rights to sell will lead to low investment decisions by workers' collectives. In fact, Yugoslav investment has been remarkably high. Some say this is due to higher-level regulation (administrative transactions), but not all of it can be so explained. Perhaps there are ideological substitutes for the individualized strivings that we have come to expect in the United States. It should be kept in mind that if every firm is making substantial reinvestments, a worker can move to another industry and not lose any benefits. Behavior under equality is different than under expectations of inequality. In a highly mobile society, people may hesitate to tax themselves for public improvements that they may not be around to enjoy. They have in effect a use right but cannot sell their claim to a new immigrant when they leave. However, if all or most communities are making similar investments, it does not matter. A person gets to use resources in the new community similar to the ones he paid for in the old community. Investment behavior is affected by the image people have of the broader system of which they are a part.

Some writers say that public ownership arises when some people do not want to bear the costs of their own actions as individuals (Alchian 1965). Everyone gets the same share of output regardless of his own personal failures. Some analysts predict that public ownership leads to inefficient, low output. People cannot capture or bear the unique outputs they create. This argument was touched on earlier in Chapter 6 in the discussion on information costs related to the form of business enterprise. The same kind of question arises in public management.

It can be observed that some managers of government agencies do not aggressively pursue cost-saving opportunities. They appear to try to maximize their budget total, which gives them larger salaries and prestige. Any cost saving benefits the general taxpayer, but not directly the income of the manager. From the manager's point of view, the savings are common property, and no one is excluded. In a private, market-oriented firm, if the manager could cut costs, future earnings would increase and the present capitalized value of the firm would increase. This situation attracts stockholders and allows the manager to earn a higher salary or realize a gain on her stock if she is also an owner.

This line of reasoning has led several authors to hypothesize that publicly owned electric utilities or airlines will have higher operating costs than private firms. Others observe that private corporations maximize sales and firm growth rather than profits, and in this sense they are similar to public firms. The above phenomenon has led William Niskanen (1971, chap. 18) to advocate that the bureaucratic director of a public agency should be awarded a share of any cost savings she creates. (It is assumed that the manager has the power to get her orders carried out and that employees have no bargaining power. In reality the problem is more complex than just changing incentives of top management.) This share would be paid after leaving office. The problem with this idea is the difficulty of specifying product quality and quantity. The manager may be tempted to cut costs by cutting product quality (something that also happens in private business and even in socialist firms when they try to meet production quotas). Niskanen's plan requires a clear definition of what it is that the public agency is to produce. Of course, clarity is essential in any calculation of efficiency. Efficiency is meaningless unless we agree on the definition of relevant output and input--that is, whose tastes count. (No distinction has been made thus far as to the kind of good to which an exchange right applies. In some contexts, the right to one resource is a vehicle for access to other rights. For example, the right to land may be the point of access to political power. Thus, the right to exchange land may be the source of still further enlargement of political power while the right to exchange other goods has no such implication. See Penn 1969.)

Lester Thurow (1985, p. 162) has also suggested that private managers' retirement benefits should depend on the success of the firm 10 years after retirement. This would lengthen the planning horizon of the firm and de-emphasize short run profit maximization.

Capital, Credit and Banking

Every economy has some system of rights by which someone is put in charge of organizing new productive opportunities--that is, put to work new potential and unused resources. In so-called capitalist countries, this is done by bankers. They are given the right to create money and to assign it to prospective entrepreneurs for a fee. The process allocates rights to command resources when a bank writes a number of dollars after a person's name. It is called a loan, but it is simply a creation of a property right in new resources. Typically, a person must already have accumulated significant resources in order to be eligible for a loan. The quality of the proposed use of new resources is not enough. In the words of David Bazelon, "Credit is status capitalized" (1959, p. 87). Note that banks create money without prior saving by individuals, though the new money if invested will result in savings.

Exchange (market) transactions are usually regarded as the distinguishing feature of capitalism, but with respect to creation of rights, the banking system essentially uses an administrative transaction. The banks cannot themselves use the new resources, so they are not exchanging one good for another. They are like a legislature distributing property rights to constituents. The fact that a fee (interest) is collected does not turn an administrative relationship between superior and inferior into one of bargained exchange between legal equals.

As an institution, capital is a system of rights allocating power. In pre-capitalist systems, investment required either individual decisions to save, trading profits, collection of surpluses by those with power, usually military. But, with fractional reserve banking, entrepreneurs can be given credit to command a share of current output to invest. Capital in the form of new debt money causes the economy to save. Part of the saving decision has become centralized in the hands of banks controlled by central banks independent of individual decisions. Surplus can be extracted involuntarily without the need for physical power under pre-capitalist tribute systems. Robert Heilbroner (1985, p. 40) emphasizes that capital is a social relationship of domination. The institutional means of this domination is credit created by banks. The future product is capitalized when an entrepreneur is given new credit created by a bank. This money commands a portion of current resources to create new commodities in the future which are re-transformed into money again. This is part of the famed M-C-M sequence of Marx. In socialist, centrally planned countries, the planning body allocates new opportunities and investment funds to government-owned industries according to a political decision. The criteria for who gets the funds are different than in capitalist banks, but the essential process of new rights creation is the same.

Heilbroner (pp. 41-2) emphasizes that an essential part of capitalism was the process that resulted in peasants and workers being excluded from ownership of their tools and their own outputs. This depending on the capitalist arises in part because of access to production credit.

There is a great deal of discussion in economics literature on monetary policy (interest rates and supply of money) but little on just who should own the rights to new potential resources. At present, the ownership of leveraged debt belongs to corporate stockholders although the collateral for the debt is the whole going concern. When the banking system is seen in property rights terms, new policy variables emerge for placing rights to new resources in the hands of various groups. This perspective has quite different implications from those of the usual redistributive welfare programs that attempt to transfer earnings after the initial rights to new opportunities have been vested. Also this view of credit creation rights raises the question of why government pays interest on public debt created at times of unemployment when private borrowing is insufficient. For elaboration see Schmid (1982 and 1984).



There is a set of property rights issues that can best be understood as boundary questions. It involves how a given person in her capacity as consumer or voter is grouped with other people. Everyone is aware of some of the issues in drawing the boundaries of political jurisdiction, whether that of nation, state, or some subdivision thereof. A similar type of issue is involved in grouping consumers, though the groupings are seldom only geographical. The question is one of defining the extent of the system over which performance is evaluated. Rights that determine subsystem identification are an inescapable choice in either the private or public sector.

Product Definition

There are a number of situations where the scale of product Y has some influence on the profitability of making product X. Such situations were discussed above in the context of obtaining demand information in the face of large cost discontinuities. The situation to be analyzed here is somewhat different. The amount of product X influences the profitability of Y because they are segments of a larger product sought by some consumers. An example is different links in a transport network such as main lines and feeder lines. A problem of interdependence arises when all people do not have the same interest in each segment of the product.

Most transportation systems are composed of segments between points. Some people ride long distance over many segments, and others only use a portion. In the case of airlines in the United States, the segments may be provided by different firms. The existence of a feeder line may affect the number of customers and profitability of the main-line firm. This interdependence can be accounted for in several ways.

The main-line firm may offer the feeder firm a payment to help it stay in business. The external effect is internalized by market bargaining. Another possibility is expansion of the size of the firm to include both segments of its interconnected route. It is possible that the integrated firm will suffer losses on certain portions of its system, but the existence of the feeder routes may still contribute to overall profitability. If contractual costs or internal information costs are not too high, selection of either alternative will provide the entire route.

But there is an interdependence between customers that raises deeper questions than the existence of the total transport system or profit maximization. Some people ride the whole route and will pay for its total cost regardless of how the portions are priced. But some riders may only ride the feeder route and some only the main route. The feeder-route riders get a reduced-cost ride compared to the margnal cost of that portion of the route. The integrated firm may charge a flat price per mile regardless of the portion of the route used. Customers are treated as one large group to the advantage of those riding the higher-cost feeder routes exclusively. The firm in reaching its profit-maximizing pricing scheme does not care about someone getting a reduced-cost ride. It might not profit the firm to try to charge more for the feeder route and lose customers on the whole system. The possibility of contracting or creating a single firm may make possible the maintenance of the whole system, but it does not speak to the question of different groups' different interests in portions of the system. Sometimes economists urge that, with nationalized or socialist firms, each separate product "stand on its own feet" (Munby 1962, p. 52).5 Thus, each portion must cover its own marginal cost for that portion, and the profits of one portion are not allowed to cover losses on another portion, which must, instead, be dropped. This situation greatly harms some groups of riders and reduces profits to the socialist firm as a whole. As already noted, private firms would try to avoid this "mistake." The word mistake is put in quotes because, as often happens, one person's mistake is another person's benefit. The person who rides only the main line would have his fare reduced if he did not have to help pay for the feeder line. Something that is a mistake from the view of total profitability may be to some group's advantage. Again, the public must decide whose tastes count, as firms, whether private or socialist, decide on service boundaries and aggregating of groups of users. The public has to decide whether to allow the private firm to integrate and maximize profit for the system, to regulate competitors that might like to provide main-line service only, or, if the service is nationalized, whether to make each segment pay its own way or not. (The latter issue came up during establishment of the ConRail reorganization of the northeast railroads in 1975.)

The problem raised by the airline case is repeated in other products. In retail stores, there are always some fast-moving, profitable items and some slower-moving goods, which nevertheless provide a variety that is very valuable to some customers and of no importance to others. When traditional stores offer a full line of goods and services, they attract competitors who undersell them by carrying only a few popular products. The full-line store finds it profitable to use profits on some high-volume items to subsidize a greater variety. The discount store upsets this practice by segmenting the market. A good example is that of bookstores. Certain high-volume paperbacks may help pay for slower-moving inventories of scholarly books. When buyers purchase some of both, the pricing practice does not matter since what you spend on paperbacks makes the others less costly. But along comes a grocery store or drugstore that stocks a few profitable paperbacks at slightly reduced prices and draws off this trade from the full-line store. The result may be that the full-line store disappears.

This result harms some consumers, and they may try to change market rules. In some states in the United States, it is illegal for grocery stores to sell prescription drugs. Such prohibition is no doubt motivated by the wish of traditional druggists to avoid competition, but it does favorably affect their ability to stock a variety of slower moving related items.

Some manufacturers realize that the availability of their full line in a store affects consumer satisfactions and the manufacturer's over-all profits. For example, Creative Playthings, a U. S. manufacturer of children's toys, requires retailers to stock either all or none of its goods. Again, this situation can be interpreted in several ways. Laws in certain localities prohibit tie-in sales,6 where the retail merchant must order not only the hot-selling selected goods he wants but also a certain quantity of related goods. The manufacturer in effect takes advantage of a perhaps temporary shortage of one good that is in great demand and uses the occasion to unload some other good. Directly raising the price of the scarce good may be impractical. There is a conflict of interest over any pricing which captures consumer surplus (see Ch. 11). Tie-in sales are not just a non-standard contract which efficiently substitutes for price.

But the full-line requirement can be of a different character. If some merchants can sell only a few high-volume items and undersell the full-line merchants, it may be that the full-line sellers disappear, and not only are the manufacturer's profits reduced but also some buyers lose the variety of choice they formerly enjoyed. Of course, the person with narrower tastes benefits. This conflict necessitates a public decision to determine whether or not a manufacturer has the right to offer only full-line contracts. The right of the manufacturer affects the rights of different consumers with different interests.

U.S. hospitals often do not charge for each individual department or set of services and equipment a price sufficient to cover their full costs.8 For example, open-heart surgery or cobalt therapy units that have sophisticated equipment and personnel and are infrequently used are often priced at less than marginal cost. Thus, other more widely used services have to pay enough to cover these specialized services.

One result of this phenomenon has been the attempted entry of proprietary hospitals, which seek to supply only those services that are in themselves profitable. They attract customers by charging lower prices for those services than the existing nonprofit full-line hospitals. Property right and institutional questions therefore arise as to whether the law should allow this type of competition or whether proprietary hospitals should be regulated in some fashion (for example, all hospitals must offer all services). A similar situation existed in the pricing of local and long distance phones before the breakup of American Telephone and Telegraph.

One's initial response is to suggest that each service should stand on its own feet. This, however, needs some examination if there is interdependence between the services so that the provision of service X affects the profitability of service Y or at least its quality and availability to consumers. Hospitals argue that to get good-quality doctors for their entire operation, they must provide specialized equipment even if such equipment does not cover its own individual cost. A further point is related to the earlier discussion of option demand in the face of information costs. In effect, hospitals make the appendectomy operation user help pay for the existence of certain rare operations in case he might need them someday. The hospitals fill this option demand even where some users may not wish it. (This is one reason why consumer groups want to be represented on hospital governing boards and not leave all decisions to medical professionals.)

Groups of consumers with different interests cannot get what they want simultaneously. The pricing practices of firms, which in effect aggregate or subdivide consumers, affect the cost to different groups as well as the quality and kind of service available. It is important to predict the consequences of different rules for different parties so that the concerned parties can choose the rules that best suit them.

A similar case involves the use of credit cards. Merchants honoring cards pay a substantial fee to the credit card company, such as American Express. Yet the person who pays cash is not given a discount equivalent to the cost of this service. There is no problem if everyone uses credit or the same frequency of credit. But those who consistently use cash subsidize the credit users.9 They may then decide to use credit also since its marginal cost to them is zero, even though they do not find the convenience of credit worth what the merchant has to pay American Express. Merchants may find that offering credit card sales increases their business and profits without differential pricing for each service offered. Do they have the right to do this even at the expense of people who do not want the service? The credit companies profit by such practices, which increase their business. Formerly, they only offered contracts to merchants that prohibited the merchant from offering cash customers a discount, but federal law invalidated such contracts.10

The law, however, does not say that cash customers have a right to a discount but says only that they have the right to be free of the bargaining power of the credit card companies in preventing merchants from giving discounts if they want to. Few merchants will want to as long as total profits are enhanced by offering a full line of services at a flat price even if it is not in the best interests of some customers who are defined in a group that they would prefer not to be in.11 Some competitor may try to develop a business out of catering to cash-only customers. But, given all of the other variables, it will be hard to convince customers that prices are lower by the amount of the credit charge. It will be hard to build a business on this basis. If cash customers are in the minority, economies of scale will keep merchants from catering to them exclusively. For purposes of unit cost, the cash-preferring minority wants to be included in the total group, but, for pricing purposes, they would prefer to be in a separate group and not have to choose between different firms, but rather between cash and credit payment to the same firm. Alas, groups conflict with each other and with the interests of the profit-maximizing firm.

The issue is how a given consumer gets thrown in with other consumers with different interests. If he is segmented out, he will be faced with different prices and available services than if he is mixed in with others. It is a boundary issue, though not geographical as in the case with political jurisdictions.

In summary, we note, as does Alec Nove, that "it is erroneous, even in elementary or abstract analysis, to present margins as if they had no context" (1969, p. 850, and 1973). We have seen that the profitability of a feeder transport link is different if considered alone than if considered as a part of the profitability of the total system. The same was true for credit services, hospital equipment, and manufacturers' full-product line availability. Economists' conventional wisdom advises against cross-subsidization and insists that each portion of the product stand on its own feet (or, more colorfully put, "each tub should stand on its own bottom"). Nove was particularly concerned that if nationalized industries would heed this advice, it would reduce total profitability. The concern here is broader and arises from the fact that people have different preferences for closely related goods and services. Not all groups have the same interest in the pricing scheme and product mix offered by the profit-maximizing private or socialist firm. The issue is not profitability versus social goals but of a public policy choice of the product relative to which profit is to be calculated.

Nove warns against the disease of "vulgarmarginalismus," which is loosely defined as being hell-bent to equate marginal cost and marginal revenue in ignorance of the context. The point of conflict in property rights terms is who gets to choose the context. In effect, private merchants define their relevant product for purposes of calculating profit. This situation is acceptable to some groups and harms others. Shall the groups affected have the right to participate in the decision? The same issue arises in socialist firms and regulated industries. Who gets to define what it is that the firm is producing? Typically, in the United States, the state public utility regulating commissions have been much more concerned with regulating the rate of return than with disputes over quality of service.

We shall examine some of the normative rules put forward by economists in more detail in Chapter 11. It is sufficient to note here that the economist's "rule of rules," that marginal cost should equal marginal revenue, has an institutional context. One is reminded of a theme from Chapter 1, where we noted that costs do not exist in nature but are related to the system of property rights. Note that the argument here is not contrary to the internal logic of marginal cost and revenue equation but that there are many such points depending on the institutional rules. We are not talking about modifying the market profit system with social values but about the basis for profit calculation. Of course, the market system can be replaced entirely with an administrative system. The argument here is not for one over the other but to point out that substantial performance difference can be obtained in either system depending on the rules defining the product and what segment is considered marginal.

Political Boundaries

The ability of a person to have government act in accordance with his tastes depends on the tastes of his fellow citizens. Who his fellow citizens are is affected by where he lives and how political boundaries are drawn. Arguments over political boundaries are as old as organized government. The drawing of such boundaries can make a person a member of a majority or minority group by throwing a person in with different mixes of other people in a decision-making unit. However obvious this fact is, it seems not to have challenged such slogans as "Power to the People" and "The government is best which is closest to the people." (See Breton 1974, p. 116.) A series of cases will be examined to see how property rights in political boundaries differentially affect people. (For a general related discussion see Dahl and Tufte 1973.)

It is instructive to observe the behavior of a political representative with reference to the same issue when he represents different groupings of constituents. The issue was involved in a proposal to create a type of national park on a peninsula in Lake Michigan. Most of the local people near the proposed park opposed it because they feared congestion and loss of private ownership of individual cottage sites. The representative to the U. S. Congress from this district reflected this dominant interest and opposed the park. Later, when he became a U.S. senator elected by the people of Michigan, he supported the park. The dominant interest in a statewide vote was the people of Detroit, who wanted a public recreational area away from the busy city. It is also interesting to note that the park was originally proposed by a senator from the neighboring state of Illinois, whose largest city was within easy driving distance of the park. The Michigan congressman and later senator was democratically elected, but he changed his mind when the the boundary changed the character of the majority. It seems that "Power to the People" does not mean too much until we inquire which people. Likewise, democracy does not mean much unless we inquire into the rules for grouping people to vote. One-man/one-vote does not speak to how that vote will be combined with others.

If the people in the vicinity of the park had the right to make the decision, they would have furthered their interests at the expense of lost opportunities for more distant people. The reverse is also true. The establishment of the federal park created costs for the local people for which they received no compensation. The interests of one group are an externality to the other.

For many years, it was a conventional wisdom in political science that the large number of local governments in the United States should be consolidated into a few super metropolitan governments. (For a critique, see Vincent Ostrom [1973, pp. 114-22] .) The most often mentioned reason was to achieve economies of scale in the provision of services. An underlying reason may have been the intellectual appeal of a neat centralized authority in contrast to the confusion of many local governments. More recently, some scholars have pointed out that economies of scale in production do not necessarily require large political units that articulate demand for these services (Bish 1971; Tiebout and Houston 1962). They have pointed to the Los Angeles (Lakewood) plan where many politically independent units contract with suppliers who embody economies of scale but who can provide just the level and kind of service the local unit wishes instead of a common service for the whole area (Warren 1966). The supplier may be either a public or a private body or person. The decision units remain small and therefore are more likely to have homogeneous members than is the large metro government. This sounds like the best of all possible worlds, where each group gets what it wants at the lowest possible production cost. But, alas, the choice of one unit can have an impact on the opportunities of another unit, as was seen in the park issue above.

For many services, the purchases of one unit are of no interest to other people. But, for unavoidable joint-impact goods, the right to make a choice according to your interests creates costs for others. An example is provided by the different responses of different government units to the closing of a major commuter highway in East Lansing, Michigan. The university students of that community blocked the highway to protest the policy of the federal government in the conduct of the Vietnam War. The highway was the major route through the university community between a large central city and still more distant suburbs. The East Lansing political authorities were sympathetic to the protest and were in no hurry to reopen the highway forcibly. They were willing to bear a little inconvenience. The state police and the governor, who represented a broader constituency, had a different view and wanted the students removed immediately, by force if necessary. Where is the boundary of the controlling political authority? If it is local, external costs are created for commuters to the larger area, but if the decision boundary is larger, the interests of the local people are a minority that can be overwhelmed. Government represents the people, but which people?

There are a whole series of such issues. They include questions of city annexation, the drawing up of legislative districts (called gerry-mandering when it goes against your interests), whether a city council should be elected at large or from wards (districts), and school consolidation (Blawie and Blawie 1973). The latter is now creating some anomalies. Some persons want to force the creation of large metropolitan districts so that schools can be racially integrated. However, the enlargement of districts means that in some central cities where blacks are nearing the majority of voters, they will become a minority again in the enlarged district.

One problem in the drawing of boundaries is that their significance for a given group changes with the type of question involved. One geographical area may include homogeneous tastes with respect to drainage or waste disposal but heterogeneous tastes with respect to police or schools. This is why the United States has so many special purpose governments providing a single service. This confusion of overlapping governmental boundaries offends anyone with a sense of neatness and orderliness, but it allows a homogeneous group to further its unique interests (and create costs for others where interests conflict). It is perhaps a quirk of human nature that arguments over governmental boundaries are always couched in terms of abstract efficiency, prevention of waste and overlapping, or bringing government closer to the people rather than in terms of the real conflict of who is going to create costs for whom and why group A's preferences should count more than group B's. Some empirical evidence on the substantive consequences of boundary alternatives is noted in Chapter 12.

A person who finds herself in a decision group where her preferences are always losing out has several alternatives. She can try to change the preferences of the dominant group, use her vote and voice more effectively, try to change the boundary, or move (exit). The property rights involved in boundary definition determine which of these are real opportunities.

The exit option deserves further comment. In the discussion of joint-impact goods (Chapter 5), it was noted that, for some goods, it is impossible for two people to consume different kinds or amounts. For example, whatever the quality of ambient air is available to A is necessarily available to and utilized by B. There is no escape, if the person remains in the air shed. There is, of course, the option of exit from the area. By moving around, people may be able to form relatively homogeneous local political units to avoid conflict in preferences (Tiebout 1956). But such a solution is more complicated than might appear. One problem has already been noted above. Often, there are interdependences between local governments. There may be no conflict within the unit, but there may be interdependences between units. A second problem is immobility. It is costly to move (Scott 1964). When preferences differ, who bears the costs of moving?

Moreover, what is the responsibility of the departing group to those who remain? That people do move to form relatively homogeneous units is clear. Relatively rich whites have left the cities, and urban problems must be solved by those left behind. The issue is the right to exit when that creates costs for others. It is involved in the right of central cities to tax nonresidents who work in the city. It is involved when large factories leave a city for a new location. A parallel case is the right to join a group. Does a group have the right to be exclusive when new entrants create costs for prior members? This was the issue in a court case testing the right of Petaluma, California, to limit its size (Kirp 1976). The possibility to limit conflict by forming homogeneous units of local government is itself limited.



Boundary control is only one way of grouping people and issues to aggregate preferences. Another has to do with the order in which items come up for vote in a legislative body. The rules committees of the U.S. Congress are very powerful because they decide which bills come to the floor for a vote and in what order. The formulation of winning coalitions of competing groups is affected by the order of consideration. This is formally demonstrated by what is called the Voter Paradox (Arrow 1963). Assume a decision unit made up of three people with a preference ordering for three different alternative ways for designing a certain program, as shown in the table below. Successive pairs must be voted on and a simple majority wins.

Individual's Preference Ordering

Individual -1st ..2nd.. 3rd

--A-------- X.... Z.... Y

--B-------- Z.... Y.... X

--C------- Y..... X.... Z

Individual A prefers policy alternative X to Z and Z to Y and so on. Preferences are not single-peaked. If the vote is first taken on the proposition to choose between alternatives Y and Z, alternative Z will be supported by a majority of two individuals, B and A. But, if the first choice is between another pair, a different winner emerges. When Z is paired with X, X will carry the majority. Few actual legislative votes take this exact form, but the general implication is clear. Control of the agenda is an important property right.12 Coalitions of groups form around the basic idea which is first put forward. A common process is to add features to a legislative bill until the necessary majority is reached. The persons who can control this sequence have a great impact on what emerges since in many situations more than one set of coalitions is possible (Ingram 1969). The voter paradox leads to cyclic instability as coalitions can be reformed. Stability is a function of whether the losers think the outcome was fair (willing participation). Appeal to the widely accepted symbols of democracy do not settle the conflict between groups over the agenda.

Referenda are usually regarded as the purest form of democracy. Yet many examples can be cited where the wording of the proposition affected the outcome and produced a different result than might be expected from majority attitudes on the basic principle of the proposition. The group that can control the wording and get its proposal on the ballot first can exercise some impact on the result.

Referenda often contain a mix of features. For example, in 1974, Michigan voters had the opportunity to vote a bonus for Vietnam war veterans to be financed by a bond. Some may have supported the idea of veteran aid but objected to the form of financing. A second proposition proposed bonding for investments in public transportation equipment and facilities. The major portion was for bus transport but some was for airports and harbors. Some people may have voted against the proposition although they supported bus transport because they objected to aid for airports and harbors. The other modes of transport were probably added in the hope that this would build a winning coalition for the whole proposal. Or it may have been added by special-interest groups, which thought they could get a free ride out of a generally popular bus transport proposal or by those who wanted the whole proposal defeated. The same problem of separating out the features you support from those you do not is common in all representative political systems. Candidates represent a package of views on different issues. You may find that parts of the package you prefer are advocated by several different candidates for the same office but embraced wholly by no one candidate. This situation is referred to by Bish (1971, p. 70) as the "blue plate menu problem." Voters are forced to choose among packages of issues where the packages presented to them are not directly under their control. They are free to choose, but the choices given to them are not of their liking. This is the now-familiar distinction between the freedom to choose within one's opportunity set and freedom to choose the set.

The blue plate menu problem is not unique to public choice. It also occurs in the market. Many goods are made up of several features and not all combinations are feasible. Buyers are often forced to select among brands that represent different mixes of these features. This "indivisibility" can also be inherent in goods. Not all combinations of features in a good are feasible. This is related to what was termed pre-emption in the case of joint impact goods above. (Also as noted above in the section on product definition, buyers are often forced to buy a manufacturer's full line of products. This is why Albert Breton [ 1974, p. 50] refers to the blue plate menu problem as one of "full-line supply." It is the same issue as that of tie-in sales discussed above.) How often have you wished someone would combine the specific features of brand X and brand Y into the good that you really want? The fact that you buy brand X in the market is not confirmation that you approve of all features of the product any more than the fact that you voted for candidate X shows you approve of all of the positions taken by the candidate. Consumers and voters only have the exit choice of buy/do not buy, or vote/do not vote, which hides much information about their preferences. The problem is that people do not agree on the packages of features that should be presented for choice. Thus, people agree on neither the property rights in the market nor the politics that controls who gets to put the packages together.

Breton (1974, pp. 115-16) argues that federal (as opposed to unitary) governments better reflect individual preferences since the representative from the smaller district must reflect the more homogeneous tastes of constituents. In small districts, though the opportunity for blue plate menus is reduced, the eventual conflict is only postponed or its location transferred. In a unitary system, the minority loses out to the majority at the time of a popular election. In a federal system, the loser is the one in a neighboring district who bears the cost of choices that are made within a district where she does not live but that have external effects. Care must be taken not to speak of preferences in the abstract. The point is not that one system or another better serves preferences but which group's preferences are better served. This is not to deny that there may be homogeneous preference in some cases.

In national decisions, preference aggregation is different with an electoral rule of plurality vs. proportional representation (Schumpeter 1942 and Commons 1907). These rules affect transactions and resulting winning coalitions and compromises between and among citizens, party leaders and parties in the legislature. Breton and Galeotti (1985) hypothesize that proportional representation results in ideologically purer party platforms which facilitates trust between citizens and representatives. Voters exchange votes for effort by politicians to achieve the voter's interest. The information cost of a citizen is less when the party platform is narrow. If a party adds features to attract a plurality, it dilutes its position and makes it harder for its original supporters to monitor its performance and preserve its trust.

In proportional representation, there are more parties representing various views. These conflicting views are not compromised within the party. In parliamentary systems, the major compromises are made when the government is formed rather than on a day to day need to form a winning coalition for a particular piece of legislation such as in the U.S. Congress. Hypothesis: The character of the winning coalition in different (different preferences count) depending on whether the compromise is made to achieve a plurality, to form a parliamentary government from various parties where none has a clear majority, or to achieve a majority vote among political representatives in a non-parliamentary, weak party discipline system. The transaction and monitoring costs of any particular interest groups and thus its ability to achieve its goals is affected by the rules for preference aggregation. However, groups may derive utility from seeing a politician wearing their label even if they don't affect legislative outcomes. Depending on the distribution of preference and the rules, a minority can influence outcomes by providing the margin of victory in forming a parliamentary government, or on a particular legislative vote, or by making a party modify its platform to achieve a plurality (Galeotti 1980). As always in economics, the margin is critical but the rules define the relevant margin.

Voting rules affect the aggregation of the preference of people with different intensities. The table above illustrated an outcome with binary voting, but other rules are possible. Point voting for example give a certain number of points or votes to each person which may be distributed among various alternatives. If a person has an intense preference for a candidate in a multi-seat contest, all of the person's points can be given to one candidate. This can lead to strategic voting when an individual tries to influence voting outcomes by not voting her true preferences.

Another alternative affecting communication of the intensity of preferences is vote trading (log-rolling). A person attempts to get their preferred alternative by offering to support the other voters preferences where the person has only mild preferences. As Fry (1978, p. 76) notes, "Vote-trading always increases the utility of the voters participating ... The voter(s) not participating in the trade, on the other hand, suffer a utility loss ..." In the case of multiple vote trading it in possible for each voter to be worse off after trading (Bernholz 1974). While the individual trades are rational, the overall result may not be. The degree of uninimity needed for public choice is discussed in Chapter 11 in the context of welfare economics.

To conclude, it has been seen that rights relating to boundaries, agenda, and vote trading control many aspects of interdependence not addressed by the rules of pure competition in markets or simple democracy and majority voting in politics.


In this section, some goods that combine several of the characteristics noted above will be examined in the context of alternative institutional frameworks. What are the conditions for group formation and growth? When will a group form under private contract alone? When will groups welcome new members, and when will they try to be exclusive?

Group Formation Related to Information and Exclusion Cost and Size

Consider a good that is not divisible in the sense that the payment or effort by one person will obtain any usable quantity of the good. The total cost is such that it requires the contribution of many people to provide a usable amount of the good. (However, the physical amount of the good is variable.) Further, the costs of exclusion are relatively high. If the good is to exist for one person, it will be hard to exclude others. Examples are national defense, voting and lobbying to secure government action such as price support for farmers, and a significant change in ambient air quality. (Note: some of these are also joint-impact goods and others are not, so this feature is not decisive in the discussion to follow.)

With low exclusion-cost goods, a group of customers forms without anyone being particularly conscious of it. Anyone who does not in effect pay to belong to this "group" will not receive any of the good. Failure to buy (contribute to the good's cost) provides immediate feedback consequences. What is the case for a small group of people interested in a good as described above with high exclusion costs? Imagine a group of ten people each willing to pay $100 for this good, which happens to have a total cost of $1,000. If no one offers to buy the good, it will become relatively obvious to all. The effect of one person's choice on others is easy to detect compared to the case of large groups to be discussed next. It is imaginable that the ten potential consumers are scattered over the country and may not be known to each other (information cost). But let us suppose that this kind of information is not a problem or has been supplied by some private or public entrepreneur or leader. In this instance, the individual can perceive the effect of his not buying or the effect of any other individuals not buying. It is easy to see that his choice will make a difference. If he wants the good, he had better make a purchase or contribution, except if he hopes that some of the others might be induced to give even more and thus give him a free ride. The result may be strategic bargaining and game-theoretic considerations, already noted in the sections on exclusion cost and joint-impact goods. About all that can be said is that sometimes the good will be obtained through private market bargaining, and sometimes it will not (or at least not in the amount that, in the absence of the temptation of strategic bargaining, people might have been willing to purchase).

The key to understanding the small-group situation is that each person can perceive relatively easily the consequences of his acts. This is not the case for large groups. It is hard to perceive that your action will make a difference. Frohlich and Oppenheimer (1970) suggest that even with high exclusion costs, individuals calculate the probable contributions of all others and make their contribution when it could make a difference. But size does affect the ability to obtain the information necessary to make this calculation. If the group is large, the individual will find it costly to know how many people would give how much. But, more fundamentally, how can a person estimate how much others are going to give and then decide on how much he will give, when the people he is trying to estimate are asking the same question? He is trying to estimate a whole series of interdependent decisions. This is not just a matter of high information cost, but simply impossible information.

Even incompatible-use goods have information problems that are not solved simply by competitive markets. In order to make good decisions, a producer of an agricultural product needs to know not only demand but also the probable supply to be offered by the sum of other producers. But this information is impossible to obtain since everyone is simultaneously making interdependent decisions. If the current year's output is high relative to demand, prices are depressed and losses suffered. This is a signal to reduce output. But which and how many producers will get out? If everyone else reduces output, the individual can maintain or even expand output and take advantage of next year's higher prices. But if others are going to maintain output, A should get out of production. The inability of a firm to obtain the necessary information results in a cycle production pattern referred to as the "cobweb." Output alternately overshoots and undershoots an equilibrium price that will just cover costs of production.

In the small-group situation, experience is suggestive and gives useful feedback. For example, assume in practice that my failure to give usually suggests that it makes no difference. I fail to give, and the good comes anyway. This tells me something, and the next time perhaps I am tempted not to give again. Now the good ceases. But can I be sure why? Was my noncontribution decisive or would potential total donations still be short of total cost? Suppose I get information on the size of the shortfall. Can I assume that there are many people like me who will realize their mistake in not supporting something we really want and now make their contribution? This guessing game quickly becomes exhausting even for Ph. D. s. It is difficult to imagine a prediction precise enough to identify the utility of my $10 contribution to a million dollar project. The perfect-knowledge assumption of neoclassical economic theory misses a lot of real-world situations of interdependence.

It is a mistake to over-intellectualize the decision-making process. Just because we can imagine some set of calculations consistent with observed behavior does not mean people actually think that way. Perhaps a bit of empathy would be more useful. When I contribute to the local public television station, I have no illusion that my contribution will be decisive. I have no estimate of the probable contributions of others, of total cost, or even of the number of people that watch public television in my area. I am not aware of any evidence that other viewers make these estimates. I am aware, after writing this book, that I have the opportunity of being a free rider. But I do not reflect on it long. I make a contribution because I agree with the purposes of public television and enjoy its programs, and it just seems like the right thing to do. I have learned a certain behavior. Scholarly people make contributions to scholarly and artistic enterprises in their communities, and it would pain me not to live up to my self-image as a scholar, even if no one in town knows I made the donation. I can imagine people making the decision on a variety of bases, and perhaps even some detailed calculations of advantage, but I would not care to make any interpretations of the sum of willingness to pay from observed voluntary contributions to goods with high exclusion costs and large numbers.

The high-exclusion-cost feature may imply a conscious calculation of advantage, but it also means that there is no quick and easy perception of the consequences of individual action. In this situation, some people are calculating and some are acting out of habit, sense of obligation, the fulfillment of some personality need, and so on. I suspect that some are not even aware of whether they are calculating free riders or whether they really do not value the good enough to make any contribution. (There is much reason to believe that not all choice follows a prior precise mental picture of what is desirable but that preferences are formed in the process of consumption. Preferences can be both a dependent and independent variable. Sometimes preferences guide consumption and other times are derived from consumption. I am indebted to James Shaffer for insight on this point.)

The role of the political entrepreneur cannot be ignored.13 A person may incur certain organizational costs while offering a good to potential contributors in the hope of capturing a salary from future contributors. Such a person may seek legislation such as the union shop rule to solve a free-rider problem, find a selective good to sell to finance the group activities that have high exclusion costs, or engage in propaganda and patriotic efforts to create a noncalculating sense of community to support contributions. People who have organizational leadership positions also gain some personal reward in terms of status.

Voting in the face of high exclusion costs for political outcomes can be conceptualized as an exchange of trust (Breton and Galeotti 1985, Salmon 1983 and Coleman 1984). Trust is a substitute for policing cost. Politicians can't contract with voters to vote the voter's preferences in exchange for votes. But, if there is sufficient trust, both parties behave as the others expects.

The concept of product is again critical. Denton Morrison (1971) argues that there is a difference between a good conceived as a desired want and an action that is conceived as right and just. There is a difference between wanting a product and feeling it is deserved. When people contribute voluntarily to a reform movement, they perceive they are doing something more than the usual consumer purchase, which compares utility and cost. The person who contributes to the NAACP or a women's rights movement may perceive the product in a different emotional sense than the person who pays taxes for mosquito control. The former carries a heavy moral flavor of righting a wrong. Morrison speaks of products that have "reform utility." Albert Hirschman (1984) similarly distinguishes instrumental and non-instrumental behavior. The latter makes a person "feel more like a real person" even if not instrumental in providing a physical good. Work toward a HEC good increases one's sense of belonging to the group which may explain voting participation (Pizzorno 1983).

The knowledge that provision of high exclusion cost goods is improbable without heroic behavior can make such contributions more desirable to would-be heroes. Utility and knowledge are significant variables in predicting performance. Successful institutional design should stress morality, the heroic, knowledge, and try to get sympathy of the masses, as well as provide side payments. For a discussion of selective incentives versus ideological commitment, see Cell (1980).

Albert Hirschman (1983) has inquired into the characteristics of successful group action in providing high exclusion cost goods in developing countries. He emphasizes the accumulation of experience and trust. Past attempts at collective action, even if unsuccessful, increase the probability of future success.

The conclusion is that sometimes people make voluntary market contributions to these goods and sometimes they do not. People do make substantial contributions to churches, philanthropic organizations, and so on. But the same process of empathy tells me that not all of the present level of such goods that I now enjoy may always be provided for in that way. Some people would not get the level of these goods they are willing to pay for if private contract were the only alternative.

A commonly used institution in the United States is that of the tax referendum in a local government area. If a majority of the voters approve, all taxpayers must pay. The agenda is controlled by political officials who provide information on total cost and benefit and each taxpayer's share. There is no opportunity for strategic bargaining, and no voter needs to estimate the demand of others. The group forms, as in a market, without any conscious contract, and so transaction costs are low. The key ingredient is a preexisting organization and "contract" where most people agree that the whole procedure is legitimate. This system is widely used in the United States for public education and other local capital improvements and supplements goods provided through legislative representation at all levels of government.

Where groups are large and habits of giving and social bond are weak, we turn to government finance, rules that help hold the group membership together such as the union closed shop, or some mixtures of goods with low and high exclusion costs such as the American Farm Bureau or American Medical Association lobbying efforts, or selling advertising to finance free broadcast television.

Or, we do without. Again, empathy tells us that there are goods and services of this type that people individually might pay small amounts for but that do not seem to be provided or are provided at low levels. Even casual observation dramatically contrasts the huge amounts of money raised for lobbying by small-member groups such as oil, sugar, milk, or tobacco producers and by special interests such as gun users and sportsmen with the small amount of money spent by general consumer groups or voter groups such as Common Cause, which claim to speak for the interests of large numbers of people.

People who want national defense, environmental improvement, consumer protection, and so on have a special problem because it is difficult for the individual to perceive that his action makes a difference. Mancur Olson (1965, p. 50) refers to such people as "latent groups," because without strong habit and social bond their interests can only be mobilized with the aid of selective incentives. (Compare Chamberlain 1974.) The problem is that these selective incentives are themselves a type of good that has high exclusion costs. Somehow, someone has to get an organization started, a referendum procedure instituted, and the union shop law passed. Why then do some of these latent groups materialize and others do not?

There probably is no substitute for some minimum amount of internalized ideology (what B. F. Skinner would call learned, reinforced behavior) where people act without calculation of individual advantage. In many cases such an amount is sufficient. In other cases it is the base on which other selective incentives are built. In the formative period of many present organizations, there was a small group of dedicated (sometimes even called fanatical) people who devoted themselves to the cause. The processes that produce people like Eugene V. Debs, Gandhi, or Luther are very complex (Erikson 1958 and Salert 1976). The role of the charismatic leader is often a factor in group formation, but more complex are the processes producing a widely held ideology that these leaders can take advantage of or that energize and nurture leaders. No latent group's success can be fully explained without consideration of its leaders. It helps if leaders and their apostles are emotionally committed. They act because they must act and not because of any calculation that worries about free riders. They have an ego and a vision that their acts will make a difference. The role of the heroic cannot be denied.

For the product or cause to gain wider acceptance, it helps if it can be described simply. In more pejorative terms, it helps to have a slogan. You can get people to respond to a leader who proposes more money for national defense if the enemy can be dramatized in simple terms. Difficult as the civil rights movement was to organize, its objective can be described more easily than the many complex issues involved in something like consumer protection. It is hard to get an emotional reaction to complex goods; people's attention spans are exhausted before they are willing to reach into their pockets for a contribution or to vote.

Sometimes the difference is made by violence. The willingness of a few people to resort to violence cannot be ignored, and neither can the willingness of the rest of society to tolerate it in some instances and not in others (Graham and Gurr 1969). Social movements seem to require a certain environment of sympathy.

The role of the private entrepreneur in mixing goods with high exclusion costs and low exclusion costs is another factor. The rewards here probably are of the more material and calculated sort. The fact that the defense industry profits from national defense and that manufacturers do not profit from consumer protection expenditures is a factor in explaining the amount spent on these two activities.

The large latent groups in the United States have an important ideological barrier against them that small groups do not have. There is an ideological preference for individual market organization, and this organization is sufficient for the interests of many small groups. But, at a certain point, many large groups need governmental administrative transactions to sustain their interests. Competition between large and small has been most uneven because of the ideological commitment to the market, which is regarded as somehow natural and good while administrative actions are unnatural and therefore suspicious. When small groups compete with large groups, the prevailing ideology has the effect of a valuable property right that constrains the options of large groups.

The role of the family must be included in a discussion of the effects of group size though it can only be touched on here. There is a class of goods where exclusion cost is low but information is needed to determine who is eligible to participate in their use. Group size and interpersonal relationships can affect the cost of this information. For example, in the winter of 1977, numerous migrant workers were unemployed because of a freeze in citrus-growing areas. Some taxpayers represented by government wished to provide aid to these people. On paper, it was easy to exclude ineligible people from the disaster relief, but it was expensive for government administration to determine eligibility. Many fraudulent claims were filed by people claiming to be fruit harvesters who had never been so employed. In some other societies, disaster relief would be provided by the extended family and the local community; people able to help would personally know those needing help; it would be cheap to determine need, and would-be free riders would be rejected.

A number of factors affecting provision of goods with high exclusion cost have now been noted including group size, information cost, habits, political entrepreneurship, reform utility, transaction cost, ideology, and violence. Some of these are affected by changes in property rights, and others are not. Some can be manipulated in the relative short run by a change in the legal rules while others are the result of a complex, long-term learning process embedded in informal cultural rules.

Group Growth: Exclusive and Nonexclusive Groups

We can observe some groups that welcome new members and others that do not. The reason for this is related to the combination of exclusion cost and joint-impact characteristics of goods. Inclusive groups welcome new members because they seek a joint-impact good where the marginal cost of another user is zero. Where it is costly to exclude users anyway, the more who contribute, the lower the cost to existing members. For example, environmental groups such as the Sierra Club want as many members as possible to help pay for their lobbying efforts. However, there is an ambivalence that derives from another aspect of nonoptional joint-impact goods. That is the fact that one member of the group cannot enjoy a different physical quantity or quality of the good. Thus, if current members have homogeneous tastes and agree on the level and kind of product sought, they do not want new members with different tastes. Therefore, some organizations, as they grow, welcome new members to lower the cost per member, but become undemocratic, with policy making controlled by the old guard establishment. This is why governments sometimes regulate the internal political rules of private organizations. Even where exclusion of users is costly, groups may try to exclude formal members who could affect policy.

Exclusive groups, on the other hand, do not want new members or new users because the marginal cost of another user is not zero. The cost of obtaining benefits for another person exceeds the revenues contributed by the new member. A producers' cartel does not want any new members because the market is fixed and would have to be shared with new entrants. Of course, if the existence of the unwanted new entrant cannot be prevented, it is better to have it a member than an outside competitor. The new member lowers the utility of old members, but not as much as destruction of the cartel. Exclusive groups need 100 percent membership of the people who cannot be excluded from their product, if the product exists at all, and thus exclusive groups are hard to maintain._

Some lobbying groups are inclusive up to the point where another member raises costs more than revenues. For example, a group of churches seeking exemption from the property tax will welcome all churches as members of their lobbying organization. But they may not welcome more and more varieties of not-for-profit firms because at some point the resistance of the rest of the taxpayers stiffens and the cost of obtaining the exemption exceeds the dues from new members. In other cases there may be economies of scale in lobbying. Larger groups may raise more resources to compete effectively with other groups, but they also raise the resistance of others. Pincus (1977) documents the history of U.S. tariffs where at some point high tariffs achieved by special interests cause free trade advocates to run for public office.

Other varieties of ambivalence also occur where the organization has multiple goods. Labor unions, for example, want more members to support their national lobbying activities but may erect high entry barriers to workers in some industries to avoid competition.

There are many consumer goods where exclusion is possible, but they are nevertheless owned by a group rather than an individual. The size of the group is related to the degree to which the good is a joint-impact good. (Where there is no jointness, the good likely will be owned and used individually.) For instance, an individual golfer realizes that she cannot afford to have a course of her own. Another user does create extra cost, but there is a degree of jointness. The individual will seek other member-owners in a club up to the point where the congestion causes her total benefit to reduce faster than the average cost of membership declines because of the revenue from new members (Buchanan 1965 and Rowley and Peacock 1975, p. 118). Certain goods then will be used in common with or without control on frequency of use by an individual. For some goods with a degree of jointness, the "club" is an alternative to government or individual ownership (Sandler and Tschirhart 1980). Group size will increase up to a point, and then new members will not be accepted. Persons can reduce conflicts by moving to the club or local community that provides their preferred mix of goods and prices (Tiebout 1956). The right of the club to exclusivity and the right of the community to control its size and homogeneity by zoning is thus instrumental.

Economies of scale create a type of jointness in that size affects the average return to each individual factor. Another firm "member" can be added without subtracting from the returns of previous members sharing equally in total output. A labor-controlled firm (or cooperative) will add new worker-members to maximize enterprise income per head of the existing labor force rather than add workers to the point where marginal cost (wage rate) equals marginal revenue, as would a capitalist firm (Wiles 1977, p. 70). (See the discussion of employer surplus in Chapter 7.)

To conclude, many public decision bodies are exclusive groups and do not welcome new members. Membership on bodies that decide rights issues between A and B is itself an incompatible-use good. Power has been defined here as participation. Membership on a rights creation and distributing body is equivalent to resource (factor) ownership in that it is the source of the opportunity to participate in decisions.


Interrelationships of economy and polity are of concern to many institutional scholars. One view is to see votes and money as parallel ways to fulfill consumer demand. In the polity, parties compete for votes in the same way as firms compete for dollars and similar equilibrium conditions apply. Downs (1957) uses the spatial equilibrium model developed by Hotelling to predict that parties starting with different programs will maximize votes by moving toward the center (the median voter). Furthermore, it is argued that the result in Pareto-optimal (Riker and Ordeshook 1973, Chapters 11 and 12). In this view parties (as firms) have no independent policy goals, but must serve the median voter. Institutions seem to be of little significance except the usual ones of competition and free entry. This theory has concentrated on axiomatic propositions with little empirical testing.

Another view is that parties and the governments they form pursue independent objectives (Fry 1978, p. 155). They must give the voters enough to get elected, but there can be slack within which to pursue their own interests. And, this slack is not just the result of legal barriers to competition. It is hypothesized that voters have information costs and are myopic, forgetting performance during the early years of a Presidents' administration, but highly impressionable by income and inflation levels in election years. There is then no equilibrium of party programs and performance. Rather there is an evolutionary process as parties try to capture the sequence. Parties pursue some of their own ends during their early years in office and increase public spending and transfer payments close to elections, thus contributing to economic instability. Government instead of just compensating for the usual business cycles cause cycles of its own as the supply of government goods fluctuates. Empirical work by Fry (1978, p. 154) suggests that the U.S. and U.K. "governments pursue an expansionary policy before elections, when they are unsure about their re-elections." This model speaks of institutional variables for ideology and legal constraints, but the simple proxies used to represent them are questionable. A dilemma of empirical work using the usual econometric techniques is that available data is imperfect but the use of richer data turn the analysis into a qualitative story. The so-called institutional variables used to date in econometric studies are not instrumental in the sense that they clearly contrast performance being affected by institutional alternatives. The results are called explanatory and might even predict changes in public spending, but do not yet readily identify how one would change institutions to get a different performance.


Voting and other forms of participation in public choice can be conceptualized in terms of transactions that are ordered by structural rules relating to situations of interdependence. First comes the right to participate. Then comes a host of other rules that condition that participation and influence eventual performance. For example, each voter may have the right to try to persuade other voters, much like goods producers use advertising to attract customers. In goods production, we saw that A's welfare is often a function of how many others share his tastes. Similarly, A's vote is worth more if many B's have the same tastes. A is motivated to grant persuasive information to B, and vice versa.

In a sense, the time to participate and vote by a citizen is exchanged for some governmental rule (or product). This product may be a high exclusion-cost good and thus invite free riders who will not bear the participation costs (applicable to lobbying activities discussed above as well as direct political participation of all kinds).

To summarize, different rules affecting A's participation in making property rights are highlighted by the paradigm (these are stated negatively, but could be stated obversely):

1. Deny A the right to vote, hold office, and so on (a variety of factor ownership related to the incompatible-use characteristic of political participation). Exclusion cost determines whether the denial can be made effective.

2. Put A in a unit where he is in a minority (boundary issue).

3. Arrange the decision rule (unanimity, majority, and so on) to affect a given person's veto power and decision cost (a variety of economy of scale).

4. Increase A's transaction costs via ballot design, physical location, and registration requirement (information and contractual cost factors).

5. Allow the use of economic resources to persuade. A more subtle version is the impact of organization of economic activity on tastes and ideology (related in part to factor ownership and competitive rules).

6. Control the agenda to facilitate formation of coalitions excluding A.

7. Regulate private contributions to political lobbying.

In short, the paradigm suggests there is much more to rules for making rules than is captured in vague concepts of democracy and the nominal right to vote.



It is one thing for a group or individual to fail to meet its preferences because it loses to a conflicting group; it is quite another to defeat itself. How is it that like-minded people fail to act in concert to obtain their objectives? Why do people who appear to prefer state of affairs X to state Y nevertheless act to bring about Y? The previous discussion of group formation had to do with provision of goods and services with money costs of production. Following the conceptualization of Olson, the large-group, high exclusion-cost problem was conceived in terms of (1) one person or few persons never having a sufficiently large gain to be able to afford the total cost and (2) no one noticing when an individual exits the group. As stated, this is not directly applicable to situations where there is no money cost. (However, the following discussion may illuminate some of the elements underlying the Olson conception.) Here we note that people do not cease a behavior that all agree is leading everyone in the less preferred direction. It does not make sense to talk about one person paying another to do something that both agree leads to a superior situation. The problem is not simply to pay for the production costs of some common good, but to cease self-defeating behavior. The subject is not the kind of product that governments usually provide with tax dollars, though it is the object of government policy. Perhaps some examples will make the topic clearer:

1. Depression: As total demand declines, each firm tries to solve its cash flow problem by laying off employees. If all firms do this, total demand declines, creating still further layoffs. Most agree that the macro-result is undesirable, but no one firm will unilaterally keep its employees. Each firm says, "I will keep my employees if you do," but no agreement seems forthcoming.

Inflation has similar characteristics, but in a different direction (Maital and Benjamin 1979). A labor union pushes for higher wages only to discover others have done likewise and the wage increase is defeated by inflation in consumer prices. No one union can practice unilateral restraint and survive. Firms are part of the same process. They try to solve their problem of increasing input costs by raising prices only to find that input costs have risen again. Of course, some gain from both inflation and depression.

2. Cold war: Each country tries to increase its security by buying more armaments. If each one does this, relative security does not improve. The cost of arms just goes up in an arms race. The only winner is the arms manufacturer. Most others agree that the macro-result is undesirable. Each country says, "I will reduce armaments if you do," but no agreement is forthcoming.

3. Price war: The case is essentially similar to that for the arms race. Each seller meets the competition and no one gains except consumers. (See Chapter 4. )

4. Population pressure: As population increases, resources per capita decrease. The decision to have children is probably governed by tradition and habit. To the extent that this is a calculated choice, each child is worth what the parents must give up to support it, and many decide to more than reproduce themselves. Some agree that this macro-result is undesirable. Some may say, "I will not have another child if you do not," but no agreement is forthcoming.

5. Transportation deterioration: Decline in the quality of public transportation seems to snowball. When quality changes (or other options become more attractive), some former users will exit the system. Where there are economies of scale, each one who exits raises the costs for those who remain. As cost increases, others leave, quality further decreases, more leave, and so on in a familiar sequence. The riders try to solve their problem as best they can, given their opportunity set. If each does this, public transport disappears.14 Some people agree that the macro-result is undesirable. If the first bus rider to exit had been paid the difference between riding and his next best alternative, the whole spiral might have been avoided. Some might say, "I will contribute to prevent a few people from exiting if you will," but no agreement is forthcoming.

6. Agricultural surpluses: In a purely competitive industry such as agriculture, the early adopters of new cost saving technologies increase their returns. As output increases, price falls and other producers must adopt the new techniques to survive. As they do so, the early adopters no longer earn profits. Most farmers understand this process that Willard Cochrane (1958) termed the agricultural "treadmill" but are powerless at the margin to extract themselves. When the treadmill is combined with immobile assets, and fluctuating demand and supply, farmers produce too much with low return and capital losses.

These are enough examples to illustrate the problem. They are some of the most vexing of our time and are not the bees-and-honey or even pollution case usually used to illustrate that there are a few, special problems where marginal decisions of each person in a market lead to situations that few want. Rather, they are representative of a wide range of problems.

What do these cases have in common? First, they all seem to have a group with a common interest--a group that may conflict with others but will not differ among themselves on the performance they prefer. Each individual tries to choose the best alternative open, but the result is not what each really wants. Seen this way, the following discussion cannot help but comment on the conventional wisdom, which asserts that welfare is best served when each person acts in his own self-interest without concern for others. This is Smith's fabled "invisible hand."

To a considerable extent, countries with market economies have built institutions based on mutual selfishness. Consumers, for example, may have a common purpose, but they do not see themselves as a group.Any standard economic principles book can be consulted for the model that adds selfishness and lack of group consciousness to the rules for perfect competition and shows how the result is the loswest possible costs for everyone. Suffice it to say that in a market everyone lookls for the best buy and makes marginal adjustments in inputs and outputs. For example, a firm is always on the lookout for cheaper production methods. But, with competition, profit is short lived as others follow suit. The producers do not act as a unit, but each follows in turn.

In the case of consumers, if A finds that the price of a given product is too high, she exits to her immediate benefit. There is a micro-motive leading to individual action (Schelling l978). The consumer does not concern herself whether there are others who share her view. If others do share the same view, they individually will perform the same act. If enough exit, the producting firm will get the message and alter quality or price or go out of business, releasing the resources to someone who will try to please the consumers. The individual consumer increases welfare by acting alone, and if others join in, so much the better; the firm is reformed. The micro-motive and the macro-result are consistent.

But this consistency is not always present. Take any one the the five examples noted above. In the case of a drop in aggregate demand, exit (that is, firing some employees) helps the firm's short-range balance sheet, but as other firms also make the same marginal calculation and adjustment, the macro-result is one that none desire; the micromotives result in a sum of behavior that no one in any group wants. The lack of group consciousness prevents concerted action. It is not enough to act unilaterally and let the group form unconsciously. If everyone is not to fail to achieve her most desired position, it requires agreement on propositions like, "I will do X if you do X." If A suspects others may cheat, then A must cheat too or be greatly disadvantaged.

Why are conscious group agreements so hard to achieve? The answer is essentially an economic one. Exit is cheap (low transaction costs), while voice is expensive. It is cheap to buy or not, sell or not, hire or fire, and so on all by yourself. But it is expensive to voice your concern and communicate explicitly with another person. It is expensive to persuade, argue, cajole, negotiate. There is always the chance that what began as a benevolent or just selfish relationship will deteriorate into malevolence as voice proceeds. Of course, in the large group case, voice to the individual seems unlikely to have any effect if it is thought about.

The term "social trap"15 has been applied to the following situation: The micro-motives are not consistent with what individuals who share a common preference want to obtain. In behavioral terms, it is when the micro and marginal reinforcers are not consistent with the macro performance wanted. If each person makes the best marginal choice within his opportunity set, each will find that the goal he seeks evades him. This is not the usual individual versus group problem where the individual has objectives inconsistent with the welfare of a larger group. It is even more tragic; it is the individual versus himself. This author is loathe to use the adjective "social" because it is so often used in the global sense when the facts are one of interpersonal conflict. Yet it seems appropriate to call this a social trap in the limited sense that in some groups, the members fail to pursue their own interests. This is truly a trap with characteristics of a Greek tragedy. The trap is confounded if the micro reinforcers are immediate while the macro results are long run, but this is not the key distinction. But remember, that the trap that causes one group to fail is often to another groups advantage (as in the case of cartels and consumers).

If the individual is to be successful, she must be aware of the group and have confidence in predicting the actions of others. Marginal choice within the available opportunity set is often not sufficient. It requires the non-marginal choice of a different kind of opportunity set. In many market transactions, it is possible to further common interests without being conscious of each other. Not so in social-trap situations. For the individual to get out of the social trap requires a consciousness of others. This consciousness need not be benevolent or self-sacrificing, though it probably helps. It does involve some measure of trust. If malevolence is present, the trap deepens.

Under what conditions does the social trap occur? What do the cases cited above have in common? The answer lies in the character of the production function and not simply the lengthy time between short run benefits and long run costs. The problem of the social trap is not lumpy inputs or outputs, although, as we shall see below, these conditions often confound a solution.

The definition of social trap to follow differs somewhat from other literature on the topic. The problem of micro-motives conflicting with desired macro-results is due to the existence of competing lines of action whose production functions are such that the dominant activity has net benefits at the margin of effort while the dominated choice has less or none. The micro-motive is supplied when there is some act under the individual's unilateral control that promises to produce some welfare improvement for that individual. The alternative line of action that would be consistent with the more preferred long-run result is marked by the fact that no matter how hard the individual tries, alone he can produce no net benefits or fewer than in the dominant activity. Higher net payoff from the mutually desired product cannot be achieved by any input change under the unilateral control of an individual. The altruistic, sacrificing person will achieve nothing for those loved unless the actions of others can be counted on, and thus love may not be enough. The preferred, but dominated, choice produces superior net benefits only when a certain threshold of inputs is reached which requires trust that others will act in a consistent manner. Collard (1981, p 37) suggests that a Kantian ethic is often needed to reach the necessary threshold. The Kantian would act only in a way that everyone could act and still achieve the desired result (regardless of concern over thresholds).

The social trap as here defined is related to the classical case of the "prisoner's dilemma" (Collard, 1981, Ch. 4). Mutually beneficial results for those caught in a prisoner's dilemma depend on mutual trust. If one person's trust is not reciprocated, that person is a sucker.

Where the social trap does not exist, it is possible for an individual to put in a little and get out a little. Reward is sufficient for some individual action. The shared objective of many individuals so acting takes care of itself without anyone's being conscious of those with similar objectives. But, in the social trap, no one can achieve any of the mutually preferred macro-net benefits by any unilateral act. It takes the concerted inputs of many actors before there is any output with a net positive value.

Where a choice along a dominated payoff function competes with choice along a dominant function, the dominant activity always wins unless there is an overriding habit, ethical code, or corrective institution offering reinforcements in a different direction. Choice of dominated alternatives is always more costly to organize because conscious joint action is required. The unseen hand and detached, impersonal selfishness work only when simple exit or effort is consistent with the individual's long-run preferred position. This is why people who want peace get caught in an arms race, and why all of the social- trap situations noted above occur.

So much for why social traps occur. Why do they persist? How can people get out? Is there an institutional corrective? The problem sounds like the usual problem of common property, as discussed in Chapter 3. The conventional answer to common property is to dissolve it in favor of individual property. This is only possible if the trap were institutionally created in the first place rather than being inherent in the situation. The other conventional approach is to give up some democracy for authoritarian social control. It is conceivable that the right to choose along the dominant production function could be made a right held by some super-private firm or government. For example, the right to have arms could be sold by a world government (similarly for the right to have children and so on). Even assuming that those who do not pay could be excluded, there is the problem of who gets the receipts of the sale. Could the people ever agree on a super-ordinate authority? Not unless they are already committed to equality (Edney 1981, p. 16). If they don't agree on equality they will cheat or overthrow the authority. Keohane (1984) explores the possibilities of sustaining international cooperation without hegemony (a state powerful enough to enforce rules on other states) regarding money, trade and oil.

The problem of exclusion cost cannot be ignored. Certainly, in the case of the arms race, detection costs are substantial for the present parties and would plague any new superpower. In the case of public transportation, it is conceivable that a market could be organized to buy off existing patrons. Conceivably, it would be possible to exclude future riders who did not help pay to retain current ridership, but in practice this would be costly to organize. Even where exclusion is possible, it may be costly to distinguish whom to exclude for what purpose. It is one thing to conceive of a group whose members share a common purpose, but it may be difficult to know who is in it unless they can be trusted to reveal themselves by voice. How do you tell who has better options than the public transit from those who are lying in order to get the payoff to remain as riders? When you have to rely on voice in order to decide whom to pay, you incur high transaction costs, unless none are opportunistic. The problem is not just to decide whom to exclude from the bus, but whom to exclude from a subsidy to keep riding. The existence of conflicting groups is a barrier to organizing a group with a common purpose. The same sort of problem arises in the depression example. Suppose that some unit were given the right to sell the right to reduce employment. Even in times of drop in aggregate demand, those who share common purpose may want to let some firms who are having other types of problems exit and release resources. There might be substantial costs in determining whom to charge and whom to let go for nothing.

The problem of the social trap does not necessarily arise originally because of exclusion costs, but its solution is hampered by them. But though exclusion costs are formidable, the above discussion ignores a key problem. If a group's members cannot agree to pursue their shared preferred goal, they surely can never agree to let some external unit own the right to control one of their present options. Perhaps this is too pessimistic. People do support government regulations, especially if there is some pre-existing procedure and organization for voting and coercion of the minority.

An area of fruitful research would be to analyze successful solutions to social traps. For example, one variety of trap mentioned above is the price war, which, while not unusual, is controlled. As noted in Chapter 4, a firm with economies of scale is tempted to use marginal cost pricing to attract new business and expand market share. But, if everyone follows suit, any advantage is temporary and the new price may not be such as to pay total costs. Such businesses fear cut-throat competition, which spoils the market. In oligopolistic markets, the quick reaction of competition reminds the wayward firm of its folly and leads to corrective action. In addition, there seems to be a widely shared informal covenant that overcomes the temptation to choose the "dominant production function," which in the long run is destructive to all firms. This covenant is irequently reinforced by exhortations at trade association meetings. There may also be certain accounting conventions that attribute a share of overhead to each unit sold so that managers are not reminded of strict marginal cost.

The role of group size is suggested above. In small groups people can see the difference they make and social pressure is practical. Experiments support the conclusion that small groups whose members can communicate are more likely to extricate themselves from traps (and to supply high exclusion cost goods). See van de Kragt, (1983) and Bonacich, et. al. (1976). Research on actual case studies indicate that the upper limit may be about 150 people (Acheson 1975, Bullock and Baden 1977).

The emotive components of social traps also affect outcomes. Memory of past accommodations affects people's consideration of current alternatives. There are cumulative effects in the degree of trust. If trust deteriorates, communication ceases or becomes distorted leading to further deterioration (Rotter 1980).

The social-trap situation can be summarized. Its essence is the divergence of micro-motives and preferred performance arising out of competing options--one with a dominant production function and another with a dominated function.16 The long-run mutually preferred result is dominated by an alternative choice with higher payoffs to marginal, individual action. The preferred choice has a higher pay-off only when the number of individuals choosing it reaches a certain threshold. In the social trap, the invisible hand becomes the invisible fist that pushes people away from their shared objective. The resolution of this trap must remain speculative. Even the most casual observer cannot miss seeing the persistence and intractability of many of these traps that everyone is caught up in (though everyone is not in the same one).

Though the resolution must be speculative, the alternatives can be briefly noted by references to the concepts discussed in earlier chapters. The role of group size cannot be ignored. Perhaps society might be organized into smaller groups. The other factor that emerges is the role of learned, shared behaviors and status transactions, where the individual does not calculate advantage apart from others in his group but rather has learned a standard operating procedure (ethical code) that is consistent with shared objectives and does not think to consider seriously the otherwise tempting dominant alternative with its rewards to individual, marginal effort. There are definite limits to externally applied property rights no matter how cleverly conceived and applied. The best way to spring the social trap may lie in learning internalized habits of non-calculating behavior, trust, and honesty that are reinforced by cultural practices so that short-run rewards are ignored. Where work patterns and the mode of life create alienation, this learning will be very difficult.


1. In a status system, people may not have the option of exchange, but there may be a traditional serial movement of goods. In an administrative system, the administrator may command a goods movement although the producer or recipient may have no option of exchange.

2. Maurice Dobb (1947, p. 32) makes a related point when he notes that when the focus is on market exchangeables abstracted from physical reality, the proprietor of land contributes as much to output as labor does.

3. The right to participate in a use decision by making a bid is different from the right of part-owner to veto a use and be the recipient rather than sender of the bid.

4. See the empirical study of mutual savings and loan associations, where shares are nontransferable, in Chapter 12.

5. This subject is discussed by Samuelson (1969, p. 117) and Nove (1969); also see Shepherd (1968).

6. An example of a tie-in sale in the public sector is a university requirement for students to live in its dormitories which reduces demand for private housing.

7. I owe this example to Jim Ward.

8. Mark Pauly and M. Redisch (1973) suggest that in not-for-profit hospitals managed from the point of view of the physicians, the resource mix is chosen to enhance physician income rather than firm net worth. Capital equipment is "overused" because it enhances physician income. Even if this is true, it does not speak to consumer differences in desire for different qualities and kinds of equipment.

9. If credit sales reduce robbery losses, the credit user should receive a discount for saving the credit firm money. Opportunity for reciprocal externality affects net consequences of the rule allowing product definition (grouping of services) to the parties.

10. The Fair Credit Billing Act effective in 1975. Also see "American Express Agrees to Allow Cash Discounts," Consumer Reports, June 1974, pp. 432-33.

11. American Express buys advertising urging people to become card holders and to utilize the card at named firms. Such advertising is paid for by a fee based on credit sales. If firms encouraged non-credit use by offering discounts, they would get the benefits of advertising without paying. O'Driscoll (1976) argues that prohibition of discounts prevents free riders. Like any rule that prevents free riders, it creates unwilling riders (customers who have the cost of credit included in prices but who would prefer discounts rather than credit and information services).

For further discussion see Davis (1974). who says "The definition of the alternatives is the supreme instrument of power" (p. 176).

12. Preferences are "single-peaked" when the alternatives can be measured with an agreed-upon one-dimensional variable. For example, if the three alternatives in the text are dollar levels of a budget and the proportions paid by each individual are predetermined such that alternative X is $100, Y is $90, and Z is $80, then though individuals B and C have different preferences for budget size, they accept the dollar-dimensional relationships among X, Y, and Z, while A does not. Individual A prefers the two extreme budgets of X ($100) and Z ($80) to the available intermediate level Y ($90). This jumping over of the intermediate alternative suggested by the dollar dimension indicates that A rejects the scale, and thus preferences are multiple-peaked. Multiple peaks might be expected when tax shares, choice among budget allocations to competing public projects, and choice or the quality of a nonoptional joint-impact good are at issue. Preferences would have been single-peaked if A had preferred X to Y and Y to Z. When preferences are single-peaked, majority voting will produce the same unambiguous result regardless of the agenda, and this result will be at the median of the preferred choices of all individuals. See Black (1958), Ochs (1974, ch. 3), Winfrey (1973. pp. 58-62), and Musgrave and Musgrave (1974, pp. 83-90).

13. Acknowledged by Olson in Appendix to 1971 edition of his book. Also see Wagner (1966), Frohich, Oppenheimer, and Young (1971), Gamson (1975), and Breton and Breton (1968).

14. It is common to argue that global welfare is increased when the number of alternatives to choose from increases. But, in an interdependent world, this is deceptive. From the starting place of primary reliance on public transport, the introduction of more choice in the form of the automobile meant that the end result was only one form of transport. The original choice disappeared in the process. For other examples see Dworkin (1982).

15. John Platt's (1973) use of the term is broader than that employed here and includes many interpersonal conflicts which are the domain of psychologists. Cross and Guyer (1980) confusingly apply the term to cases of incompatible use where the reinforcers create conflict in preferences. Most cases they discuss arise primarily from the difference is short and long run.

16. The previous discussion of Chapter 8 focused on why, for example, a group that hates mosquitoes might not get organized to obtain mosquito spraying. It focused on the situation where no one has sufficient benefit to cover total cost. The current section is parallel but in slightly different terms. No one has sufficient inputs to achieve any net benefit. When this situation is tied to a competing action that presents micro-motives working in the opposite direction, a different problem arises. The parallel to the mosquito case might be where the mosquito-hating group not only failed to buy spray but actually set about increasing the number of mosquitoes.

If you have any questions or comments, please email

Back to Main Page