DEATH AND THE TRANSFER OF WEALTH: BEQUEST PATTERNS AND CULTURAL CHANGE IN THE EIGHTEENTH CENTURY.

Central European laws of succession have their origins in ancient German times when decedents had only bodily heirs--their own children--and nullum testamentum. [1] By the middle of the eighteenth century, however, wills had long been a well-established element of the legal system. [2] During this period, the rules that governed the making of wills in the Habsburg Monarchy followed Roman law in largely giving testators a free hand in their choice of heir. But these rules also contained elements that would have sounded familiar to the ancient Germans, such as strict regulations about minimum portions. These had only been introduced into Roman law relatively late.

The Austrian governments of Maria Theresa (1740-80) and Joseph 11(1780- 90) left the existing laws regulating the making of wills mostly intact, although they supported the codification of civil law of which the law of succession forms a part. In substance the law of succession remained the same, with a few exceptions concerning priests and members of religious orders. [3] Rather than forbid clauses that were deemed inappropriate by the government, the authorities tried to persuade people to exclude certain contents from their wills and to include others.

In addition to this administrative interference in bequest behavior, however successful it may have been, Phillipe Ari[grave{e}]s has suggested an autonomous change in the making of wills. Ari[grave{e}]s maintains that the functions of wills changed in the context of a transformation of family customs during the second half of the eighteenth century, [4] and that certain of these functions, particularly those concerning charitable and religious matters, became less important as mutual affection within families grew in intensity. Instead of providing for charitable bequests themselves, testators would leave these issues to their families and hope that loving spouses or children would fulfill voluntarily what previously would have had to be ordered in a will. Ari[grave{e}]s argues that the hitherto wary relationship between testators and heirs became trustful in the course of the eighteenth century, and that it would increasingly have been considered unbearable to lay down explicitly and by legal act what woul d have been performed out of mutual love anyway.

Bequest behavior in the eighteenth century thus raises a number of questions concerning loyalties and sentiments within families and the impact of political change on everyday culture. This paper addresses these questions as follows: In the first section we describe the sources used in this analysis. The following four sections discuss the main factors that determined whether bequests to relatives, friends, other persons and charity were made in eighteenth-century wills; these factors are testators' professional status, their family status, the formal characteristics of wills, and large-scale changes in popular culture in the political context of religious and church reforms. The penultimate section examines the ways in which bequests to various recipients, were interrelated in addition to the factors discussed previously. In the final section we summarize the results.

This analysis is based on a sample of Upper Austrian wills from the period 1700 to 1820; the sources come from urban as well as rural settings and cover all strata of the Upper Austrian population. [5] Apart from wills, they include other documents of probate proceedings, which were necessary for inventorying the wealth of deceased persons and distributing it among heirs and legatees. These proceedings were carried out by manorial and municipal courts. In principle, obtaining probate was mandatory in any case of death, although the proceedings were summarily closed when the deceased had left no wealth. Apart from probate inventories, which were necessary for determining the amount of wealth the deceased had left, wills were included in the documents because they determined the distribution of the heritage among heirs and legatees. In some series of documents, however, wills and inventories were stored separately, and not all of them can be linked successfully.

* * * * *

A major part of the original sources was destroyed deliberately or has been lost accidentally. For this reason, complete series from certain periods or all sources of certain courts have disappeared, and the remaining series are often incomplete. However, it does not seem that documents were discarded systematically according to criteria that are crucial to our analysis, for example, the contents of wills or the level of wealth of the testators. It should be stressed that for obvious reasons our analysis includes only those cases where wills were left, whereas cases of intestate succession have been omitted, although even in these latter cases the decedent implicitly acknowledges a certain legal succession, namely that provided for by law. [6]

Originally the sources used in this analysis were stored in manorial and municipal court archives all over the crown-land. Today they are kept in the Upper Austrian State Archive (Ober[ddot{o}]sterreichisches Landesarchiv). The sample established thus far includes 2,828 cases. [7]

Wills like the ones used in this paper have gained some importance in research on popular culture. [8] They are a good source for the history of popular culture, especially of religion because they cover a large part of the population from the late seventeenth century on, even propertyless persons and persons who left no other written records. In addition, they normally make available crucial data that reveal the environment and personal characteristics of the individuals involved. Religious behavior as reflected in these documents includes religious rhetoric as well as preferences for liturgical functions and charitable clauses. However, wills were clearly not written specifically as documents of religious self-definition, but served numerous other functions, such as distributing one's wealth among surviving relatives and friends, defining one's place in a community, or giving one's view of the world. The contents of wills were influenced by these purposes as well as by the conditions under which the testat ors lived, and by accidental circumstances connected with the act of making a will. This diversity of contents and external factors offers the chance of obtaining insights into the functions of pious bequests and bequests in general. Our understanding of bequest behavior increases if we compare various kinds of bequests such as bequests to family members or friends or pious bequests: for instance, pious bequests may constitute an autonomous sphere showing no association with one's behavior toward one's personal environment, or, on the contrary, may be systematically connected with other bequests. In this paper, we distinguish between several alternatives, namely bequests to spouses, children, other relatives, godchildren, servants (including household servants and farm-hands), and other unrelated persons, and pious bequests. We ask which of these alternatives were chosen by testators, which factors determined the choice, and in what way secular and religious bequests were interrelated.

Table I gives an insight into who profited from different kinds of bequests in wills in our sample. We list the proportions of wills which contain inheritance rights and legacies [9] received by spouses, children, relatives, friends, servants, godchildren and charitable recipients. We see that inheritance rights were most frequently given to spouses, although only about half of the sample testators were married at their death. Among married persons, two thirds left their fortune to their spouses. Widowed persons mostly favored their children; unmarried persons favored relatives. The situation becomes more complex for recipients of legacies. Pious clauses are present in about 60 percent of all wills in the sample. A third of all wills contain legacies for relatives; legacies to servants, friends or children appear fairly frequently as well. Again, bequest patterns differ according to the marital status of the testators.

These numbers suggest that bequest patterns were influenced by a number of factors that determined who became a heir and who received a legacy. Among these factors, we distinguish between social and political factors such as professional and family status, secular changes in behavior, and factors connected with the circumstances of the declaration and the juridical form of wills.

The main results for the whole sample and a subsample are given in Tables 2 to 5. The tables display results of probit analyses in which legacies to specific recipients and inheritances of these recipients are modeled as dependent dummy variables, and indicators of circumstantial factors and personal characteristics of testators are regarded as independent variables. The right-hand part of Table 4 shows the results of a tobit analysis in which the amount of bequests for pious purposes is the dependent variable. [10] For bequests to family members, we use the variables Spouses and Children; for other relatives, the variable Relatives. In Table 2, Friends includes bequests to godchildren, servants and other unrelated persons; in Table 3, bequests to Godchildren and Servants are measured separately and are not included in Friends. For pious aims, we use a collective term although it was quite common for single testators to make bequests to several churches and monasteries, a number of fraternities, several hosp itals and so on; [11] for the purposes of this paper, we treat all these bequests as if they belonged to one category. In the tables, they appear as Pious inheritance and Pious legacies and include inheritance rights and legacies, respectively, for hospitals, almshouses, orphanages, poor and sick people in general, religious brotherhoods, churches, monasteries, and the celebration of Masses.

In order to get a more precise picture of how obligations toward family members, relatives, friends and charitable aims interacted, in Tables 2 and 3 we restrict the sample to those persons who had been married at any time in their lives. Since unmarried persons could not possibly leave anything to a spouse or to legitimate offspring, their inclusion in the sample would distort the results concerning the relative importance of different recipients as named before.

* * * * *

Which factors determined whether testators named a spouse, a child, a relative or a friend as heir, and which factors determined the likelihood of legacies for spouses, children, other persons and charitable aims?

Wealth, profession, education and status determine both testators' mentality and their financial scope. Education can be expected to exert a strong influence on religious attitudes because it determined how people came into contact with and reacted to current and new ideas; unfortunately, direct information about the education testators had received is lacking in most cases. Similarly, the data on the wealth of testators are not entirely satisfying: although in many cases exhaustive information is given about their wealth and its composition, in many other cases this information is lacking completely. Therefore, using wealth data in analyses would lead to the exclusion of a major part of the data set. However, both wealth and education are closely related to profession, which is well-documented and can thus be used as a substitute for those variables. In order to limit the number of professional groups, we distinguish between priests, free professionals, farmers and lower class persons (located in all sector s); the reference group includes mainly artisans and officials. In addition, we examine the importance of noble status: although aristocrats could live in quite varied circumstances, many of them were either officials or without profession (living off their private means), and many belonged to the more affluent parts of society.

Professional and status factors had some impact on the choice of both heirs (Table 2) and legatees (Table 3), although in different ways. Some of the differences between inheritances and legacies may be explained by a tendency to give either a portion of one's estate or only a legacy to someone, but not both (in principle, a person could of course receive a legacy and be an heir to the estate at the same time).

This is true of bequest behavior among the nobility: persons of nobility are less inclined to leave their wealth to their spouses than to their children. Obviously, persons of nobility were careful to keep their wealth for their own offspring in case a surviving spouse might leave part of their wealth to persons other than the testator's children, such as children from a previous or a subsequent marriage, or other relatives. [12] However, although an aristocrat's spouse was less likely to receive an inheritance the same person had no worse chances of receiving a legacy--compared with other groups, aristocrats show no significantly different behavior with respect to legacies left to spouses. Thus persons of nobility provided their spouses with some means, but they wanted to control the amount of wealth possibly transferred to their spouses' lineage. On the other hand, an aristocrat's child, being the heir anyway, was significantly less likely to receive a legacy in addition than were the children of artisans and officials. Regarding other differences between professional groups, we see that within families, professions are mostly irrelevant to the probability of a certain behavior. The farmers here form an exception; compared to the reference group (artisans and officials), farmers who chose an heir show a specific pattern in deciding between spouses and other persons: they are more likely to appoint friends and less likely to appoint their spouses. It seems that farmers tended to connect matters of inheritance with the retirement arrangements for themselves and their spouses, which relatively often involved friends. Farmers who wanted to retire usually handed over the major part of their property to other persons and received a life-annuity in compensation; sometimes their contractors in these deals additionally became their heirs even if they were not relatives. [13]

The situation changes where bequests to friends and especially to servants are concerned. Relations between servants and masters must have been of a peculiar nature indeed, as more affluent testarors, such as free professionals and the nobility, show a strong inclination to remember their servants in their wills, whereas farmers and lower-class persons are disinclined to do so. Clearly the circumstances of farm-hands and household servants and their relationship to their respective masters were quite different. Servants of the nobility enjoyed the most extraordinary position: in Table 3, serving in an aristocrat's household is the most important factor determining whether a servant received a legacy. This paternalistic behavior on the part of the aristocracy worked for the benefit of other unrelated legatees as well, although less strongly: persons of nobility are still more likely than other people to give to unrelated persons who are not their servants, whereas farmers are significantly less prepared to do so. Concerning paternalism, the nobility seems to have set a standard for other upper-class groups like free professionals as well. [14]

Among professional and status groups, we find also some significant differences in attitudes to charity (Tables 2 and 4): lower-class people in all sectors, as well as farmers, were less inclined to make pious bequests than were the artisans and officials who constitute the reference group, while priests show significantly more willingness to provide for those matters. The values of pious bequests too are lower for lower-class people and farmers, and higher for priests. Persons of nobility and free professionals do not differ from artisans and officials with respect to the frequency of pious legacies in their wills; however, the value of their bequests is significantly higher, reflecting the fact that aristocrats and free professionals usually had more to give.

Marital status and family structure are important marginal conditions for bequest behavior. Married persons have the option to give to their spouses, and parents are likely to give to their children. Bequests to other persons and charity had to compete with these aims, and they may be different if no spouses or children are present. In part, marital status and the family structure in general are associated with profession, sometimes on legal grounds, as is the case with Roman-Catholic priests. The available information about the marital status of testators seems to be quite reliable; in this analysis, we distinguish between married, widowed and unmarried persons. Information about the number of children still alive is probably less accurate, because in about a third of our sample we must rely on the texts of the wills themselves as a source for this information and these texts do not necessarily mention all living children.

Family structure influenced bequest behavior both in the secular and in the religious sphere. Apart from somewhat trivial results, [15] we see how marital status and the number of children determined in which way testators distributed their wealth within their family. Children competed with spouses-the more children testators had, the lower the probability that they would appoint their marriage partners as heirs. This result is not trivial, because married testators who had children might be expected to distribute their wealth among their children and their marriage partners together. On the other hand, the number of surviving children, while lowering a spouse's chances for the inheritance, raised the same spouse's chances of receiving a legacy. However, the children's prospects grew better in both respects: the number of children is positively connected with both inheritances and legacies given to children.

Family structure mattered in the religious sphere, too, showing a highly significant independent influence on charitable bequests--even unmarried people who were not priests were more likely than married people to give for pious purposes, and the more children testators had, the lower the probability that they would leave pious legacies (Table 4). It seems that charity had to compete with family needs, which is why testators who had no family could better afford to give to the church and the poor. This interpretation seems likely if we look at the values of these bequests: testators who had families were less likely to give anything to the church or the poor, and the value of what they gave was lower (Table 4).) If we confine our analysis to those testators who gave to charity, we see that the number of children they left has a highly significant negative effect on the value of these legacies. [16] Obviously these testators did not just trustingly leave charitable spending to their heirs as Philippe Arias ma intained; otherwise they would not have made any charitable bequests at all. Rather, they wanted to be cautious in order to preserve their children's wealth and therefore limited charitable spending more strictly than persons who had no family.

Wills, though requiring certain minimal formal standards in any case, were not uniform acts. There are at least four criteria to distinguish among wills of different kinds: the distinction between testaments and codicils, the differences between written and oral wills, the functions of reciprocal wills, and the role of witnesses.

The first distinction is made between testaments and codicils: wills may nominate one or several heirs to the estate, i.e. persons who will receive all of the testator's belongings that are not explicitly left to someone else; in this case, the will is called a testament. A will nominating no heir, but only legatees, i.e. persons who will receive certain objects or rights, is called a codicil. A person leaving only a codicil implicitly acknowledges intestate succession for all belongings that are not left to legatees. Although in principle, testaments and codicils need not differ in making bequests to different groups of legatees, in fact we observe such differences (Tables 2 and 3). Codicils usually concentrate on the persons closest to the testators such as spouses and children, who receive legacies significantly more often in codicils than in testaments. On the other hand, testaments are significantly more likely to include pious bequests.

Secondly, we distinguish between different kinds of written and oral declarations: wills may be written either by the testators themselves, in which case they are called holographic wills, or by other persons, or they may be made orally. Oral wills must be spoken in the presence of other persons, for instance neighbors, which creates a situation completely different from the privacy of writing with one's own hand without any witnesses present. These different forms of wills were not independent of social factors: holographic wills clearly required the ability to write, which was not common in rural areas even several decades after the introduction of compulsory education in the 1770s. Although oral wills were not completely unknown among educated persons, they were more common in the lower classes and the rural population: among farmers and lower class persons, about half of the testators made their wills orally, whereas among free professionals and priests as well as the nobility, 80 to 90 percent left writ ten wills. Wills that were written by the testators themselves or spoken in the presence of other persons, show specific bequest patterns compared to wills written by other persons (Tables 2 and 3). For spouses, both holographic and oral wills have a significant positive effect on their likelihood of receiving an inheritance, and a negative effect on their chances of receiving a legacy; for children, oral wills have the opposite effect. It seems that oral wills were often spoken in the presence of the testators' spouses, which may have raised the spouses' chances of receiving everything that was not explicitly left to legatees. Concerning charity, oral wills are less likely, and holographic wills are more likely, to include pious bequests. Thus, privacy in making one's own will favored pious behavior.

Thirdly, we must consider the importance of reciprocal declarations: wills may be made by just one person or by two or even more persons together. Wills made by more than one person usually stipulate mutual rights, for instance, the husband is appointed heir by his wife and vice versa; these wills are called reciprocal wills. Reciprocal wills were made almost exclusively by married persons who often had no children, although we know of a few such wills left by siblings. A reciprocal declaration is likely to restrict the contents of a will to the issues that seem necessary for an arrangement between the persons involved, i.e. the distribution of the couple's wealth between the surviving spouse on the one hand and the relatives of the pre-deceased spouse on the other. [17] Other issues may be left for subsequent declarations or to the discretion of the survivors. In fact, reciprocal wills are significantly less likely than other wills to include bequests to children, godchildren, servants, friends or charity ( Table 3).

Fourthly, not only oral wills but also all other wills except holographic ones required the testimony of two or three persons. Although witnesses were not necessarily told the contents of written wills (they simply signed the documents), some witnesses such as lawyers and notaries were usually involved in writing down or even wording the text. The same may be true for other persons who had some professional relationship with people who were on the brink of the grave, particularly physicians and priests. Wills witnessed by physicians, clerics or lawyers generally reveal no significant preferences in the choice of heirs (at least not at the 5 percent level of significance), but show some specifics in making legacies, especially to persons outside the family, who are more likely to receive legacies when certain witnesses were involved (Tables 2 and 3). [18] These results seem consistent: witnesses had no reason to suggest that a distant relative should become heir to the estate, but they may have reminded testa tors to include legacies for their servants and other distant persons.

In 1771, the authorities restricted the participation of priests in drawing up wills because it was felt that they might exert undue influence in their own interest on the provisions made in those wills. Priests were forbidden to write down wills for testators who were unable to write, and regulars were even forbidden to act as witnesses. [19] As a matter of fact, wills witnessed by priests are more likely to contain charitable clauses, and thus the fears of the Theresian government seem to have been not entirely unfounded (Tables 2 and 4). But this interpretation does not fully explain the phenomenon, since lawyers seem to have worked in the same direction: legacies for pious purposes were, ceteris paribus, more frequent and more valuable in wills made with the assistance of lawyers. Charitable bequests just seem to follow the same rules as bequests to servants, friends and relatives.

Generally speaking, we can distinguish on the one hand between wills that served specific purposes or were made rather quickly, and full and solemnly declared wills on the other. This distinction is crucial for the explanation of bequests in favor of distant persons and charity. Codicils and reciprocal declarations were wills that served specific purposes, oral wills were made quickly; these wills do not use the full repertoire of rich rhetoric and various contents that is typical for a solemn baroque will, but concentrate on the core issues. Solemn wills, on the other hand, show a taste for extravagant rhetoric and a marked preoccupation with doing what was considered proper, including posthumous acts of charity and favors to distant persons. Full wills were often written well in advance, and frequently by the testators themselves, which suggests that pious behavior was deeply ingrained and not merely a result of the undue influence of professionals like lawyers or priests; however, other solemn wills were indeed written with the assistance of lawyers and priests, and it may well be that these professional administrators reminded testators to have charitable clauses included in their wills.

Apart from variations due to the specific situation of individual testators, we observe large scale changes in bequest patterns that happened in the course of decades. The time factor shows some influence both on bequests within the family and on charity, It has a negative effect on the probability of inheritance between spouses, and a positive impact for children (but not relatives) and for friends. These effects become somewhat clearer if we compare inheritances and legacies (Tables 2 and 3). For legacies, the time factor works differently, favoring spouses and putting children (and relatives and servants) at a disadvantage. It seems that in the course of time testators began to reverse their bequest behavior toward their families, increasingly choosing their heirs among their children and giving only a legacy to their spouses. This means that neither spouses nor other family members were completely forgotten in any period examined for this study, but there was a change in the way assets were measured and handed over to them-when all bequests to marriage partners are put together, i.e. if we use a dummy for inheritances and legacies together, the time factor no longer has a significant effect on the results. [20]

Changes over time had a more pronounced impact on pious behavior, with secular changes in the attitudes toward church and religion acting independently of individual circumstances. In early modern times up to secularization, and to a lesser extent even later, Roman Catholics preserved traditional views on the usefulness of charity, [21] which explains the occurrence of charitable bequests in wills in all parts of the Austrian population during that period. However, in the eighteenth century, Austria and other countries like France and the Catholic parts of Germany [22] showed a considerable change in religious attitudes: baroque piety, which was marked by the sensuous and theatrical expression of religious beliefs, by strong contrasts and ostentation, gave way to the more sober Reform Catholicism. [23] In addition, a small portion of the population even showed signs of dechristianisation, which means a complete disappearance of religion from certain spheres of life that had been marked by religion previously . The decline of baroque piety began in the 1760s, and the transformation period lasted some thirty years.

The transformation of mass mentality coincided with the aims and measures of governmental church policy in the same period. Under Maria Theresa and Joseph II, Austria pursued a strict course of reform in many fields, including, for instance, health policy, education, agriculture, commerce, and civil and criminal law. In this context, religion and the Catholic church were intended to serve as means of reform in fields that were not inherently associated with spirituality and Catholicism. The government regarded an appropriate religion as a prerequisite for a virtuous and productive life, and thus as necessary for everyone. According to reform-minded writers and bureaucrats, proper piety meant a kind of spirituality that concentrated on the core issues of Catholicism and abandoned those seemingly shallow and incidental exercises that were typical of baroque piety.

Although the government cannot be made responsible for the changes in mass behavior [24] the new ways of religious life in the population nicely matched the political aims of the administration. Decreasing bequests for purposes like chapels, altars or monasteries, which were regarded as useless by the bureaucracy, were in line with the government's program; and although holy Mass was regarded as valuable by most influential bureaucrats, cuts in private spending for Masses were still regarded as desirable. An explanation of the transformation of mass mentality must take into account these programs and considerations, which are characteristic of Austrian cultural history in the second half of the eighteenth century.

Religious behavior changed during the second half of the eighteenth century, with Reform Catholicism taking the place of baroque piety. We represent the decline of baroque piety by the factor Time, which is a transformation of the calendar year; it was specified to show little variation up to the 1760s, a monotonous decline in the decades centering around 1785, and little variation from about 1800 on. [25] Large-scale changes in mentalities, expressed by the variable Time, show the most significant influence on charity: ceteris paribus, charity becomes less frequent during the reform period and remains at a low level afterwards. Even more significant is the connection between the time factor and the value of pious legacies; pious bequests become not only less frequent but also less valuable in the course of the period under consideration (Table 4).

The impact of time on testamentary provisions should not be interpreted as a shift from charitable acts ordered in wills to charitable acts done voluntarily by loving spouses and children, as has been suggested by Philippe Ari[grave{e}]s. The time factor applies regardless of family structure, among married as well as unmarried persons and among people who had children as well as those who did not. This suggests that the Ari[grave{e}]s hypothesis does not offer an exhaustive explanation for the shrinking number and value of charitable legacies in the eighteenth century. Otherwise the time factor would not show any influence on the wills of unmarried and childless persons, since these persons did not have any spouses or children to make charitable donations in their memory. In fact, a separate analysis of charitable bequests in the wills of unmarried persons shows that, during the eighteenth century, charitable and pious bequests became less frequent and less valuable in the wills of these persons as well; it is the most important of all the factors used in the analysis (Table 5).

Now, was there any connection between inheritances and legacies of different kinds? The results presented thus far suggest that the independent variables worked in specific ways for different kinds of recipients. We may assume that at least four factors influenced bequest behavior during the period under consideration.

The first factor is the situation in which the will is made. Holographic wills and wills drawn up with the assistance of professional experts such as lawyers and clerics were made at times when the testator had leisure for long and careful consideration. In these more thoughtful documents, testators did not concentrate exclusively on the central issue of providing their nearest relatives with the necessary means, but included incidental matters like gifts for friends, servants, and pious institutions as well. Oral wills, on the contrary, were mostly made at a time when testators worried about the most important things and did not care as much about small gifts for friends; thus these documents often confined themselves to the basic issue of appointing an heir.

The second factor is related to the first. Wills are not necessarily comprehensive regulations for everything to be done after the testator's death, but may sometimes resemble a settlement among the living, similar to a marriage contract. Typically, reciprocal wills were made by husband and wife together, often prior to the birth of common children; thus they provide for the rights of relatives on either side and seek a compromise between the rights and needs of the spouse on the one hand and the relatives on the other. These concerns do not entirely preclude stipulations of religious acts, pious legacies and so on, but such issues are of secondary importance.

A third factor is the difference among professional groups, which is more pronounced in matters that do not concern the family. More affluent groups were more likely to invest large sums for religious purposes: free professionals, persons of private means, and clerics spent more, farmers and the lower classes spent less than artisans and officials. Persons of nobility too, who often lived off private means, were likely to spend more for religious purposes. And we have already observed that the upper classes were more likely than middle and lower class persons to leave legacies to their employees.

The time factor has a detrimental effect on religious behavior and works in various ways with respect to family, relatives and friends. The interpretation of this factor is easy with respect to the religious indicators, where Austrians translated the new ideas of a purified, reformed Catholicism into action. For other persons, changes in bequest behavior over time appeared often as changes in the way bequests were made, rather than in the substance of the bequests. For instance, restators continued to favor certain groups of people, but increasingly gave legacies instead of inheritances to them or vice versa.

Thus bequest behavior was obviously not uniform towards family, friends, and religious purposes. But the question remains whether there is still systematic covariation in the unexplained remainder of bequests to spouses, relatives, servants and religious institutions. Such covariation among several items of behavior would be independent of time, social factors, and other factors as specified in the previous sections. We conjecture that, independently of circumstantial factors, bequest behavior showed a consistent pattern in that testators did not make their choice randomly among different groups of legatees and heirs, but rather tended systematically to combine gifts to specific groups with gifts to specific other groups. The best way to test for this hypothesis seems to be a test for correlation of the residuals of the dependent variables in Tables 2 and 3.

Results are given in Tables 6 and 7. They suggest that there are indeed still correlations among different kinds of bequests apart from what has been explained by the models discussed so far. We can observe two situations where bequest decisions take a clear shape:

The first issue about which testators had to decide was how to treat their spouses compared to other people, especially their children. As far as inheritance rights are concerned, testators could choose between appointing several heirs together or only one. For married testators who had children, the law of intestate succession would have allowed for spouses and children to become joint heirs. We see that persons who made a testament chose exactly the opposite way: they usually gave the full inheritance either to their spouses or to other persons (generally to their children). The coefficients of correlation are significantly negative for any combination of spouses on the one hand and children, relatives and friends on the other. It should be stressed that this effect is independent of the marital status of the testators. Concerning legacies, the correlations between gifts for spouses and other recipients are not significant at all. Thus people who made their wills were not merely going through ritual motions but rather wanted to arrange things in an individual way that differed from what was provided for by law.

The second issue was how to deal with a host of potential beneficiaries outside the family. Testators could choose among people whom they knew personally, such as relatives, servants and friends, and people who might profit from their gifts by way of a charitable or pious institution. The question is whether testators picked legatees among these groups randomly or systematically. The results suggest that legacies were made consistently either to all extra-familial groups alike or to none of them. Persons who made bequests to charity usually favored relatives, godchildren, servants and friends as well; persons who favored relatives tended to give also to servants and friends, and there is also a correlation between bequests to servants and friends. This is true of legacies, but not of inheritance rights; clearly, for married persons, inheritance was mostly an internal family matter and had little to do with persons outside the family. Again, these effects work independently of the family status of the testato rs.

This latter result is particularly interesting with regard to the history of mentalities in the eighteenth century. Charity was a particularly important manifestation of what we may call altruism--working for the benefit of people outside one's family. But charity, though important, was only one among several manifestations of altruism. Together with favors done for relatives, servants and friends, it formed a coherent pattern of behavior that cannot be interpreted as specifically religious or pious, but rather as altruistic.

Eighteenth-century wills show a great variety of bequest patterns. In the previous sections, we have shown that a major part of this variety can be traced to determining factors like the professional background of testators, their family environment, the way wills were made, and long-term changes in mentalities. These factors influenced bequests to persons inside and outside the family and to charitable institutions.

Apart from these mechanisms, however, we can still observe systematic correlations in the favoring of different groups of recipients. Concerning inheritance rights, it seems that, regardless of their personal characteristics and accidental circumstances, testators had to decide between their spouses on the one hand and all other potential heirs on the other hand. A joint inheritance of spouses and children or other heirs, while occurring occasionally, was not the predominating state of affairs.

Concerning legacies, we see that testators who favored persons outside the family usually favored pious institutions, friends, relatives, and servants in equal measure rather than randomly selecting recipients from one of these groups. Thus charitable legacies did not form a group entirely separate from other recipients outside of testators' families, but rather profited from the same kind of altruistic behavior that worked in a favorable way for other recipients as well.

This is not to say that charitable legacies lacked any religious background. They clearly played a special role in eighteenth-century wills, as evidenced by their frequent occurrence and the emphasis testators laid upon them. In becoming less frequent in the second half of the eighteenth century, they reflect changes in popular mentality which are characteristic of several continental European societies.

But religion was not the only aspect relevant to charitable behavior. Charity as manifested in wills was influenced not only by a long-term change in religious attitudes and altruistic motives, but also by the family status of testators. Those who had spouses and children restricted charitable expenditures in order to preserve their family's wealth.

Eighteenth-century wills provide one of the rare opportunities to hear the voices of ordinary people of the time, often in their own words. An examination of the wills reveals that these documents, while reflecting the external circumstances of their making, are far more than ritual declarations without any connections to their authors' inner beliefs. In fact, in many ways they provide us with our clearest picture of the ideas of the people of the time.

ENDNOTES

This research has been supported by the Austrian Science Foundation, Grant No. P6317. The author wishes to thank Peter Solar, John Komlos and participants in the XIIth International Conference of the Association for History and Computing (Glasgow, June 30, 1997) and the IIIrd Conference of the European Historical Economics Society (Lisbon, October 30, 1999) for helpful comments.

(1.) Tac., Germ. cap. 20.

(2.) For the law of succession in Austria, see Gunter Wesener, Geschichte des Erbrechts in [ddot{O}]sterreich seit der Rezeption (Graz/K[ddot{o}]ln 1957).

(3.) For instance, the participation of these persons in making other persons' wills was limited, and members of orders were granted only limited scope in making their own wills. See, for instance, Verordnung, 4. September 1771, Patent, 30. August 1782, Sammlung der Kaiserlich-K[ddot{o}]niglichen Landesf[ddot{u}]rstlichen Geserze und Verordnungen in Publico-Ecclesiasticis vom Jahre 1767 bis Ende 1782 (Wien 1782), pp. 41-42, pp. 136-137).

(4.) Philipp Aries, Geschichte des Todes (M[ddot{u}]nchen 1982), pp. 596-599.

(5.) Upper Austria, which is part of the Republic of Austria today, was one of the crownlands of the Habsburg Monarchy. During the eighteenth century, its population was about 625000 persons.

(6.) For a further examination of the methodological problems connected with wills, see Michael Pammer, Glaubensabfall und Wahre Andacht: Barockreligiosit[ddot{a}]t, Reformkatholizismus und Laizismus in Ober[ddot{o}]sterreich 1700-1820 (M[ddot{u}]nchen/Wien 1994), pp. 34-77.

(7.) We used the following archival sources for sampling: Ober[ddot{o}]sterreichisches Landesarchiv (O[ddot{O}]LA) Alte Statthalterei, 1-17. O[ddot{O}]LA, Landesgerichtsarchiv: 5-16 (Landeshauptmannschaft), 32-49 (Landrecht), 77-103, 106-110 (Stadtrecht), 530 (Aichberg), 532 (Aigen), 533-535 (Aistersheim), 540 (Altm[ddot{u}]nster), 543-544 (Aschach), 550-552 (Aurolzm[ddot{u}]nster), 565 (Ebenzweier), 606 (St. Florian), 626 (Freistadt), 632 (Pfarrhofund Dek. Freistadt), 633 (Kirchenamt Freistadt), 634 (Schulprovisoramt Freistadt), 649 (Freiwald), 655 (Gleink), 663 (Landgut Oberwies), 667 (Gramastetten), 673 (Pfarrhof Grieskirchen), 683 (Hagenberg), 685 (Rigading), 686 (Hall), 697-698 (Markt Hall), 702 (Kremsm[ddot{u}]nster), 708-709 (Markt Haslach), 725 (Holzheim, Weikertsberg), 728 (J[ddot{a}]germayrleiten), 729 (PfarrhofKalham), 735 (Spital a. Pyhrn, Starhemberg. Lehen WeilBenberg), 743-744 (K[ddot{o}]ppach), 751 (Kremsm[ddot{u}]nsrer), 769-771 (Stift Lambach), 796 (Landshaag), 797 (Landshaag),800 (Spiralamt Freistadt, St. Peterskirchenamt Freistadt), 802 (Leonfelden), 812 (Lichtenegg), 817 (Losensteinleiten, Trinit[ddot{a}]ts Benefizium Hall), 820 (Luftenberg), 823 (Lustenfelden), 829-837 (St. Martin), 881 (Mondsee), 887-890 (M[ddot{u}]hlwang), 896-897 (Neuhaus), 905 (Benefizium Peuerbach), 906 (Markt Neumarkt, Hofkirchen), 908 (Oberneukirchen), 911-912 (Traunkirchen), 917-918 (Ottensheim), 922-923 (Markt Ottensheim), 925 (Parz), 947 (Puchberg), 958 (Kirche Ried, Windischgarsten), 961 (Pulgarn), 973 (Magistrat Ried), 977 (Pfarrhof Kematen, Pfarrhof Pfarrkirchen), 980 (Riedau), 987 (Pfarrhof Aistersheim), 993 (Sch[ddot{u}]rding), 1006 (Schaumburg), 1007 (Schiferisches Erbstift), 1015 (Schl[ddot{u}]sselberg), 1020 (Schwanenstadt), 1026-1027 (Sierning), 1032 (Pfarrhof Waldneukirchen, Kirche Sipbachzell, Kirche Viechtwang, MarktKirchdorf), 1033 (Spielberg), 1035 (Spielberg), 1040 (Starhemberg), 1043-1045 (Steinbach), 1050 (Steinbach), 105 1-1058 (Magistrat Steyr), 1074 (Steyr), 1082 (Gotteshaus Sierning, Gaflenz), 1082 (Ostermiething), 1084 (Steyregg), 1092 (Sunzing), 1102 (Vichtenstein), 1116 (Almegg), 1128 (Waxenberg), 1130 (Waxenberg), 1140 (Weidenholz), 1150 (Weissenberg), 1152 (Weissenberg), 1156 (Magistrat Wels), 1167 (Weyer), 1173 (Wildberg).

(8.) See, for instance; Michel Vovelle, Pi[acute{e}]t[acute{e}] baroque et d[acute{e}]chrisdanisation en Provence au XvIIIe si[grave{e}]cle. Les attitudes devant la mort d'apr[grave{e}]s les clauses des testaments (Paris 1973); Philipp T. Hoffman, Church and Community in the Diocese of Lyon 1500-1780 (New Haven, CT, 1984); Rudolf Schkgl, Glaube und Religion in der S[ddot{a}]kularisierung. Die lcatholische Stadt-K[ddot{o}[ln, Aachen, M[ddot{u}]insrer-1700-l840 (M[ddot{u}]nchen 1995); Leslie M. Moscow, "Charity and the Bequest Motive: Evidence from Seventeenth Century Wills," Paper given at the 36th Annual Cliometrics Conference, Nashville, Tennessee, May 17-19, 1996.

(9.) An inheritance right of one of the recipients mentioned in the tables is the claim to either the overall estate of the testator or a certain quota of this estate (regardless of the extent of the estate). A legacy left to one of the recipients is the claim to a certain thing or a specified sum to be taken out of the heritage.

(10.) A linear regression would nor be the appropriate method for such an analysis because the variable is censored at zero, since bequests are always positive numbers.

(11.) For details, see Pammer, Glaubensabfall.

(12.) Similarly, compared with artisans and officials, free professionals show a significantly stronger tendency to leave their wealth to their children.

(13.) In addition, farmers are less likely than other professional groups to leave legacies to their children.

(14.) Free professionals followed aristocrats in leaving legacies both to friends and to servants, although they did not match the nobility's activities entirely.

(15.) E.g., according to our data, testators with children are less inclined to look for an heir among relatives and friends and are more likely to leave their wealth to their children. Widowed persons are less likely to leave their wealth to a spouse, and more likely to leave it to children, relatives or friends (Table 4).

(16.) These numbers do not appear in the tables.

(17.) Therefore reciprocal wills favor inheritance rights of spouses and relatives but not of other groups of recipients.

(18.) Wills witnessed by a priest are more likely to contain legacies for relatives, friends and servants. For relatives and servants, the same is true when lawyers were the witnesses. The presence of physicians as witnesses had no influence on the wills.

(19.) Secular priests, however, were allowed to do so. Verordnung, 4. September 1771, Sammlung, pp. 41-42.

(20.) Mutatis mutandis the same is true for relatives and friends which suggests that the economic crisis in the second half of the eighteenth century did not wipe our bequests outside the family entirely. For the economic development of Austria in the eighteenth century, see; John Komlos, Nutrition and Economic Development in the Eighteenth-Century Habsburg Monarchy. An Anthropometic History (Princeton 1989).

(21.) In the early period after Christianization, charitable bequests may have been seen as God's share of the estate; cf. Harold J. Berman, Recht und Revolution: Die Bildung der westlichen Rechtstradition (Frankfurt 1991), pp. 378-379. For the time from the "birth of purgatory" on, it seems clear that charitable acts were intended as a means of mitigating the punishment that awaited sinners in the other world; Jacques LeGoff, La Naissance du Purgatoire (Paris 1981).

(22.) See Vovelle, Pi[acute{e}]t[acute{e}] baroque; Schl[ddot{o}]gl, Glaube und Religion.

(23.) Reform Catholicism favored the modest expression of one's religious beliefs and the internalization of Catholic values instead of ritualized exercises. In addition, it allowed the instrumentalization of religion for the practical needs of the state: the Catholic church had developed a network of dioceses and parishes that covered the whole territory of the Habsburg Monarchy and was regarded as a potentially effective instrument of local government. Many of the reforms of enlightened absolutism in Austria relied at least partly on parishes for their implementation. Therefore both ideological issues and practical exigencies contributed to making the church and religion a matter of primary importance to politics.

(24.) Mass behavior changed prior to the reform decrees issued by Maria Theresa and Joseph II.

(25.) The transformed time variable [TIME.sub.tr] is calculated according to the following formula:

[TIME.sub.tr] = arctan(0.33 * (1785 - YEAR )),

where YEAR is the calendar year of the testator's death. For the transformation see Michael Pammer, "Modeling Religion. Bureaucratic Reform and the Transformation of Popular Piety in the 18th Century," Historical Social Research 19/4 (1994): 4-25, pp. 9, 18.