THE ENFORCEMENT OF PROPERTY RIGHTS IN CHILDREN, AND THE MARRIAGE CONTRACT

THE ENFORCEMENT OF PROPERTY RIGHTS IN CHILDREN, AND THE MARRIAGE CONTRACT 


MARRIAGE, as a form of contract for the enforcement and transfer of rights between husband and wife, must rank among one of the major economic activities in the history of civilisation. It involves diversified decisions and immense values. Moreover, few contracts have been more universally and severely regulated by the church or the state. Some thorough economic analysis of the marriage contract is long overdue. It is no easy matter, however, to discover a case in which the terms of the marriage contract have historically been freely dete:cmined in the market. If such a case need be made, then traditional China seems to qualify most closely in essential aspects despite the fact that marriage decisions there were made, not by the marrying parties, but by the parents who held property rights in marriageable children. Thus a study of the marriage contract in China necessarily entails a corollary study of the gains and costs of enforcing property rights in the children. Towards this joint venture this study is directed. In the analysis, we shall be looking for economic explanations of phenomena hitherto regarded as purely sociological or anthropological. It is hoped that the study will raise some pertinent questions on our modern marriage practices.
I.


THE CONSTRAINTS OF PROPERTY RIGHTS AND THE CRITERION OF MAXIMISATION


Just as dogs were raised to hunt for their masters before they were pets, so in early traditional China children were raised as a source of income and a store of wealth. Provided that a child obeys his parent, he is a relatively secure asset holding. 2
1 It requires courage to be idiosyncratic, and I should not have turned this paper into a serious piece except for the encouraging remarks of John S. McGee, Axel Leijonhufvud, Theodore W. Schultz and GeorgeJ. Stigler. I am also grateful for comments by Yorem Barzel, Thomas Borcherding, Jack Hirschleifer, Harry Johnson, Benjamin Klein, Walter Oi, Eugene Silberberg, Gordon Tullock and Dean Worcester. The study is part of a proposed research in the general area of contracts, financially supported by the National Science Foundation. Information which initiated this paper was graciously offered by my own mother, who was blindly married yet who surrendered property rights over her sons and did not demand negative dowry for her daughters. 2 Where wars or famines are frequent, the assets chosen by individuals may take peculiar forms based on their potential security under different kinds of risk. One example is the popularity of gold teeth in Southeast Asia two or three decades ago. I do not hold that social instability is a necessary or a sufficient condition for the raising of children as capital assets. However, instability tends to increase the relative value of human capital, since not even land possesses its attribute of " the dog that follows."


642


THE ECONOMIC JOURNAL


[JUNE


Until only a few decades ago, parents in China held property rights in children: " The head of the family was its eldest male member. . . . He held the title to all family property and he alone could dispose of it, as well as of the earnings and savings of all the family members. He settled the marriages of his children and signed the marriage contracts. . . . Furthermore, the law exonerated father or grandfather who killed his son or grandson unintentionally when chastising him ' in a lawful and customary manner.' Nobody disputed the right of the head of the family to sell its members into slavery." 1 . That it was the father, not the mother, who held these rights is frequently noted,2 although in the available sources it is not always clear which parent had what right of action. However, for observations which :ve here seek to explain, a detailed separation of their rights is not essential, and we shall arbitrarily use the word "parents" synonymously with "head of household." " Exclusive rights" in children implies the right to extract both pecuniary and non-pecuniary income from them. Yet the actual relationship within a traditional Chinese family was complicated in two important respects. First, an intricate property-rights structure existed within each family. While an identified head of household had rights over all members, his subordinates might also exert rights over their inferiors. For example, a son was ranked superior to a daughter, a wife superior to a concubine, and so on. Some implications of this complex structure will be discussed in the relevant context. Second, children in China were not treated in every respect as slaves. For example, they were granted the right to inherit wealth, and to raise children of their own. When one takes into account such factors as " love" and " concern" in rearing children-which are not transferable and therefore not measurable in terms o?wealth-it may be essential that a criterion of utility be used for the postulate of constrained maximisation. Owing, however, to problems to be discussed in the concluding section, we hold for the most part to the wealth-maximising postulate. The apparent conflict between the use of the
1 Olga Lang, Chinese Family and Society (New Haven: Yale University Press, 1946), pp. 26-7. See also G.Jamieson, Chinese Family and Commercial Law (Shanghai, 1921), Chapter i; Hsu Chaoyang, Chung-kIlO Ch'in-shuFa Su-yuan (The Origin of Chinese Kinship Law) (Shanghai, 1933). Even after the establishment of the Chinese National Government in 1911, the liberation of property rights in children was gradual. See Marc van der Valk, An Outline of Modern Chinese Family Law (Peking, 1939). 2 As we shall see, every mother was once a property acquired by the father's family. This, together with an assertion that inheritance given to one's own offspring yields a higher level of utility to the giver than if the inheritance was given to an outsider, explains the dominant role of the father in holding property rights. The general impression (see Lang, op. cit., p. 29) of" dignity and sternness" of a Chinese father, in contrast to the" kind and tender" mother, may also be explained. Given the different property rights constraints of the parents, the price of" love" and " affection" was relatively high for the father and low for the mother. Thus the quantities of the non-pecuniary goods demanded by parents of their children vary between the parents.


1972]


ENFORCEMENT OF PROPERTY RIGHTS IN CHILDREN


643


wealth criterion and the failure to treat children as slaves is resolved in the following way. (1) We posit that unexplained non-slavery features are recorded and did exist. (2) Given these conditions, we pursue the analysis via the wealth-maximising postulate. The lack of explanation for these conditions suggests only incompleteness in analysis, which may be filled with a generalised utility theory; it does not imply any internal inconsistency in the argument. 'The general hypothesis may then be stated as follows. In seeking wealth maximisation, the decision-making head of a family was subject to three identified sets of constraints: (1) the existing property rights of the family, embodying endowment, market and production opportunities; (2) the posited non-slavery conditions in child rearing; (3) the costs of enforcing and transferring the rights in children. Taking the first two constraints as given, we shall concentrate on wealth maximisation subject to the third constraint, namely, the costs of enforcing and transferring property rights. The hypothesis is thus wealth maximisation subject to the constraints of transaction costs, operated with the implicit aid of other elementary economic principles. Test implications are largely drawn from the tendency of parents to reduce these costs. One implication of particular relevance is the predictable rise of a fourth set of constraints, namely, those of custom. All implications are tested against the factual background of the experience in traditional China.


II.


MARRIAGE PRACTICE IN CHINA EXPLAINED


Consider the parents of a family who raise children as a source of income. Allow that in the course of time the grown children would be given the right to marry and to raise children. Assume that the rights held by an older generation over a younger generation repeat in the same pattern through time. Any grown son or daughter could attain greater wealth, if, after having obtained" feedings" from his parents, he alone received all the income he generated. The cost to the parents of enforcing payments from the child would be higher if the child left home and worked on his own. To protect their investment return, the parents therefore prohibit the establishment of separate families following the child's marriage. 1 Indeed, one suspects that in China the repeated emphasis on the virtue of filial piety 2 was intended to reduce the cost of enforcing property rights in children rather than to enhance the gain in terms of morality. The fact that the renting or leasing of brides was not observed is explainable by the complexities which would have been introduced by pregnancy
1 See Lang, op. cit., Chapters. ii-iv; Francis C. K. Hsu, Under the Ancestor's Shadow (New York: Columbia University Press, 1948), and Chen Ku-yuan, Chung-kuo Yun Yin Shih (A History of the Chinese Marriage) (Shanghai, 1936). 2 See Lang, op. cit., and Hsu, op. cit.


No. 326-vOL. 82.


UU


644


THE ECONOMIC JOURNAL


DUNE


and child raising. Since the formation of separate families was prohibited, every marriage therefore involved an outright transfer; the parents of a family either yielded their rights to one child or acquired property rights over another through a marriage contract with another family. The parents, not the marrying children, participated in the contract. Note that a purchased bride might also be sold by the groom's parents, thus ending the marriage at any time. In every such marriage, then, one child was retained and another acquired. It is not easy to explain why, in China at least, the popular choice was to retain sons and marry daughters away. If the preservation of the family line was a prime consideration there is no a priori reason why the retention of sons would serve the purpose better. Granted, male labour was often more valuable than female labour, but why not marry the son away in exchange for a higher dowry? The explanation advanced here depends on the cost of forcing children to work and of deterring possible runaways after the marriage contract was formed. For two reasons it would cost less to bring home a girl and to retain a submissive son upon his marriage. First, the cost to parents of preventing runaways would be lower for their own child than for another acquired by marriage contract, since years of strict discipline 1 would have yielded an obedient character. Second, it would cost less to "tame" a girl than a boy; the difference in physical strength was one reason, and another will be offered later. Neither reason, standing alone, would dictate a choice as between retaining a son or a daughter. But together the two cost components would make it more advantageous to retain the son. In every such marriage, a transfer of property rights was made from the bride's parents to the groom's parents. In the words of Maurice Freedman: " Her marriage cut her off economically and as a legal person from her own family and transferred the rights in and over her to the family receiving her. . . . The right and duties in respect of her would lie with her husband's people. . . . She had no secure base outside this family from which to operate, because, while she might try to bring in support from her family of birth to moderate oppression, she could not rely on it. To a large extent physically, and in all degrees legally, she was locked within her husband's gates." 2 Payment for the outright transfer of property rights to the bride was made by the groom's parents to the bride's parents. This sum was, in effect, a negative dowry, the western term for which is "bride price." 3
1 See Lang, ibid., Chapter iii; and P. L. K. Tao, "The Family System in China," The Sociological Reviews,January 1913, pp. 47-54. I "The Family in China, Past and Present," Pacific Affairs (Winter, 1961-62), p. 328. a See Arthur P. Wolf, " Adopt a Daughter-in-law, Marry a Sister: A Chinese Solution to the Problem of the Incest Taboo," American Anthropologist, 70, 1968, p. 867.


1972]


ENFORCEMENT OF PROPERTY RIGHTS IN CHILDREN


645


The amount of the bride price might indeed be substantial, and it could be in money or in kind. A Chinese writer describes the practice thus: " The term' exchange marriage' refers to the situation where girls were treated as commodities; a wife or a concubine was acquired by forgoing other goods in exchange for her. It was the chief practice after the practice of marriage by force. . . . The practice [of exchange marriage] was developed into . . . 'pecuniary marriage' which, disguised in [customary marriage procedures], was in fact a sale and a purchase. There were also Ii which recognized the transactions, and laws which permitted special sales." 1 Historical records indicate that such market transactions for marriage were universal during at least some periods in China: "With no exception, every marriage is conducted through pecuniary means. When bargaining for more and competing for less, shamelessly people do not find"the practice strange at all." 2 Partly because a bride's running away after marriage might be viewed as a breach of contract, partly because a bride with a lower inclination to run away would yield a higher bride price, the bride's parents had incentives to discipline her with ethics and other devices consistent with wealth maximisation. Thus in educating a young girl, heavy emphasis was placed on the merits of loyalty to the husband and his parents. She was taught that after her marriage her childhood home would be called" foreign home," and that the mere thought of voluntary divorce was sinful. 3 Her participation in social activities was discouraged or prohibited. 4 Bedtime stories to a young girl included tales of runaway brides who met disaster. To reinforce all this teaching, the girl was subjected to foot-binding at about five years of age. 5 To the groom's parents, the physical handicap thus imposed was more than counterbalanced by the girl's lessened ability to run away. They would therefore be willing to pay a higher market price for the bride with bound feet, and this served as a signal for the bride's parents to bind her feet long before she was pledged through a marriage contract. The relevant cost of bound feet was the value of production
Chen Ku-yuan, op. cit., pp. 83-4. The translation is mine. Quoted ibid., p. 95. The translation is mine. For other similar evidence, see ibid., pp. 83-97. 3 And in a divorce, to be distinguished from the wife being sold, a woman was not necessarily freed from the groom's family (ibid., pp. 233-51). The associated complexity does not concern us here. Furthermore, numerous rules prohibited a woman's right to remarry after her husband died (ibid., pp. 223-32). And marriage to the spirit of a dead man was not unknown, as cited by Florence Ayscough, Chinese Women (Boston: Houghton-Mifflin Company, 1937) pp. 32-4. , See, in particular, Chen Tung-yuan, Chung-kuo Fu Nu Sheng Huo Shih (A History of the Lives of Chinese Women) (Shanghai, 1937). See also Asycough, Chinese Women, and Li Yieni Tsao, " The Life of a Girl in China," in American Academy of Political and Social Science, China Social and Economic Conditions (January 1912), pp. 62-70. 6 See Li Yieni Tsao, op. cit., p. 67. In Lang, op. cit., p. 46, it is stated that" bound feet kept women at home, made them safer, less movable property. This was, in fact, often recognized." An alternative explanation frequently supported by Chinese writers will be discussed in the concluding section. The practice offoot-binding was recorded as early as the Tang Dyuasty (A.D. 618A.D. 907). See Chen Tung-yuan, op. cit., pp. 125-8.
1 2


646


THE ECONOMIC JOURNAL


UUNE


forgone. In such activities as weaving and other domestic work, bound feet were not a great impediment. They were, however, a great disadvantage in the kind of work where a male had a comparative advantage, as in farming; and this restriction on possible activity offers another reason for the lower cost of taming a girl than a man. Following the principle of investment in human capital, the girl was naturally taught those skills-sewing, cooking, weaving 1-in which she could be most productive and thereby bring the highest bride price to her parents. In any such marriage, contractual negotiation would have been vastly more complex if the marrying children had also participated. Their preferences for good looks and love might have come into conflict with the goals of parents! The custom which evolved, therefore, was blind marriage, with the groom allowed to see the bride's face for the first time enly after the marriage. II The ethics of filial piety specified unconditional acceptance of the parents' choice of mate. The formation of a traditional Chinese marriage contract therefore differed from the modern western marriage contract in at least three ways. (1) The groom and bride did not themselves engage in searching and the cost of searching was borne by the contracting parents, not by the marrying couple. (2) The marriage contract was an unregulated market contract, whereas numerous legal rules govern today's marriage. (3) The property rights transferred under the blind marriage contract constituted an outright purchase. For all these reasons, great effort was needed to form such a contract and to enforce it. As in any other transaction where costs are high, cost-saving devices were brought into play. We have already mentioned as enforcement devices the practice of foot-binding and the disciplinary emphasis on submissiveness and filial piety. Another useful device was the employment of a middleman, the marriage broker. In the only source that I could find giving substantial details, the middle- . man of the marriage transaction, usually a woman, is described as one who served jointly as a searcher, a negotiator and a referee in enforcing the contract. 3 Her multiple offices are explainable by the lower cost in joining these services. During the contractual negotiation, the middleman would provide information on the skills and other attributes of the bride. The bride price was determined by competitive bidding among potential buyers,'
See Ayscough, op. cit., Li Yieni Tsao, op. cit., and Chen Tung-yuan, op. cit. Even during the ceremony, the bride wore a heavy mask. See Chen Ku-yuan, op. cit., pp. 151-64; Ying-Mei Chun, " A Wedding in South China," in American Academy of Political and Social Science, op. cit., pp. 71-3; and Shang Ping-ho, Li Tai She Hui Peng Su Shih WuKao (A Study of Traditional Chinese Customs) (Shanghai, 1938), pp. 235-37. Stories are abundant of how a certain boy attempted to take a glance at his future bride, and how the parents took precautionary measures. 8 See Chen Ku-yuan, op. cit., pp. 147-50. , Ibid., Chapter iii.
1 B


1972]


ENFORCEMENT OF PROPERTY RIGHTS IN CHILDREN


647


and the middleman received a competitive share of the bride price as well as gifts from both families, and sometimes a bonus when the information on the bride turned out to be reliable, which was generally the case, it is said, except that the beauty of the bride was usually exaggerated, a tactic understandably endorsed by both sets of contracting parents. 1 Specialisation in handling marriage contracts was evident. It was common for a middleman to obtain the negotiation right from the parents years before the child's marriage. In all essential aspects, then, she ran not only a brokerage firm but an escrow firm as well. Since the delineation and enforcement of rights under any escrow relationship depends on the legal structure in force at the time, it is not surprising to find that the role of the middleman as a referee in a traditional Chinese marriage differs from period to period. 2 ? Perhaps because of uncertain prospects, futures contracts were sometimes obtained. The parents of a young boy might draw a contract with the parents of a young girl, with a bride price presently determined and a future marrying date specified. In this case, however, the boy's parents usually participated in monitoring the training of the girl, for the girl's parents might have been expected to " under-invest" once a futures contract was formed. Written source materials on such futures contracts are not available. s However, we may turn to the interesting practice of t'ungyang-hsi, which involves a present marriage contract for infants. T'ung-yang-hsi means a "daughter-in-law raised from childhood." That is, " when a girl is born . . . she is often given away or sold when but a few weeks or months old, or one or two years old, to be the future wife of a son in the family . . . which has a little son not betrothed in marriage. "4 Such practice is recorded as early as the Sung Dynasty (A.D. 960A.D. 1279),5 and surveys conducted a few decades ago reveal that t'ung-yanghsi constituted roughly 20 % of the marriages at that time. 6 The considerably lower bride price in purchasing a t'ung-yang-hsi than in acquiring a grown bride 7 is what one would expect, since the costs of feeding and training the girl are now to be borne by the boy's parents. In a world uncomplicated by transaction costs and risks, at any moment in time the present values of the bride price, net of feeding costs, are identical under the two options. How would we explain the choice of one option over another? Following the Chinese, I shall call the purchase of a t'ung-yang-hsi a
Ibid., pp. 149-50. Historical records indicate that in ancient times the middleman was a legal or government official, but at later periods she operated on a private basis. Little information is available on the earlier type. See ibid., pp. 147-8; and Shang Ping-ho, op. cit., pp. 233--4. 3 Much of the information contained in these two final paragraphs was obtained from my mother. 4 Justus Doolittle, Social Life of the Chinese (New York: Harper and Brothers, 1865), p. 195. 6 Chen Ku-yuan, op. cit., p. 109. 6 Wolf, op. cit., pp. 864-6. The frequency oft'ung-yang-hsi in different surveys varies greatly. 7 See ibid., p. 867, and Chen Ku-yuan, op. cit., p. 109.
1 2


648


THE ECONOMIC JOURNAL


[JUNE


" minor marriage" and the purchase of an adult bride a " major marriage." Recognising the existence of transaction costs and risks, l the preferred choice of the minor marriage contract may be attributed to the saving of transaction costs of two different types. One type exists in the financial market. The parents of an infant girl, wishing to borrow so as to invest in raising the child, may find that the borrowing rate is higher than the lending rate. 2 Under the relevant borrowing rate of discount, they will be confronted with a lower present value, net of feeding costs, in keeping the child than that attainable by selling out under a minor marriage contract. Another and perhaps more important type of cost saving under the minor marriage contract lies in the lower cost of enforcing property rights in the bride. As a rule, a t'ung-yang-hsi was reared in the groom's family. Submissiveness and loyalty are more easily assured in a girl brought up from childhood. 3 Therefore the costs of forcing the bride to work and of deterring her running away will be lower. This also implies that, other things equal, the fee paid to the middleman will be lower with a minor than with a major marriage contract, since the middleman's services as a referee will be less required. While transaction costs made the minor marriage contract a preferred option, however, the purchase of a t'ung-yang-hsi did involve more risk than the purchase of an adult bride. The mortality rate for infants was high, and the I.Q. and other productive potential of a bride could not be easily revealed in early childhood. The very existence of risk served to reduce the price of a t'ung-yang-hsi, and risk aversion reduced the price further. Since more risk generated a lower bride price for a t'ung-yang-hsi, and since the identified transaction costs were lower for a minor marriage, it is puzzling to note that minor marriage contracts were generally less frequent than major contracts. One may resort to a variety of tentative explanations. Love of one's own child may have generated utility which made some parents elect to keep the child until marriageable age. Again, and perhaps more important, in many periods in China the selling of t'ung-yanghsi was prohibited by law. 4 Or consider the explanation suggested by Professor George J. Stigler: with a high mortality rate for infants, why bother with the minor marriage if it might well have to be done over again? In spite of the difficulties, some implications of the foregoing analysis on the choice of a marriage contract may be tested. The existence of transaction costs in the money market implies that it would mainly be the poor who sold t'ung-yang-hsi, owing to the lack of collateral for borrowing
1 In a different context the same approach is used in my "Transaction Costs, Risk Aversion, and the Choice of Contractual Arrangements," Journal of Law and Economics, April 1969. 2 See Jack Hirshleifer, Investment, Interest and Capital (New York: Prentice-Hall, Inc., 1970), Chapter vii. 3 This point is recognised in Wolf, op. cit., pp. 868-9. 4 See Chen Ku-yuan, op. cit., pp. 109-10. During the period under which the aforementioned surveys on t'ung-yang-hsi were conducted, the selling of t'ung-yang-hsi was officially illegal.


1972]


ENFORCEMENT OF PROPERTY RIGHTS IN CHILDREN


649


money. This has been the usual observation. 1 The higher costs of negotiating and enforcing a major marriage contract (particularly the costs of maintaining the bride's obedience) imply that marriage procedures were more involved, hence more costly, with major than with minor marriages. This is also supported by the facts: the complexities are evident in drawing up documents, finding witnesses, and various customary steps in transferring property rights to the bride in a major marriage. 2 As with any capital asset, the acquisition of a bride as a factor of production and as a stock of wealth must accord with the laws of diminishing returns and of portfolio selection. A custom restricting marriage to a one-to-one basis would have been inconsistent with the maximisation of wealth by the decision-making unit of the household. A solution to this problem, of course, was to allow the accumulation of concubiq.es. Concubinage is a complex subject. A concubine was not a wife; she usually had less rights than, or was subordinate to, the wife. 3 A concubine was not a mistress or a maid; her acquisition entailed a marriage contract implying the right to bear and raise children, although the associated contractual procedure was less complicated than that for the wife. 4 It is not easy to explain why a concubine was given inferior rights in a family, but in fact the wife (usually the man's first marriage partner) was frequently assigned a commanding position over concubines, monitoring the job performance of the later brides. 5 In spite of all claims that concubines were" inferior" I have been unable to find any evidence that their productive abilities were less. The indisputable fact of their lower rank can perhaps be best explained in light of the inheritance rule under the property-rights structure of a traditional Chinese family. Inheritance, defined here as the transfer of property rights from an older to a younger generation, was not executed abruptly at the death of the head of the household. Rather, it was a progressive process, with each family member gradually freed and allowed to command higher income as time went on. 6 In general, given a certain position, the seniority rule
See ibid., p. 109, and Wolf, op. cit. See Wolf, op. cit., pp. 866-7, and Chen, op. cit., pp. 141-60. Our discussion also implies that if the practice of foot-binding were to be eliminated, the process would begin with t'ung-yang-hsi, but I have been unable to find evidence to refute or confirm this implication. Furthermore, while the retention of grown sons was a rule, the lower enforcement cost in bringing up a boy or a girl from childhood implies that the selling of infant sons was not unusual. This is verbally confirmed by living Chinese elders. 3 See Chen Ku-yuan, op. cit., pp. 62-8; and Shih Yi-yun, " Kuan Yu Wu Kuo Ching Tai Fa Chih Ch'i Chih Yen Chiu" (Concubinage Under Recent Chinese Law) Journal of Social Science (National Taiwan University, August 1956), pp. 137-76. 4 See Shih Yi-yun, op. cit. A bride price was required for the transfer of a concubine, and the practice was recognised by law. 5 In the event of offences committed against the senior authority of a family, the concubine was subject to heavier punishments than the wife. See ibid. 6 We are not concerned here with the intricacies of which types of rights were first transferred, nor with the sequence of their distribution to different persons in different positions.
1 2


650


THE ECONOMIC JOURNAL


(JUNE


prevailed not in terms of physical age but in terms of the length of association with the family. Thus the wife would automatically be entitled to more rights than the concubine. Under the concubinage system, some men had no mate and some several.1 The bachelors' demand for sex was met in an unrestricted prostitution market,2 which had the flexibility of short-term rental contracts. The female entertainers were usually purchased outright by prostitution firms, sometimes from the girl's parents, sometimes from the portfolios of concubine holders. The middlemen summoned for these transactions might be the same ones who negotiated the marriage contracts. The ethics ofloyalty and decency emphasised when a bride was acquired no longer held when she was discharged.


III.


TRANSACTION COSTS AND THE RISE OF CUSTOM


An individual in any society must face constraints far more complicated than those encountered by Robinson Crusoe. For the most part, the economic literature analyses two types of constraints which do not exist in a Crusoe economy. Legal constraints are embodied in property rights and a variety of government regulations. Market constraints, conditional to existing legal arrangements, include all contractual arrangements that in effect stipulate and restrict the behaviour of the contracting parties. Because the enforcement and transfer of property rights frequently entail considerable costs, economic theory which earlier ignored such costs has failed to explain a large class of observations. That, perhaps, is why analyses which explicitly recognise the constraints of transaction costs have recently come into favour. 3 However, another reasonably important type of constraint has also been neglected: the constraints of custom. A " custom" may be defined as a set of implicit social rules restraining the behaviour of one or more classes of individuals on a simultaneous basis and generally not subject to negotiation. The word " class" as used here may be illustrated by a Chinese traditional distinction: at the death of a father a son must mourn according to strict rules which more remote relatives need not follow. If customs are to be effective, those who violate them must somehow be penalised. In traditional China, penalties might be imposed by the court, by the commanding parents, or by society in reducing the violator's alternative earnings.
1 According to surveys conducted in the 1930s, the average size of a Chinese family was about five. Numerous reports on the existence of large families suggest that the variance must be very large. See Francis Lang-kwang Hsu, "The Myth of Chinese Family Size," American Journal of Sociology, March 1943, pp. 555-62. 2 See Shang Ping-ho, op. cit., pp. 517-27. 3 The term" transaction costs" is used here arbitrarily to include all costs which do not exist in the Crusoe economy. For the difficulty, for example, of separating the cost of enforcing from the cost of transferring property rights, see my" The Structure ofa Contract and the Theory ofa Nonexclusive Resource," Journal if Law and Economics, April 1970.


1972]


ENFORCEMENT OF PROPERTY RIGHTS IN CHILDREN


651


The interpretation advanced here is that social customs take form because they serve to reduce the costs of delineating and transferring property rights. 1 This interpretation may be illustrated by briefly discussing the role of ethics in association with customs, and by showing the relevance of customs in delineating and enforcing rights in a traditional Chinese family. The development of a custom can usually be traced to a society's normative value standards. These standards, forming an ethical code, in turn stem from the interpretations of certain religious principles or the " teachings" of some men. Filial piety, the importance of clearly delineated relationships in forming family and society, and decency and obedience in woman-kind were the main principles of the Confucian ethical code imposed by Chinese parents on their children. 2 To understand the customs in the Chinese family which were.in harmony with the Confucian ethical code, a brief description of the family structure will help. Except for the head of the household, who was once a " slave" himself, every family member was a slave with varying degrees offreedom. 3 They differed in their rights to make decisions and to receive income. This complexity stemmed partly from the gradual inheritance practice and the associated seniority rule. It also stemmed from discrimination against women, owing both to the lower rents they generally yielded to their parents 4 and to the fact that they were either purchased or would be sold and thus yielded less non-pecuniary gains to their superiors.
1 This interpretation, of course, does not exclude the possibility that some customs may generate utility. I could not, however, think of any clear case where a custom satisfying the given definition was originally intended solely for the purpose of" enjoyment." R. H. Coase has taught us that when costs are associated with the delimitation and transaction of rights the courts, through legal action, may decide on an economic issue while reducing the costs of transactions. (See Coase, " The Problem of Social Cost," Journal of Law and Economics, 1960.) But even the operation of courts is costly; and when customs which differ from the courts are adopted to guide certain behaviour for a general class of individuals, they may also serve as rules for court decisions. Implied in our present discussion is that customs differ under different systems of property rights, and vary as the efforts of legal enforcement of law vary. And because the costs associated with delineating and transferring rights involve different natures and dimensions, the customs in many societies are complex indeed. These complexities do not concern us here. For our present purpose we shall confine our attention to traditional Chinese customs pertaining to the enforcement of property rights in children. 2 See Ayscough, op. cit.; David S. Nivison and Arthur F. Wright, eds., Confucianism in Action (Stanford; Stanford University Press, 1959), pp. 1-24; James Legge, The Chinese Classics (Hong Kong; Hong Kong University Press, 1960), Vols. I and II; and Tseng Tu, Hsiao Ching Hsin Ta Tu Pen (A New Textbook of the Filiality Classic) (Shanghai, 1906). 3 This point is sometimes recognised. See, for example, Liang Chi-Chao, " Chung-kuo Nu Ti Chih Tu " (The System of Slavery in China), The Tsing Hua Journal (Peking; December 1925), pp.527-53. 4 The predominance of agriculture in traditional China and the associated demand for male labour may offer an explanation. The term" poor investment," which survives even today, was popularly used in reference to a daughter. For evidence of discrimination against women in terms of rights, see Chen Tung-yuan, op. cit.; and Yong Mon-Chuen, " Chin Wu Shih Nien Lai Chung-kuo Chia Ting Ti Pien Hua" (Changes in the Chinese Family System during the Last Fifty Years), Academic Review Quarterly (Taipei,June 1955), pp. 66-72.


652


THE ECONOMIC JOURNAL


[JUNE


Imagine a family with several sons each of whom had a wife and several concubines; each woman was raising children of both sexes and was employing different maids; and there was a host of cousins of different degrees. Imagine further that each person was born a slave, some being purchased and some retained, that each was being freed gradually but at different paces, and that each was allowed to command different aspects of rights. Now the total real income of the family was to be shared among all members according to who had what rights of action. The delineation and enforcement of rights among all members of the family would be very costly indeed. How would customs come into play? To simplify the discussion, view the income attainable by any member of a family as a measure of the degree of his rights. With this view we accept the transitive rule implied in the literature on the Chinese family: 1 if A is superior to Band B superior to C, then A is superior to C. The problem of property-right delineation must now satisfy the condition that, for any randomly selected pair of family members, the rules must be able to tell either which member was the superior one and to what extent, or that the pair was equal in terms of rights. Two general delineation rules existed. One was the seniority rule described earlier. The other, less familiar to us, was the kinship rule: members positioned closer to the" cult of ancestors" (i.e., by male descent) were superior to those of more distant degrees; and, given the same degree, males were superior to females and wives superior to concubines. The kinship rule also held that all members of a higher kinship rank were superior to all members of a lower kinship rank; for example, a younger brother was superior to his older sister by the same parents. Thus the ranking of every randomly selected pair in a family was determined. One natural outcome of such delineation was the designation of names and kinship terms of a complexity unmatched in other societies. A Chinese might have as many as eight names and terms designating seniority, sex and kinship, and his rights with respect to all other members of the family.2 Strict rules determined who must call whom by what name and term. The custom of kowtow and other gestures which served as admission of inferior rights by one family member to another were an integral part of family discipline. 3 The assignment of rights through naming and gesturing was apparently
1 Statements of fact contained in this and the next paragraph are distilled from Lang, op. cit.; Freeman, op. cit.; Shih Ye-yun, op. cit.; P. L. K. Tao, op. cit.; and Marc van der Valk, op. cit., Pt. 1. 2 See Chinga-Chao Wu, "The Chinese Family: Organization, Names, and Kinship Terms," American Anthropologist,July 1927, pp. 316-25; and A. L. Kroeber, "Process in the Chinese Kinship System," American Anthropologirt,January-March 1933, pp. 151-7. 3 Offences committed by one member against another were punished by different measures according to rank, and they were enforced either by the court or by the head of the household. See Shih Yi-yun, op. cit.


1972]


ENFORCEMENT OF PROPERTY RIGHTS IN CHILDREN


653


not enough. Ceremonial rites also served as formal occasions for the exercise of rights by each member.! The rites of mourning and of ancestor worship were in accord with filial piety; thus the occasions also served as disciplinary measures emphasising the respect due to the rights of an elder. The following account is obtained from discussions with living Chinese elders. At the death of a family member, all members inferior to him were obliged not only to mourn and to wear special costumes for the occasion, but also to assume designated positions during the funeral ceremonies. The lengths of mourning time and a host of other considerations were likewise designed so that all mourning members were systematically ranked. 2 In ceremonial mourning, it would have been inconsistent with the doctrine of filial piety for one of superior rank to mourn an inferiqr; therefore members superior in status to the dead were not ranked. Moreover, funeral occasions were irregular. More general ceremonial observance was therefore met by ancestral worship, held regularly at the time of festivals and intended primarily for the higher-ranked members of a family.3 Sacrificial offerings to the dead ancestor were in harmony with filial piety; paper houses and false money were burned to ashes. To induce a reluctant member to undergo the personally costly process, food and drinks were served after the worship. 4 Two side remarks are needed. The use of ethical code and customs is not necessarily the cheapest way of enforcing property rights in man. The practices used in China are not observed in a slave economy. Perhaps less gentle methods would cost less. But given the non-pecuniary and nonslavery aspects which existed in a family, the observed practices in China might very well have been efficient. A second comment is that the described delineation and customs of ranking give the impression of prima facie inflexibility in the system. Might not a highly productive member be placed at such a low position that his incentive to produce was reduced? This could be true, but records indicate that promotion of one member at the expense of another was not uncommon. a As for the non-negotiable customs, it is natural that they take
1 See Feng Yu-Ian, "Ju Chia Tui Yu Hun Sang Chi Li Chih Li Lun" (Confucianism in Marriage, Mourning and Worship), Yen-Ching Hsueh Pao,June 1928), pp. 343-58. See also Marc van der Valk, op. cit., pp. 17-19. 2 The observed practice that higher-ranking members were subjected to more costly rules of mourning behaviour than were lower-ranking members poses an economic puzzle for which I have no answer. 3 I am told that in some villages pork and grain shared during the worship were apportioned according to the different positions and gestures designated for each member. ? It was usual for a cooked chicken to be placed in front of the ancestral tablet which represented the ghost of the dead. When a young son was reluctant to kneel and kowtow to the tablet, a witty mother might urge, " If you don't believe in the ghost, you should at least believe in the chicken." 5 For example, the rank of a wife might be abolished so as to promote a concubine; see Chen Ku-yuen, op. cit.; and Shih Yi-yun, op. cit.


654


THE ECONOMIC JOURNAL


[JUNE


a long time to die. Some of the old festivals still survive today, although they no longer serve the old purposes.


IV.


SOME CONCLUDING REMARKS


Any lack of neatness in this paper may partly be attributed to the large class of observations which I have sought to encompass but is mainly due to some analytical problems which I have avoided. In this concluding section, I would like to bring these problems to light, with the hope that they may stimulate further studies in the general area of the marriage contract. The most difficult problem encountered in economic analysis of what have been alleged to be sociological phenomena is the choice of a criterion for the postulate of constrained maximisation. For the most. part I have used the postulate of wealth maximisation. Yet few could deny that activities such as marriage or child rearing entail, at least in part, non-pecuniary gains and costs which may only be measurable in terms of utility. "Love," testifies Stigler, " is no zero-sum game." 1 It would be absurd to say that in traditional China love and concern played no role. For the explanation of the kind of observations encompassed here, a generalised utility analysis would yield more comprehensive results. Yet, compared with the wealth criterion, the utility criterion lacks operational simplicity in obtaining implications refutable by facts. A prerequisite for the application of the utility function is that we should be able to identify those non-pecuniary options (or their associated activities) that have utility and to specify the relative costs of acquiring them. 2 Refutable implications can then be obtained by allowing changes in situations which affect the gains and costs, with the aid of the substitution effect and the law of demand. But in specifying and assigning values to nonpecuniary options, the hazard of unwittingly making tautological statements is uncomfortably frequent. Consider, for example, foot-binding. A number of Chinese writers have attributed that practice to a desire to enhance the charm and sex appeal of women. In other words, bound feet generated utility to men. But why does the practice exist no more? The usual answer is that there has been a change in "taste." Such an argument, of course, explains nothing. For when the shifted variable (taste) is unpredictable, no theory is implied. 3 On the other hand, the explanation for foot-binding adopted in this paperthat the practice is intended to reduce the costs of enforcing property rights to a bride-yields a statement subject to refutation by facts: if bound feet
1 GeorgeJ. Stigler, " Opportunity Cost of Marriage: Comment," Journal of Political Economy, October, 1968, p. 863. B See Armen A. Alchian and William R. Allen, Exchange and Production (Palo Alto: Wadsworth Publishing Company, 1969), pp. 26-7. 8 See Donald F. Gordon, " Operational Propositions in Economic Theory," Journal of Political Economy, April 1955.


1972]


ENFORCEMENT OF PROPERTY RIGHTS IN CHILDREN


655


continue to exist after property rights in grown children have been lost, the hypothesis is falsified. The advantage of applying the wealth criterion therefore lies in the ease of evaluating pecuniary options. My reluctance constantly to employ the utility criterion left a great deal unexplained. Thus I merely accepted the observed conditions that inheritance took place gradually, that offspring were allowed to marry and to raise children, and that children were not treated purely as slaves. The propensity to love, however, is not ruled out. On the contrary, throughout sections II and III a demand function for love for one's own children is implicit. The price parents have to pay for "love" is higher under a system of enforced property rights in children, therefore in traditional China the quantity of " love" demanded by parents was relatively low. No change in " taste" is implied. Thus the explanatory power of.the analysis is in spite of, not because of, the neglect of a generalised utility theory. Professor Stigler has suggested to me that for the problem at hand wealth maximisation on a lineage basis would seem to work better than on the individual basis. With wealth maximisation by lineage the observed granting of inheritance and marriage rights to children might also be explained. But I have found it difficult to draw a dividing line at which a lineage must end; in light of the complex delineation of rights in a Chinese family, the definition of a decision unit for such a postulate might be ambiguous. In certain modified forms, however, lineage maximisation seems appealing for analysing clan formation and capital accumulation. Among the problems which I have only implicitly avoided, two are important. The first is the question why in some other societies dowries are reported to be positive. The answer, I believe, is that women were relatively scarce in China. Consider infanticide, a practice which is said to have existed in China. With this practice girl infants expected to yield poor investment returns would be eliminated. Consider further concubinage: with this practice the aggregate demand for women would be higher than if marriage was restricted to a one-to-one basis. In other words, prohibitions of infanticide and bigamy in other societies tend to reduce the economic value of women. Yet these prohibitions are inconsistent with private property rights in children. The positive dowries and different marriage customs in other societies, therefore, can be analysed by identifYing the relevant constraints of property rights. A more difficult problem, of course, is to explain why the basic structure of property rights within families differs from one society to another. A second problem implicitly avoided is why the system of enforced property rights by parents over grown children has, for the most part, vanished during the past few decades. Granted, the First Peking Draft of a new Chinese family law (1916) marked the beginning of a series of changes in law that would help to end the traditional Chinese family structure. 1 1 See Marc van def Valk, op. cit.


656


THE ECONOMIC JOURNAL


[JUNE


But it is not clear why the law changed in the way it did. As in the case of slavery in other parts of the world, it would seem to me more difficult to explain its decline than its rise. Perhaps the problem could be better understood if we ask a different question: Why did a traditional family not grow indefinitely large? Part of the answer, similar to that offered for the extent of firm size, is that at some point the unit costs of monitoring inputs rose beyond the gains from co-operation. Again, the costs of delineating and enforcing property rights increased as the structure of a Chinese family expanded. It is natural to expect that at the death of the head of a household the sons frequently would choose to split and to establish separate families. l The difficulty of explaining the decline of property rights in grown children would therefore seem to lie in specifying the changes which affect the gains and costs of maintaining those rights. In any event, the first generation who freed their children at mature age would be substantial net losers. The transition appears to have been gradual, with every succeeding generation gaining a small net wealth from the predecessor. 2 Other changes associated with declining property rights in children occur as our analysis predicts. The practices of blind marriage, bride price and foot-binding disappear; mistresses replace concubines; and prostitutes own themselves. Kowtow is an event of the past, and surviving kinship terms now serve only to confuse. The Confucian ethics are quoted only for laughter. As clearly observable in Hong Kong, the music bands that were once summoned for marriage and funeral marches, and the paperproduct shops which specialised in producing sacrificial offerings, have all but gone out of business. The temple where ancestral worship was conducted is now a place only for the ghost. 3 In view of the Chinese experience, it seems clear that marriage was originated as a form of contract for the transfer of rights in the bride. The terms of contract, while recognised by the court, were competitively determined in the market place. As noted earlier, however, current marriage practice differs in essential aspects: the children make their own marriage
1 See Francis Lang-Kwang Hsu, op. cit. The impression prevails that a traditional Chinese family also assumed the role of a firm, although the contractual relationships among firm members do not seem to exist in the family. Professor Walter Oi has suggested to me that the growth of firms outside of families might have attracted family members, and this development would tend to increase the cost of enforcing property rights in children. Hence, the family structure was weakened. 2 What effect would this change have on population growth? In modern practice, the absence of property rights in grown children implies negative pecuniary rents to the parents, which can only be compensated by the utility gains in " consuming" children. Other things equal, however, the demand for children is a positive function of the rents the parents expect to capture. The argument thus suggests that population growth in China was prompted by the capturable returns from raising children. 3 The last two sentences are perhaps slightly exaggerated. Property rights in grown children are still partially enforced in some families in Taiwan and Hong Kong. In mainland China, however, the abolition of the old family structure seems complete.


1972]


ENFORCEMENT OF PROPERTY RIGHTS IN CHILDREN


657


decisions, and the contract, under the guise of a licence, has its terms defined by law. It is difficult to believe that the demand for love and sex accounts for the survival of marriage in general. Perhaps one has to seek explanations in the gains of co-operation and interdependence between man and wife and in their joint responsibility for children. Yet a marriage contract intended for these purposes could, like any other contract, be developed in the market-if only the law would recognise and protect the free choice of terms. Much is to be gained, therefore, from a study of the implications of the costs of searching for a mate and of divorce; of the laws on taxation, adoption, and abortion; and of the arbitrary terms of contract, currently imposed on marriage by various governments, such as those concerning the sharing of property.