We turn now to the use of positive reinforcement in the practical control of behavior. This consists in general of the presentation of food, clothing, shelter, and other things which we call “goods.” The etymology is significant. Like the behavior of the individual which is positively reinforcing to the group, goods are “good” in the sense of being positively reinforcing. We sometimes speak of them also as “wealth.” This term has a similar etymological connection with positive reinforcement, but it also includes generalized conditioned reinforcers, such as money and credit, which are effective because they may be exchanged for goods.
As a simple example of economic control an individual is induced to perform labor through reinforcement with money or goods. The controller makes the payment of a wage contingent upon the performance of work. In actual practice, however, the process is seldom as simple as this. When we tip a man or pay him for performing a small service and thereby increase the probability of his performing a similar service in the future, we do not depart far from the laboratory study of operant reinforcement. Behavior has occurred and has been strengthened by its consequences. This is also roughly true when a man is steadily employed. His performance at a given time is mainly determined by the contingencies of reinforcement which have prevailed up to that time. When an explicit agreement is made, however, prior verbal stimuli must be analyzed in order to account for the effect of the economic contingency. Thus when we agree to pay a man a given amount for a given piece of work, our promise to pay is not far from the command analyzed in Chapter XXII, except that reinforcement is now positive rather than negative. Payment is contingent upon the verbal stimulus of the promise to pay and upon a correspondence between the topography of the behavior and certain verbal specifications. The offer, “I’ll pay you two dollars if you mow the lawn” specifies (1) behavior (“mowing the lawn”), (2) a reinforcement (“two dollars”), and (3) a contingency (“if”). To the prospective employee the whole remark serves as an occasion which, if the offer is to be effective, must be similar to other occasions upon which similar contingencies have prevailed.
Fixed-ratio schedules. With the exception of payment “by the job,” the economic control of behavior follows certain schedules of reinforcement. When a man is paid in terms of the number of units of work completed, the schedule is essentially that of a fixed ratio. It is usually known in industry as piecework pay. The same principle applies to commission selling, to the craftsman who makes and sells a standard product, to the writer who is paid by the story or book, and to the small private contractor. Fixed ratio is, in general, a very effective schedule of reinforcement. If the ratio is not too high—that is, if the amount of work required per unit of pay is not too great—and if each reinforcement is of a significant amount, the individual will characteristically work at a high rate. This is as true of the pigeon in the laboratory as of the man in industry. An employee who has been paid on some other basis and then transferred to piecework pay will usually show a considerable increase in speed. The increase is partly the automatic result of the increasing frequency of reinforcement which follows on a fixed-ratio schedule as the rate increases. Some of it is due, as we have seen, to the fact that a high rate of responding tends to prevail at the moment of reinforcement under such a schedule. Progress toward the completion of a given number of responses also has the effect of a conditioned reinforcer. The schedule is more effective if this progress is emphasized—for example, by a visible counter.
A fixed-ratio schedule may, in fact, be too effective. It leads not only to high levels of activity, but to long working hours, both of which may be harmful. A bricklayer paid in terms of the number of bricks laid may “burn himself out” in a few years. Another objection to the use of the schedule in industry is that the increased return to the worker which follows conversion to such a schedule often seems to justify increasing the ratio. Let us suppose that an employee producing a hundred items per week is paid fifty dollars on a weekly basis and that the management offers to pay this instead on a piecework basis of one dollar for every two items. The effect upon the employee is a rapid increase in production. Let us suppose that he is able to increase his weekly wage to a hundred dollars. In terms of current rates of pay this may appear to justify increasing the number of items required per dollar to, say, three. As the piecework schedule remains in force, production may continue to rise. In the long run a very much higher rate of work may be generated by only a slight increase in weekly pay. This is precisely the way in which in the laboratory a high rate of responding is generated under a fixed-ratio schedule.
When the ratio is high or the reinforcement trivial, a fixed-ratio schedule characteristically produces a period of inactivity just after each reinforcement. At very high ratios these periods may be greatly prolonged. They represent, as we have seen, a condition of abulia similar to that in complete extinction in which, although the deprivation is severe, the individual simply “has no behavior available.” He finds it impossible to start on his next assignment. He may report this by saying that he is discouraged, that he can’t face his job, and so on. A typical example of fixed-ratio pay is the salesman selling on commission. When “business is not good,” the amount of work which must be done per unit of reinforcement is high, and abulia is common,
The ratio and the magnitude of the reinforcement show a subtle relation. Is a reinforcement of ten dollars per thousand items as effective as one dollar per hundred, or one cent per item? If a man places a fixed economic value upon his labor, there should be no difference, but this is not the case. One can advance to a high ratio only after a long history of reinforcement at lower ratios. Especially with uneducated labor the ratio may be crucial. Thus a contractor who employed peasant labor to move earth with wheelbarrows found it most effective to pay a small amount each time a full wheelbarrow was delivered to the proper point. The use of piecework pay in industry or elsewhere presupposes a considerable history of economic control.
Fixed-interval schedules. Labor is most commonly paid by the day, week, month, or year. These appear to be fixed-interval schedules. The size of the interval, like the size of the ratio, is a rough function of earlier contingencies affecting the individual. The wages of the day laborer are not only calculated on a daily basis, they are often paid daily also. Substantial reinforcement at shorter intervals is needed before payments spaced as much as a month apart are effective. To analyze such a history in detail we should have to investigate certain subsidiary kinds of behavior, some of them verbal, which are generated by schedules of reinforcement and which bridge the gap between working on, say, the first day of the month and being reinforced on the last. Such an analysis would have to include the effect of agreements or contracts between employer and employee.
In any case, however, wages received at fixed intervals do not parallel the intermittent reinforcements described in Chapter VI. In human behavior certain prominent stimuli, commonly correlated with the time of payment, make a temporal discrimination possible. The performance of a pigeon or rat under fixed-interval reinforcement changes dramatically when a stimulus is arranged to vary in some way with the passing of time between reinforcements. Clocks and calendars are verbal devices designed to supply stimuli of this sort to the human subject. When such stimuli are available, the worker—whether human or subhuman—waits until the reading on the clock is very close to that at which behavior is reinforced. If there were no other factors involved, payment for work at the end of each week would generate only a small amount of work just before pay-time.
It is necessary, therefore, to supplement fixed-interval schedules with other techniques of control. The supervisor or “boss” is a source of aversive stimulation contingent upon any behavior which falls below certain specifications, including a minimum rate of production. Some of the power available to the supervisor may be derived from his position in the ethical group—he may condemn laziness or poor work as bad or something to be ashamed of—but insofar as he “can’t do any worse than fire a man,” his main aversive stimulation is the threat of dismissal. Wages serve in such a case simply to create a standard economic condition which may be withdrawn aversively. The boss threatens dismissal, or some measure which is effective because it is a step toward dismissal, whenever the employee slows down; he removes that threat when the employee speeds up. Eventually the behavior of the employee generates comparable aversive stimulation; he works at a rate just above that at which he feels guilty or threatened. The use of an aversive boss is an excellent example of the general principle that when punishment is abandoned in favor of positive reinforcement, there is a tendency to turn to other forms of aversive control. The threat of withholding an accustomed positive reinforcement is always available for this purpose. Payment of wages is an obvious advance over slavery, but the use of a standard wage as something which may be discontinued unless the employee works in a given manner is not too great an advance.
A production line moving at a set rate makes the contingency between speed of work and aversive stimulation more clear-cut. This “pacing” of behavior is by no means a modern achievement. The galley slave pulled his oar to avoid the whip, which was contingent upon his failing to pull in unison with others. A line of reapers swinging scythes in unison paced each other—the basic rhythm being determined in part by a leader but also in part by the length and mass of the pendulum composed of man and scythe—because any deviation brought aversive stimulation, often dangerous, from the scythes of other reapers. The production line has the effect of reducing some of the personal attributes of aversive stimulation by a boss, but a danger inherent in any pacing system is the temptation on the part of the controller to increase the pace.
Combined schedules. Fixed-interval schedules are also supplemented in industry by various sorts of “incentive pay.” These are combinations of fixed-interval and fixed-ratio schedules. Each of the component schedules corrects some of the shortcomings of the other. Supplementary aversive stimulation from a supervisor is not needed if the ratio component is effective. At the same time the ratio component may not be enough to lead to dangerously high rates or long hours of work. When a salesman is paid partly on salary and partly on commission, the combination is designed to correct the abulia which might otherwise follow reinforcement at a high ratio.
Variable schedules. Laboratory studies have shown that variable-interval and variable-ratio schedules are superior to fixed schedules in sustaining performance, but it is not easy to adapt such schedules to the payment of wages. A contract between employer and employee which guarantees a given return, either per interval of time or per unit of work, rules out a genuine variable schedule. Such schedules may be used, however, in the payment of money—such as a bonus—not specified in a contract or contingent upon behavior in any other way. The bonus would usually be classed as an emotional variable which predisposes the individual favorably toward his work or his employer, but it may also act as a reinforcer. Its effect as such is considerably reduced if it is given on a fixed-interval schedule. The standard Christmas bonus, for example, eventually functions primarily as part of the pay which may be withdrawn as a form of aversive stimulation in dismissal. An unpredictable bonus, given in smaller sums on a variable-interval schedule but in approximately the same amount annually, would have a much greater effect.
Differential reinforcement of quality of work. Wages are usually contingent upon specified behavior at a specified level of quality or skill. In general the performance of an employee, like that of the laboratory animal, adjusts quite accurately to the exact contingencies of reinforcement. Both “do no more than they need to do.” Additional economic reinforcement may be made contingent upon work which exceeds minimum standards. Bonuses, raises, and promotions, when contingent upon exceptional performance, shape the topography of behavior in the direction of quality or skill (Chapter VI).
Extra-economic factors. It is now generally recognized that the employee seldom works “just for the money.” The employer who relies exclusively on economic control overlooks the fact that the average worker is reinforced in other ways. The individual craftsman not only constructs something which he can sell for money, he is reinforced by his success in dominating the medium in which he works and in producing an article for which he receives approval. These additional reinforcements may have a substantial effect in sustaining his level of work. They are often lost in mass-production methods in which the worker receives only an economic reinforcement for his achievement. To say that the craftsman is motivated by “pride in his work” does not help us to understand the problem. In order to deal effectively with the behavior of the employee we must in any given case be able to specify the actual circumstances which are reinforcing, and perhaps how they have come to be reinforcing.
The effect of the reinforcement of the worker is not shown in his rate of production if that rate is determined by an aversive pacing system. Extra-economic factors in industry usually have a more direct effect upon the behavior of the worker either in coming to work or in staying on one job. Quite apart from his rate of production while at work, the worker who “likes his job” shows little absenteeism and a history of few changes of employment. He likes his job in the sense that he is reinforced for coming to work—not only by an effective wage schedule but by the conditions under which he works, by his fellow workers, and so on. He dislikes his job insofar as it has aversive properties. If he is kept at a high level of work through constant aversive stimulation in the form of a threat of dismissal, the whole task will become aversive and, when his economic condition permits, he will remain absent or, if possible, change jobs. Conditioned aversive stimuli associated with sickness, unemployment, or hardship in old age may also have important aversive effects. It does not help much in dealing with these problems to say that the employee wants “freedom” or “security.” In the design of optimal working conditions, considered with respect not only to productivity but to absenteeism and labor turnover, we need an explicit analysis of actual reinforcing and aversive events.
That part of the behavior of the worker which is under economic control generates aversive stimuli—from the nature of the work itself or from the fact that it prevents the worker from engaging in activities which would be reinforcing in other ways. These aversive consequences are roughly offset by the economic reinforcement which the worker receives. When the worker accepts or rejects the offer of a job, he may be said to be comparing positive and negative reinforcers. A similar comparison is made by the employer. Since those who use economic control must give up the goods or money with which they reinforce behavior, economic reinforcement is by definition aversive to the controller.
If these conflicting consequences are roughly equal, the individual may engage in behavior leading to a decision in the sense of Chapter XIV. Shall a man mow his own lawn or pay someone else to mow it for him? This will depend in part upon the aversive properties of mowing the lawn and the aversive properties of giving up the money needed to hire someone to mow it. It will also depend upon the behavior of making a decision in which the man may review other possible consequences of mowing the lawn himself—the exercise may be good for him—or the kinds of things for which the money which must be paid could otherwise be exchanged, or ways in which he might earn that amount of money less aversively than by mowing a lawn, and so on. The prospective employee may alter similar conditions affecting his behavior in accepting or rejecting an offer.
A “deal” is made in such a case if in avoiding the aversive consequences of mowing the lawn, the employer offers an amount equal to or greater than that which matches the aversive consequences to the employee. The amount offered will also depend upon the aversive consequences of giving up money. The amount offered by the employer is what the job is “worth” to him in his current economic circumstances; the amount accepted by the employee is what the job is “worth” to him in his current economic circumstances.
The “economic value” of labor or other personal services thus has to do with the matching of positive and negative reinforcing effects. The reinforcing effects of two tasks could be directly compared, but money provides a single scale on which the economic values of many different types of labor or services may be represented. We have already seen that money has certain advantages as a generalized reinforcer; it has fairly simple dimensions, it can be made contingent upon behavior in a clear-cut way, and its effects are relatively free of the momentary condition of the organism. Money has a special advantage in representing economic value because different amounts can be compared on a single scale; one amount may be equal to another, twice as great as another, and so on. This standard scale is so effective in comparing reinforcers that it is often taken to represent some sort of independent economic value not associated with positive or negative consequences. The monetary scale is regarded as a primary dimension of value. But the scale would have no meaning apart from the comparison of other consequences.
To the employer the economic value of labor is just that amount of money which he will give up in return for that labor. This depends upon the results of the labor. We pay a man for mowing a lawn if a mowed lawn is reinforcing. We pay him for making shoes if shoes are personally reinforcing or can be exchanged for money or goods which are reinforcing for other reasons. Sometimes behavior itself is directly reinforcing, as in entertainment; we have seen that the entertainer is in the business of making his behavior positively reinforcing so that it will have economic value.
To the employee the economic value of labor is just that amount of money for which he will supply that labor. The aversive consequences against which he places a value upon his services may be of many sorts. Hard labor is directly aversive, as is confinement at a given task for long periods of time regardless of the energy required. Some tasks are aversive for special reasons. Thorndike found that people were in general willing to name a price for engaging in a wide variety of aversive tasks—such as letting a snake coil around one’s arms and head, eating a dead earthworm, or spitting on a picture of George Washington. Money which is paid for behavior which, although not especially aversive in itself, may possibly lead to punishment, is usually called a bribe. The bribe supplies a measure of the economic value of a given probability of punishment.
Behavior has “nuisance value” when a man is paid for not engaging in it. When a solicitous parent gives an allowance to his son so long as he does not smoke or drink or marry before a given age, the behavior which the son foregoes may have substantial reinforcing properties for him. He “earns” his allowance by accepting the aversive consequences of giving up the stipulated reinforcements. When the behavior which is given up has no substantial reinforcing consequences but would be highly aversive to the man who pays to suppress it, the money paid is referred to as blackmail. When the behavior is verbal—for example, testifying to or otherwise reporting censurable behavior—it is commonly called hush money. A similar controlling relation is exploited by the underworld gang which sells “protection”—in other words, agrees not to damage person or property in return for payment. Blackmail and protection represent unstable social systems in the sense of Chapter XIX. Such control is opposed by the ethical group or by religious and governmental agencies which make aversive consequences contingent upon engaging in such transactions.
Buying and selling or exchanging in barter are so commonplace that we are likely to overlook several of the processes involved. The basic transaction or “deal” is expressed by the offer, “I will give you this if you will give me that.” As in transactions involving personal labor, such complex stimuli are effective only after extensive economic conditioning. The process is easy to observe as a child learns to swap toys with his playmates or to buy penny candy at the corner store. Before such behavior reaches a relatively stable state, the child must be affected by the full aversive consequences of giving up a toy or a penny and by the reinforcing consequences of obtaining another toy or candy. When such conditioning has taken place, similar behavior with similar objects and similar money may become relatively automatic, and it may be easy to overlook the complex relationships involved. Whether a sale is made quickly or after long deliberation depends upon whether the aversive properties of giving up money or going without the object are matched by the positively reinforcing properties of the money or the object. In “a good bargain” the object bought is more highly reinforcing than the money given up, and the sale takes place quickly. In the doubtful bargain, positive and negative consequences are relatively evenly matched, and the sale may take place only after long deliberation.
The economic value of goods. The use of money in buying and selling permits us to evaluate goods as we evaluated labor—on a simple one-dimensional scale. An object is “worth” to an individual just that amount of money which he will give up in exchange for it, or in exchange for which he will give it up. Before an exchange or a sale can occur, certain critical values must be reached or exceeded. A will give the article to B if the aversive consequences of this act are roughly matched by the positively reinforcing consequences of the money which B will give to A. B will give this amount of money to A if the aversive consequences which are thus involved are matched by the positively reinforcing consequences of receiving the article from A.
Several other conditions affect economic transactions. Since the money which a man will give in exchange for goods is a measure of the reinforcing effect of the goods, it will vary with the level of deprivation. The value which a man assigns to food depends upon how hungry he is. By keeping food in short supply he may be induced to pay a high price. In the population as a whole this is reflected by the fact that the price commonly paid for an object can be manipulated by manipulating the supply. But how much a man will pay for food also depends upon the aversive consequences of giving up money, and this depends roughly upon how much money he has. If “money is no object,” he may pay a high price. In the population as a whole the price of an object will therefore be determined in part by the supply of money. These two factors, the supply of goods and the supply of money, have, of course, a prominent place in traditional economic theory. They are not, however, the only determiners of economic transactions.
An important consideration is the history of reinforcement of the behavior of acquiring or giving up goods or money. The behavior of buying or selling may be strengthened or weakened apart from the particular nature of a given transaction. When the reinforcing consequences to the buyer greatly exceed the aversive consequences of giving up the price of an article, the simple behavior of buying is strengthened. In the technique of the bargain store some objects are sold at a low price so that others, which are not bargains, can also be sold. The “buying habits” of the public often reflect the same principle. Whether an individual readily engages in buying also depends in part upon previous aversive consequences of giving up money. “Learning the value of a dollar” is the effect of the aversive consequences of parting with a dollar.
The reinforcing effect of an article, and hence the price which can be obtained for it, is enhanced by many techniques of merchandising. The article is made “attractive” by design, packaging, and so on. Properties of this sort make an object reinforcing as soon as it is seen by the prospective purchaser, so that a previous history with similar objects is not required.
Imitative behavior is relevant in buying and selling. An object may be bought simply because other people are buying objects of the same kind. This is the principle of the bargain crush and the public spending spree. Testimonial advertising sets up imitative patterns for the potential buyer by portraying other buyers or possessors of goods. Imitative nonbuying is characteristic of periods of deflation.
The balancing of positive and negative consequences may be offset by altering the time which elapses between these consequences and behavior. Sales are encouraged by promises of immediate delivery. The same effect is felt, in the absence of an agreement, when a mail order house by filling its orders as rapidly as possible gains an advantage over a rival house with a longer average delivery time. The behavior of mailing in an order is probably not, strictly speaking, reinforced by the receipt of goods after, say, four days; any reinforcing effect of such a consequence must be mediated by verbal or nonverbal intervening steps. But these intervening steps need not change the advantage gained by reducing the time which passes between the behavior and the ultimate consequence. Another kind of time relation is manipulated when the purchaser is permitted to buy on credit. In buying on the installment plan, the aversive consequences of giving up the purchase price are postponed and distributed. The effect is to be distinguished from the effect of credit in permitting goods to be purchased before money is available.
Another important factor contributing to the probability that an individual will turn over money, either for other money or for goods, is the schedule on which he is reinforced for doing so. A faulty vending machine or dishonest vendor occasionally fails to complete the exchange of goods for money. The probability of engaging in transactions under similar circumstances is to some extent reduced through extinction. However, if a vendor characteristically offers an especially good bargain whenever the transaction is completed, the probability may remain at a significant value. In general, the greater the reinforcing effect of the object exchanged for money, the more often reinforcements may fail without extinguishing the behavior altogether. This is an example of the type of economic interchange called gambling.
One may gamble with money for money, as in playing a roulette wheel or slot machine; with money for goods, as in buying a chance on an automobile; or with goods for money, as in playing a customer double or nothing for the bill. The behavior of the gambler is under very complex control depending upon his history of reinforcement. It is sometimes possible to calculate the “chances” of a given gambling system, and these, if known to the gambler, may determine whether he will place a bet or not. How the probability that a man will place a bet of a given size varies with such factors as the size of the stake or a given history of reinforcement can be studied experimentally. The predisposition to continue betting under a given system, however, depends primarily upon the schedule of reinforcement. Gambling devices in general observe a variable-ratio schedule. From the point of view of the gambling establishment this is a safe schedule because the percentage profit in the long run is fixed. It is also an unusually effective schedule in generating gambling behavior. The gambling establishment selects a mean ratio which is a compromise between two consequences. Too high a ratio yields a large mean profit per play but a loss of patronage. Too low a ratio yields too small a profit in spite of a ready patronage. The professional gambler “leads his victim on” by building a favorable history of reinforcement. He begins with a low mean ratio under which reinforcement occurs so frequently that the victim wins. The mean ratio is then increased, either slowly or rapidly depending upon how long the gambler plans to work with a particular victim. This is precisely the way in which the behavior of a pigeon or rat is brought under the control of a variable-ratio schedule. A mean ratio can be reached at which reinforcements occur so rarely that the pigeon or rat spends more energy in operating the device than he receives from the reinforcement with food, while the human subject steadily loses money. All three subjects, however, continue to play.
Gambling devices make an effective use of conditioned reinforcers which are set up by pairing certain stimuli with the economic reinforcers which occasionally appear. For example, the standard slot machine reinforces the player when certain arrangements of three pictures appear in a window on the front of the machine. By paying off very generously—with the jack pot—for “three bars,” the device eventually makes two bars plus any other figure strongly reinforcing. “Almost hitting the jack pot” increases the probability that the individual will play the machine, although this reinforcer costs the owner of the device nothing.
Gambling, then, is a system of economic control in which the individual is induced to pay money in return for a reinforcement the value of which is too small to lead to exchange under other schedules. If a man cannot sell a car to one man for $3,000, he may still sell it to 3,000 men for $1.00 if the culture has provided the necessary history of variable-ratio reinforcement when its members have “taken chances.” If the gambling establishment cannot persuade a patron to turn over money with no return, it may achieve the same effect by returning part of the patron’s money on a variable-ratio schedule.
In summary, then, the probability that a transaction will take place is a function of the levels of deprivation of buyer and seller with respect to goods and money, upon the history of both participants with respect to good and bad bargains, upon the temporary characteristics of the object or the situation involved in merchandising, upon the behavior of others engaged in similar transactions, upon the temporal contingencies which govern the receipt of goods or the giving up of money, and upon a history of certain schedules of reinforcement. All these conditions follow from an analysis of human behavior; they are also familiar features in traditional discussions of economic behavior. They obviously affect the usefulness and precision of the concept of economic value. The reinforcing effect of either goods or money cannot be stated without taking into account many different characteristics of the history of the individual buyer or seller, as well as the external circumstances under which a given economic transaction takes place.
When millions of people engage in buying and selling, lending and borrowing, renting and leasing, and hiring and working, they generate the data which are the traditional subject matter of the science of economics. The data include the quantities and locations of goods, labor, and money, the numbers of economic transactions in a given period, certain characteristics of transactions expressed as costs, prices, interest rates, and wages, together with changes in any of these as functions of time or other conditions.
Statements about goods, money, prices, wages, and so on, are often made without mentioning human behavior directly, and many important generalizations in economics appear to be relatively independent of the behavior of the individual. A reference to human behavior is at least implied, however, in the definition of all key terms. Physical objects are not goods apart from their reinforcing value. More obviously, money cannot be defined without reference to its effect upon human behavior. Although it may be possible to demonstrate valid relationships among the data generated by the economic transactions of large numbers of people, certain key processes in the behavior of the individual must be considered. The traditional procedure has been to deduce the behavior of the individual engaging in economic transactions from the data derived from the group. This procedure led to the Economic Man of nineteenth-century economic theory, who was endowed with just the behavior needed to account for the over-all facts of the larger group. This explanatory fiction no longer plays a prominent role in economic theorizing.
Some attention to the individual transaction is often required when generalizations at the level of the group prove invalid. We have already noted many special conditions which affect economic value. In the data generated by millions of people the effects of these special conditions may strike an average or cancel each other out. But when a given condition holds for a large number of people, it cannot be disposed of in this way. Economists frequently explain the failure to predict a particular consequence from a broad generalization by appealing to special conditions of this sort. Although the supply of money and goods may suggest inflation, for example, some external condition, not otherwise related to the supply of money or goods, may generate undue caution on the part of a large number of buyers. If the science of economics were to take all such extra-economic variables into account, it would become a complete science of human behavior. But economics is concerned with only a small number of the variables of which the behavior of the individual is a function. There are many practical reasons why this limited area needs to be studied in relative isolation. This means that the economist will always need to appeal from time to time to the behavior of the real economic man.
Economic theory has been especially inclined to use the principle of maxima and minima. The freedom, justice, and security of the governmental agency, the salvation and piety of the religious agency, and the mental health and adjustment of psychotherapy have their parallels in “wealth,” “profits,” “utility,” and many other concepts in terms of which economic transactions have been evaluated. Since quantification is encouraged in economic theorizing by the useful dimensions of money as a generalized reinforcer, it may appear that these entities are more easily adapted to a functional analysis. But it has not been shown that they are, in fact, any more useful in predicting or controlling a given economic transaction than are their counterparts in the other fields. The conception of economic behavior which emerges from a functional analysis offers an alternative possibility. The present chapter has, of course, dealt with only a small fraction of the many kinds of economic transactions to be observed in any large group of people, but an adequate science of behavior should supply a satisfactory account of the individual behavior which is responsible for the data of economics in general.
The power to wield economic control naturally rests with those who possess the necessary money and goods. The economic agency may consist of a single individual, or it may be as highly organized as a large industry, foundation, or even government. It is not size or structure which defines the agency as such, but the use to which the economic control is put. The individual uses his wealth for personal reasons, which may include the support of charities, scientific activities, artistic enterprises, and so on. The eleemosynary foundation is engaged in disposing of wealth in support of specified activities. Religious and governmental agencies frequently, as we have seen, use this supplementary technique for their special purposes.
If there is any special economic agency as such, it is composed of those who possess wealth and use it in such a way as to preserve or increase this source of power. Just as the ethical group is held together by the uniformity of the aversive effect of the behavior of the individual, so those who possess wealth may act together to protect wealth and to control the behavior of those who threaten it. To that extent we may speak of the broad economic agency called “capital.” The study of such an agency requires an examination of the practices which represent concerted economic control and of the return effects which support these practices.
As in religious, governmental, or psychotherapeutic control, economic power may be used to further the special interests of those who possess it. Excessive control generates behavior on the part of the controllee which imposes a practical limit. The group as a whole usually condemns the excessive use of wealth as bad or wrong, and classifies the charitable use of wealth as good or right. Some countercontrol is also exerted by religious and governmental agencies. Under most modern governments, for example, the individual cannot legally control many sorts of behavior through economic power. Laws concerning prostitution, child labor, fraudulent practices, gambling, and so on all impose limits. Particular economic transactions are restricted, or rendered more or less probable, by tariffs, levies, taxes on profits and on transactions, price controls, changing the supply of money, government spending, and so on. All these measures alter the balance between those possessing labor or goods and those possessing money; hence they alter the frequency with which certain kinds of economic transactions take place. The effect is usually to reduce the extent to which the possessor of wealth is able to employ it in controlling others.