哈佛大学：《财产法的经济分析》课程教学资源（教案讲义，英文版）ECONOMIC ANALYSIS OF PROPERTY LAW.pdf_大学文库

Chapter 7 DEFINITION JUSTIFICATION AND EMERGENCE OF PROPERTY RIGHTS In this chapter, I first define property rights, then address the question of their justification, that is, the advantages that might be thought to explain their existence, and last discuss several instances of their emergence 1. Property rights Defined I will use the term"property rights" to refer broadly to two subsidiary types of rights possessory rights and rights of transfer 1.1 Possessory rights. What are often called possessory rights allow individuals to use things and to prevent others from using them. a particular possessory right is a right to commit a particular act or a right to prevent others from committing a particular On reflection, it can be seen that common property arrangements entail considerable agglomerations of possessory rights, but with certain limitations. The basic notion of ownership embraces a large bundle of possessory rights subject to exceptions When we say that a person owns a parcel of land, we ordinarily mean that the person can do virtually as he pleases on the land (plant crops on it, build on it, leave it idle)over time and under most contingencies. Also included in the idea of ownership is the right of the owner to prevent others from using what he owns over time and in most contingencies However, the owner's possessory rights are not absolute. If a use harms others (maintaining a compost heap that produces noxious odors), it may be proscribed; or, under certain conditions, an owner may be compelled to allow others to use his things(in an emergency, someone might be permitted to seek shelter on his land) Likewise, under rental provisions, the renter of property generally enjoys a large swath of possessory rights, most of those associated with ownership. For example someone who rents an apartment can use it more or less as he pleases and bar the entry of others, even generally the landlord. But there may be limitations on the renter's possessory rights that would not be faced by an owner-occupier; typically, a renter of an apartment is forbidden from making major modifications to it Another example worth noting is property arrangements within enterprises. When person is working for an enterprise, he will often enjoy certain possessory rights. A person may be able to do a variety of things in the office to which he is assigned and exclude others from it under a fairly wide range of circumstances Here sh- The subject of the enforcement of these and other rights will be addressed generally in Part Five all simply assume that property rights are upheld 2A completely specified act includes in its description the place, time, and contingency under which it is committed--for example, digging at a designated location, on Thursday at 4 PM, if it is not raining(the contingency) Chapter 7-Page I

Chapter 7 DEFINITION, JUSTIFICATION, AND EMERGENCE OF PROPERTY RIGHTS In this chapter, I first define property rights, then address the question of their justification, that is, the advantages that might be thought to explain their existence, and last discuss several instances of their emergence. 1. Property Rights Defined I will use the term “property rights” to refer broadly to two subsidiary types of rights, possessory rights and rights of transfer. 1.1 Possessory rights. What are often called possessory rights allow individuals to use things and to prevent others from using them. 1 A particular possessory right is a right to commit a particular act or a right to prevent others from committing a particular act.2 On reflection, it can be seen that common property arrangements entail considerable agglomerations of possessory rights, but with certain limitations. The basic notion of ownership embraces a large bundle of possessory rights subject to exceptions. When we say that a person owns a parcel of land, we ordinarily mean that the person can do virtually as he pleases on the land (plant crops on it, build on it, leave it idle) over time and under most contingencies. Also included in the idea of ownership is the right of the owner to prevent others from using what he owns over time and in most contingencies. However, the owner=s possessory rights are not absolute. If a use harms others (maintaining a compost heap that produces noxious odors), it may be proscribed; or, under certain conditions, an owner may be compelled to allow others to use his things (in an emergency, someone might be permitted to seek shelter on his land). Likewise, under rental provisions, the renter of property generally enjoys a large swath of possessory rights, most of those associated with ownership. For example, someone who rents an apartment can use it more or less as he pleases and bar the entry of others, even generally the landlord. But there may be limitations on the renter=s possessory rights that would not be faced by an owner-occupier; typically, a renter of an apartment is forbidden from making major modifications to it. Another example worth noting is property arrangements within enterprises. When a person is working for an enterprise, he will often enjoy certain possessory rights. A person may be able to do a variety of things in the office to which he is assigned and exclude others from it under a fairly wide range of circumstances. 1 The subject of the enforcement of these and other rights will be addressed generally in Part Five. Here I shall simply assume that property rights are upheld. 2 A completely specified act includes in its description the place, time, and contingency under which it is committed -- for example, digging at a designated location, on Thursday at 4 PM, if it is not raining (the contingency). Chapter 7 - Page 1

In strict logic, and for some purposes, it will be helpful to reduce such common property arrangements to their more basic elements, including the arrays of possessory rights that they encompass 1.2 Rights of transfer. The other type of right associated with the notion of property rights is a right to transfer a possessory right, that is, the option of a person who holds a possessory right to give it to another person(usually, in exchange for something) Closely related, but distinct, is the right of the recipient subsequently to transfer his possessory right to another person(and for that person to do the same, and so forth). We will assume unless otherwise noted that rights to transfer possessory rights implicitly include these subsequent-order rights to transfer It is apparent that common property arrangements entail certain rights to transfer possessory rights. The concept of ownership incorporates not only possessory rights, but also rights to transfer these possessory rights; an owner is usually presumed to be able to sell or give away his property, in which case the acquirer obtains all the possessory rights held by the owner, as well as the rights to transfer these rights. Under rental arrangements, however, the renter of a thing may or may not enjoy the right to transfer his possessory rights to another. a person who rents an apartment may be allowed to rent to another( to"sublet), or he may be prevented from doing this according to his rental agreement As a general matter, if we want to be precise about the meaning of property rights, we should describe an arrangement in terms of the entailed possessory rights and rights to ransfer 2. Justifications for Property Rights 2.1 The general question of justification. a time-honored and fundamental question is why there should be any property rights in things. That is, in what respects does the protection of possessory interests in things and the ability to transfer them promote social welfare, broadly construed? I now consider a list of factors suggesting that the existence of property rights fosters social welfare 2.2 Incentives to work. It is often said that property rights provide individuals with incentives to work, and it is worth examining this argument in some detail. For this purpose, consider initially a stylized model in which individuals produce a good and in which the measure of social welfare is the utility from the good less the disutil ity of work. The determination of the socially optimal amount of work is plain in the model. It is best for an individual to work an additional hour if and only if the increment in utility from consuming what would be produced in that hour exceeds the disutility of work from the hour Example 1. Individuals can work either 0, 1, 2, or 3 hours, and each hour an individual works he produces one unit of output. The total utility he derives from output, the total disutility of work, and social welfare per individual (utility from output minus disutility rom work) are as shown in Table 1 Chapter 7-Page 2

In strict logic, and for some purposes, it will be helpful to reduce such common property arrangements to their more basic elements, including the arrays of possessory rights that they encompass. 1.2 Rights of transfer. The other type of right associated with the notion of property rights is a right to transfer a possessory right, that is, the option of a person who holds a possessory right to give it to another person (usually, in exchange for something). Closely related, but distinct, is the right of the recipient subsequently to transfer his possessory right to another person (and for that person to do the same, and so forth). We will assume unless otherwise noted that rights to transfer possessory rights implicitly include these subsequent-order rights to transfer. It is apparent that common property arrangements entail certain rights to transfer possessory rights. The concept of ownership incorporates not only possessory rights, but also rights to transfer these possessory rights; an owner is usually presumed to be able to sell or give away his property, in which case the acquirer obtains all the possessory rights held by the owner, as well as the rights to transfer these rights. Under rental arrangements, however, the renter of a thing may or may not enjoy the right to transfer his possessory rights to another. A person who rents an apartment may be allowed to rent to another (to “sublet”), or he may be prevented from doing this according to his rental agreement. As a general matter, if we want to be precise about the meaning of property rights, we should describe an arrangement in terms of the entailed possessory rights and rights to transfer. 2. Justifications for Property Rights 2.1 The general question of justification. A time-honored and fundamental question is why there should be any property rights in things. That is, in what respects does the protection of possessory interests in things and the ability to transfer them promote social welfare, broadly construed? I now consider a list of factors suggesting that the existence of property rights fosters social welfare. 2.2 Incentives to work. It is often said that property rights provide individuals with incentives to work, and it is worth examining this argument in some detail. For this purpose, consider initially a stylized model in which individuals produce a good and in which the measure of social welfare is the utility from the good less the disutility of work. The determination of the socially optimal amount of work is plain in the model. It is best for an individual to work an additional hour if and only if the increment in utility from consuming what would be produced in that hour exceeds the disutility of work from the hour. Example 1. Individuals can work either 0, 1, 2, or 3 hours, and each hour an individual works he produces one unit of output. The total utility he derives from output, the total disutility of work, and social welfare per individual (utility from output minus disutility from work) are as shown in Table 1. Chapter 7 - Page 2

Table 1: Hours, Output, and Social Welfare Hours of work Output Utility from output Disutility from work Social welfare 0 0 0 0 0 1 1 10 6 4 2 2 18 13 5 3 3 24 22 2 Social welfare is maximized if each individual works two hours. Working the first hour is beneficial since it augments utility by 10 and involves disutility of only 6; working a second hour raises an individual’s welfare further because it increases utility by 8 and increases disutility by only 7; but working a third hour would increase utility by only 6 and involve greater disutility of 9.// Socially optimal amount of work is not performed in the absence of property rights. If there are no property rights, individuals will tend not to work the socially optimal amount and social welfare will be less than optimal in the model just described because of problems with individuals’ incentives to work. In a situation in which property rights do not exist, an individual will take into account that his output may be taken from him. In deciding whether to work an extra hour, an individual will compare the increment in utility from consuming the additional amount of his production that he will be able to keep -- rather than the whole of the additional amount of his production -- to the disutility of work from the hour. When the individual makes this determination, he may well decide to work less than optimally because he will enjoy less utility from the smaller amount of output that he will retain for his own use. Let us illustrate this point. Example 2. Modify Example1 by assuming that each individual will lose half of what he produces to others who can take that amount. Then each individual will choose not to work at all, for the situation facing each individual will be as follows.3 Table 2: Hours and Output in the Absence of Property Rights Hours of work Retained output Utility from output Disutility from work Individual welfare 0 0 0 0 0 1 .5 5 6 !1 2 1 10 13 !3 3 1.5 14 22 !8 3 The utilities for outputs of .5 and 1.5 are assumed to be as displayed and are consistent with the previous table. Chapter 7 - Page 3

Here an individual will not even work the first hour because he will be able to keep and consume only 5 units of output, and therefore enjoy utility of only 5, which is less than the disutility of work of 6, and so forth /l In this example, individuals decide not to work at all, so that social welfare is zero. If, however, the example is altered and individuals lose a smaller fraction of their output, they will work a positive amount. This also means that a positive amount would be taken from them, which raises two complications. First, whoever takes output will have it to consume - in other words what is taken will still contribute to social welfare Second the incentives to work of whoever takes from others have to be reexamined. Bu neither of these complications changes the conclusion that the amount of work and the ut level of social welfare will generally deviate from the optimal Another point to be noted in the absence of property rights is that it is possible that individuals will choose to work more than the optimal number of hours, not less Suppose that having a certain minimum amount of output to consume is important for subsistence. Then, to guarantee that he will retain this amount after others have taken some output from him, an individual might well be inclined to work more than he otherwise would. Although total output Suppose, for instance, that the individuals in the model are identical to one another and each has the opportunity not only to work and produce, but also freely to take a fraction of the output of another person. Then an individual's incentives to work will remain undesirable, for it will still be true that in deciding whether to work an extra hour, the individual will know that he will not be able to keep the full amount of what he produces in the hour Formally, we can show that individuals will work too little under certain assumptions. Let a be the fraction of the output that is taken, w be hours of work and also output produced, u( w) be utility from utput consumed, where u/ (w)>0 and u(w)0 and d(w >0. It is socially optimal for w to maximize u)-dow), implying that the optimal w satisfies u(w) d (w): let w* be the optimal w. If a fraction a is taken from individuals, however, then if w is the work the identical individuals generally choose, each particular individual will be able to take aw from another Hence, each individual will select his w to maximize u(aw +(1-a)w)-dw), implying that(1-a)u(at +(1 - w)=d(w). But in equilibrium, w=w so that we must have(1-a)u(w)=d (w ). This means that w <w*(implicit differentiation of (1-a)u (w)=d(w)with respect to a shows that w is decreasing n a, and w* corresponds to a=0) This possibility, that individuals might work more in the absence of property rights than in their presence, is occasionally -and mistakenly--interpreted as an argument against property rights. As I am about to explain, when individuals do work more in the absence of property rights, social welfare still tends to be lower than in the presence of property rights. The issue in question is whether the absence of property rights affects work incentives in such a way as to reduce social welfare, not whether work effort and production themselves fall or rise. Consider the foll situation Hours of work Output Utility from output Disutility from Social welfare 0 Chapter 7-Page 4

Here an individual will not even work the first hour because he will be able to keep and consume only .5 units of output, and therefore enjoy utility of only 5, which is less than the disutility of work of 6, and so forth.// In this example, individuals decide not to work at all, so that social welfare is zero. If, however, the example is altered and individuals lose a smaller fraction of their output, they will work a positive amount. This also means that a positive amount would be taken from them, which raises two complications. First, whoever takes output will have it to consume -- in other words, what is taken will still contribute to social welfare. Second, the incentives to work of whoever takes from others have to be reexamined. But neither of these complications changes the conclusion that the amount of work and the level of social welfare will generally deviate from the optimal.4 Another point to be noted in the absence of property rights is that it is possible that individuals will choose to work more than the optimal number of hours, not less.5 Suppose that having a certain minimum amount of output to consume is important for subsistence. Then, to guarantee that he will retain this amount after others have taken some output from him, an individual might well be inclined to work more than he otherwise would.6 Although total output 4 Suppose, for instance, that the individuals in the model are identical to one another and each has the opportunity not only to work and produce, but also freely to take a fraction of the output of another person. Then an individual=s incentives to work will remain undesirable, for it will still be true that in deciding whether to work an extra hour, the individual will know that he will not be able to keep the full amount of what he produces in the hour. Formally, we can show that individuals will work too little under certain assumptions. Let " be the fraction of the output that is taken, w be hours of work and also output produced, u(w) be utility from output consumed, where u'(w) > 0 and u''(w) 0 and d''(w) > 0. It is socially optimal for w to maximize u(w) ! d(w), implying that the optimal w satisfies u'(w) = d'(w); let w* be the optimal w. If a fraction " is taken from individuals, however, then if we is the work the identical individuals generally choose, each particular individual will be able to take "we from another. Hence, each individual will select his w to maximize u("we + (1 ! ")w) ! d(w), implying that (1 ! ")u'("we + (1 ! ")w) = d'(w). But in equilibrium, w = we , so that we must have (1 ! ")u'(we ) = d'(we ). This means that we < w* (implicit differentiation of (1 ! ")u'(we ) = d'(we ) with respect to " shows that we is decreasing in ",and w* corresponds to " = 0). 5 This possibility, that individuals might work more in the absence of property rights than in their presence, is occasionally -- and mistakenly -- interpreted as an argument against property rights. As I am about to explain, when individuals do work more in the absence of property rights, social welfare still tends to be lower than in the presence of property rights. The issue in question is whether the absence of property rights affects work incentives in such a way as to reduce social welfare, not whether work effort and production themselves fall or rise. 6 Consider the following situation. Hours of work Output Utility from output Disutility from work Social welfare 0 0 0 0 0 1 1 40 2 38 2 2 45 10 35 Chapter 7 - Page 4

will be higher when individuals in the model work more than is optimal, social welfare will tend to be lower than optimal This is for two reasons: The hours that individuals work in excess of the optimal number may be hours for which the disutility of work is high(the disutility of work may grow as the number of hours left in the day diminishes) the extra output may be unevenly distributed, contributing relatively little to an individual's utility if he already has substantial output Socially optimal outcome is achievable under property rights: where each individual has rights in his own output. If each individual has possessory rights in the output he produces, he will work the socially optimal amount and social welfare will be maximized. This is one of the classic arguments for the desirability of property rights. In particular, an individual with possessory rights will know that he will be able to consume what he produces and thus he will compare the utility from the output he produces by orking an extra hour to the disutility of work effort. In Example 1, individuals will hoose to work two hours, the optimal amount. another way of expressing why social optimality will result is that the goal of each individual will coincide with the social goal Socially optimal outcome is achievable under property rightS: where a supervisory entity has rights in an individuals output. a different regime in which the socially optimal outcome may result involves an entity that enjoys possessory rights individuals'output and that can supervise individuals'work. Such a supervisory entit might, for example, be a single-person owner of a farm. If a supervisory entity can monitor individuals'work, then, through use of appropriate rewards or punishments, it may be able to ensure that they work the optimal amount. Whether a supervisory entitys incentives will lead it to choose the optimal outcome, and how a regime with supervisory entities compares to a regime with individuals as possessors of property rights in their Notice here that an individual gains most of his utility from the first unit of output (interpreted as that of what they produce, so that it would take two hours to obtain one unit of output, an individual woul talf necessary for subsistence)and that working one hour is socially optimal. But if individuals would lose work two hours: his welfare would then be 40-10=30. whereas if he would work one hour his welfare would be only 20-2=18(assuming the utility from 5 unit of output is 20) IThis is the case in the example of the previous note, where the second hour of work creates output that would be worth only 5 but involves extra disutility of 8 STo demonstrate the latter possibility, let us assume that individuals are identical and that there is a probability p that each person will lose a fraction a of his output and an equal probability of the independent event that he will obtain a of someone else's output. Then(using the notation from note 4)an individual's expected utility will be Pu p)u(w)-dw); the terms correspond to the events that he loses a of his output and gains a of another person's, that he does not lose any of his but gains a of another persons, and so forth. Because he will select w to maximize his expected utility, and because w =w in equilibrium, we obtain the equilibrium conditionp(I-au(w)+(1- p)pu(1+a)+ (1-a)(1-p)pu((1-a))+(1-p)(w)=d (w). Now the third term on the left side of this condition can be arbitrarily higher than u (w), for(1-a)ww"*is possible; that is, the equilibrium may be such that individuals work more and produce more than is optimal. They are also worse off than in the optimal situation. To see explicitly why this is so, note hat, as is readily shown, if they work w and retain their entire output for sure, individuals are better off than in the risky situation, but we know that when they retain their output, they are better off working w* Chapter 7-Page 5

will be higher when individuals in the model work more than is optimal, social welfare will tend to be lower than optimal. This is for two reasons: The hours that individuals work in excess of the optimal number may be hours for which the disutility of work is high (the disutility of work may grow as the number of hours left in the day diminishes);7 the extra output may be unevenly distributed, contributing relatively little to an individual=s utility if he already has substantial output.8 Socially optimal outcome is achievable under property rights: where each individual has rights in his own output. If each individual has possessory rights in the output he produces, he will work the socially optimal amount and social welfare will be maximized. This is one of the classic arguments for the desirability of property rights. In particular, an individual with possessory rights will know that he will be able to consume what he produces and thus he will compare the utility from the output he produces by working an extra hour to the disutility of work effort. In Example 1, individuals will choose to work two hours, the optimal amount. Another way of expressing why social optimality will result is that the goal of each individual will coincide with the social goal. Socially optimal outcome is achievable under property rights: where a supervisory entity has rights in an individual’s output. A different regime in which the socially optimal outcome may result involves an entity that enjoys possessory rights in individuals’ output and that can supervise individuals’ work. Such a supervisory entity might, for example, be a single-person owner of a farm. If a supervisory entity can monitor individuals’ work, then, through use of appropriate rewards or punishments, it may be able to ensure that they work the optimal amount. Whether a supervisory entity’s incentives will lead it to choose the optimal outcome, and how a regime with supervisory entities compares to a regime with individuals as possessors of property rights in their Notice here that an individual gains most of his utility from the first unit of output (interpreted as that necessary for subsistence) and that working one hour is socially optimal. But if individuals would lose half of what they produce, so that it would take two hours to obtain one unit of output, an individual would work two hours: his welfare would then be 40 ! 10 = 30, whereas if he would work one hour, his welfare would be only 20 ! 2 = 18 (assuming the utility from .5 unit of output is 20). 7 This is the case in the example of the previous note, where the second hour of work creates output that would be worth only 5 but involves extra disutility of 8. 8 To demonstrate the latter possibility, let us assume that individuals are identical and that there is a probability p that each person will lose a fraction " of his output and an equal probability of the independent event that he will obtain " of someone else’s output. Then (using the notation from note 4) an individual’s expected utility will be p2 u((1 ! ")w + "we ) + (1 ! p)pu(w + "we ) + p(1 ! p)u((1 ! ")w) + (1 ! p) 2 u(w) ! d(w); the terms correspond to the events that he loses " of his output and gains " of another person’s, that he does not lose any of his but gains " of another person’s, and so forth. Because he will select w to maximize his expected utility, and because w = we in equilibrium, we obtain the equilibrium condition p2 (1 ! ")u'(we ) + (1 ! p)pu'((1 + ")we ) + (1 ! ")(1 ! p)pu'((1 ! ")we ) + (1 ! p) 2 u'(we ) = d'(we ). Now the third term on the left side of this condition can be arbitrarily higher than u'(we ), for (1 ! ")we w* is possible; that is, the equilibrium may be such that individuals work more and produce more than is optimal. They are also worse off than in the optimal situation. To see explicitly why this is so, note that, as is readily shown, if they work we and retain their entire output for sure, individuals are better off than in the risky situation; but we know that when they retain their output, they are better off working w* than we . Chapter 7 - Page 5

output, are, of course, different questions. The point here is simply that it is possible for the optimal outcome to be achieved if a supervisor has information about individuals work effort and possessory rights in output Comments. (a) The conclusions just discussed about property rights and incentives to work carry over to more general and realistic settings(as will the conclusions about the other advantages of property rights to be discussed). In particular, in an economy with many kinds of goods and money as a medium of exchange, an product of his labor. Rather, his incentive to work will involve his being able to consume the various goods purchased with his money wages. But this difference does not alter the point that an individuals incentives to work will tend toward the desirable under property rights and will tend to be suboptimal in their absence (b) In modern industrialized countries, the majority of individuals are motivated to work not because they consume or sell the output they produce, but rather because of the incentives and salary structure established by supervisory entities where they are mainly to certain farmers, artisans, shopowners, and independent professionaltricted employed. The individuals who have property rights in their own output are restricted (c) In extreme instances in which property rights break down, we observe that little productive work is done. During civil wars and other episodes of great upheaval productive work generally ceases. (Instead, people devote themselves to protecting what hey have, and some engage in looting. 2.3. Incentives to maintain and improve things. An essentially similar justification for property rights is that they are associated with incentives to maintain and improve things. If we reinterpret the model from the last section and assume that the result of work effort is the maintenance or improvement of durable things(as when individuals apply oil to machines or fertilize farm land), we may say that it is socially optimal for an individual to work an extra hour if the utility gained from maintenance and improvements exceeds the disutility of work. Further, this outcome will not occur in the absence of property rights; when durable things may be taken from individuals, they will ot benefit from improving them (if a machine would be taken from an individual, he will not have an incentive to maintain it well). When, however, individuals hold possessory they will possess the things in the future and therefore be able to enjoy the gains from o rights in durable things, they will have an optimal motive to improve the things becaus maintenance and improvements. Likewise, if there is a supervisory entity that can observe individual behavior, the supervisor can induce individuals to work optimally to maintain and improve things Comments. (a) Suppose that we take into account the possibility that durable things will be transferred in the future, for example, that farmland will change hands when the possessor becomes too old to continue to work. Then the socially optimal amount of effort to improve durable things will reflect the future use of things by new possessors as well as present ones; the socially desirable amount of fertilization of farmland will reflect the enhancement in its productivity for both current and future users Moreover, the socially optimal improvement in durable things will be promoted not only by possessory rights, but also by rights to transfer things. Notably, possessors will often transfer things by selling them and sales prices will generally reflect improvements made in things. Because present owners will anticipate this, they will be led to make Chapter 7-Page 6

output, are, of course, different questions. The point here is simply that it is possible for the optimal outcome to be achieved if a supervisor has information about individuals’ work effort and possessory rights in output. Comments. (a) The conclusions just discussed about property rights and incentives to work carry over to more general and realistic settings (as will the conclusions about the other advantages of property rights to be discussed). In particular, in an economy with many kinds of goods and money as a medium of exchange, an individual’s incentive to work will not inhere in his being able to consume the literal product of his labor. Rather, his incentive to work will involve his being able to consume the various goods purchased with his money wages. But this difference does not alter the point that an individual’s incentives to work will tend toward the desirable under property rights and will tend to be suboptimal in their absence. (b) In modern industrialized countries, the majority of individuals are motivated to work not because they consume or sell the output they produce, but rather because of the incentives and salary structure established by supervisory entities where they are employed. The individuals who have property rights in their own output are restricted mainly to certain farmers, artisans, shopowners, and independent professionals. (c) In extreme instances in which property rights break down, we observe that little productive work is done. During civil wars and other episodes of great upheaval, productive work generally ceases. (Instead, people devote themselves to protecting what they have, and some engage in looting.) 2.3. Incentives to maintain and improve things. An essentially similar justification for property rights is that they are associated with incentives to maintain and improve things. If we reinterpret the model from the last section and assume that the result of work effort is the maintenance or improvement of durable things (as when individuals apply oil to machines or fertilize farm land), we may say that it is socially optimal for an individual to work an extra hour if the utility gained from maintenance and improvements exceeds the disutility of work. Further, this outcome will not occur in the absence of property rights; when durable things may be taken from individuals, they will not benefit from improving them (if a machine would be taken from an individual, he will not have an incentive to maintain it well). When, however, individuals hold possessory rights in durable things, they will have an optimal motive to improve the things because they will possess the things in the future and therefore be able to enjoy the gains from maintenance and improvements. Likewise, if there is a supervisory entity that can observe individual behavior, the supervisor can induce individuals to work optimally to maintain and improve things. Comments. (a) Suppose that we take into account the possibility that durable things will be transferred in the future, for example, that farmland will change hands when the possessor becomes too old to continue to work. Then the socially optimal amount of effort to improve durable things will reflect the future use of things by new possessors as well as present ones; the socially desirable amount of fertilization of farmland will reflect the enhancement in its productivity for both current and future users. Moreover, the socially optimal improvement in durable things will be promoted not only by possessory rights, but also by rights to transfer things. Notably, possessors will often transfer things by selling them and sales prices will generally reflect improvements made in things. Because present owners will anticipate this, they will be led to make Chapter 7 - Page 6