LK INSHIP AND CLASS: A STUDY OF THE WEYERHAEUSER FArm_!_J by rtiARVIN GLENN ~ l A DISSERTATION Presented to the Department of Sociology and the Graduate School of :he University of Oregon in partial fulfillment of the requirements for the degree of Doctor of Philosophy March 1977 ii APPROVED: (Albert ~ Sz~~) iii Copyright © 1977 by Marvin Glenn Dunn iv VITA NAME OF AUTHOR: Marvin Glenn Dunn PLACE OF BIRTH: Madison, Wisconsin DATE OF BIRTH: March 14, 1944 UNDERGRADUATE ANO GRADUATE SCHOOLS ATTENDED: Wi s cons i n S ta te Uni ve rs i ty -R i ve r Fa 11 s Garrett Theological Seminary Northwestern Uni ve rs i ty University of Oregon DEGREES AWARDED: Bachelor of Arts, 1966, Wisconsin State University-River Falls Master of Arts, 1970, Northwestern University AREAS OF SPECIAL INTEREST: Social Stratification Sociology of Religion Sociology of Development (Latin America) Politi ca 1 Sociology PROFESSIONAL EXPERIENCE: Teaching Assistant, Department of Sociology, University of Oregon, Eugene, 1970-1971, 1972-1973 Instructor, Department of Sociology, University of Oregon, Eugene, 1973-1974 Instructor, Department of Sociology-Anthropology, Whitman College, 1975-1976 Visiting Assistant Professor, Department of Sociology-Anthropology, Lewis and Clark College, 1976-present V AWARDS Mm HONORS: North Central Jurisdictional Fellowship, 1966-1967 National Science Foundation Fellowship, 1971-1972 vi ACKNOWLEDGEMENTS Many people have helped to make this dissertation possible. would 1 i ke to express my appreciation to some of them. The members of my dissertation committee--Don Van Houten, Ben Johnson, Al Szymanski, and Bill Domhoff--provided me with encouragement, criticism, and invaluable suggestions throughout the process of gathering data and writing the dissertation . Bill Domhoff 1 s empirical work on th e upper class and hi s methodological expertise were especially helpful . Collecting material for a bfographical and historical study such as this can be like working a pu zzle of disconnected or missing pieces, sorting and choosing, and always searching for that extra bit of informa- tion. Although the results of my research are not the final word on the Weyerhaeuser family, I am part icularly indebted to two groups. for helping me to find some missing pieces: The first is the Weyerhaeusers -- the Weyerhaeuser Company public relations depa_rtment for infomation on the Weyerhaeuser Company, and the Weyerhaeuser family members who graciously consented to interviews and who opened their family papers to me, provid i ng unique ins i ghts into the upper class family being studi ed. The second group includes the research librarians and staff at the University of Oregon, Whi tman College , lewis and Clark College, the Minnesota Hist or ical Society , the Washington Historical Society and the Forest History Society. Others have helped me develop some of the ideas presented here. Conversations with colleagues at Whitman and lewis and Clark Col'leges- - especially Ely Chertok, Dick Adams, and Ray Horsworthy--helped me to vii clarify my points and provided new leads and new arguments. I wish to thank the members of the Pacific Northwest Research Center for their help in collecting information, assisting · with charts in the final draft, and for access to their files and friendship. A grant from the Student (Advisory) Committee on International Affairs provided support in the early stages of the research, and is acknowledged. I also wish to thank Connie Ingram and Kathy Maple, graduate student secretaries in the sociology department at the University of Oregon, for the help they gave me throughout my graduate studies. My own kin are particularly important. Ardy not only read the first drafts and made numerous suggestions for improving them, but typed the final copy. Malcolm and Philip have contributed to the dissertation in many ways appropriate for five year olds. The presence and advice of all three have helped me over some slow and hard periods . viii TABLE OF CONTENTS INTRODUCTION .................................... . Chapter I. THEORETICAL ISSUES AND LITERATURE SURVEY ... . ..... 6 II. METHODS: CASE STUDY AND NE1WORK ANALYSIS . . . . . . . . 46 I I I. KINSHIP AND UPPER CLASS ROOTS . . . . . . . . . . . . . . . . . . . . 70 IV. WEALTH AND ITS MAINTENANCE: FAMILY OFFICE AND MEETING, INHERITANCE, TRUSTS, AND HOLDING COMPANIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 V. CORPORATIONS I: HISTORICAL ROLE AND FAMILY INVOLVEMENT (THREE GENERATIONS OF CONTINUITY) .... 170 VI. CORPORATIONS II: CONTINUED INVOLVEMENT AND SHIFTING FORMS . .. . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . 199 VII. PHILANTI-IROPHY, RELIGION & IDEOLOGY ............... 229 VIII. POLITICAL INVOLVEMENT AND INFLUENCE: CAMPAIGN CONTRIBUTIONS, TRADE ASSOCIATIONS, AND OTI-IER INVOLVEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 CONCLUSIONS AND THEORETICAL IMPLICATIONS ......... 330 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346 ix LIST OF CHARTS Chart Page 1 . Weyerhaeuser Family Tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 2. Heyerhaeuser Family Interests . . . . . . . . . . . . . . . . . . . . . . . . 165 3. Weyerhaeuser Corporate Interests in 1934 ............. 194 4. Members of the Second and Third Generations Involved in the Weyerhaeuser Family Business ...... 196 5. Weyerhaeuser Corporate Interlocks, 1950 204 6. Weyerhaeuser Corporate Interlocks, 1960 2li 7. Weyerhaeuser Corporate Interlocks, 1975 224 8. Members of the Third and Fourth Generations Invtilved in Weyerhaeuser Family Business --1975 . . . . . . . . . . . . . 225 9. Flows of Money To and Fran Foundations ............... 248 10. Educational links to Weyerhaeuser Foundations, 1970-1975 .............•....................•. , . . • . 254 11. Religious Links to Weyerhaeuser Foundations, 1970-1975 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 12. Flows of Weyerhaeuser Political Contributions ........ 275 13. Weyerhaeuser Family and Corporate 1972 Presidential Campaign Contributions . . . . . . . . . . . • . . . . . . . . . . . . . . . . 288 X LIST OF TABLES Tables Page 1. Weyerhaeuser Probate Records, Minnesota and Washington . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . . 133 2. Weyerhaeuser Trusts Reported in TNEC Study .......... 144 3. Statement of Dividends Received by the St. Paul Office of Weyerhaeuser & Denkmann For the Year 1902 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 4. Firms In Which the Weyerhaeuser Family CMned an Interest Between 1900 and 1914 . . . . . . .. . . . . . . . . . . . 188 5. Assets of Weyerhaeuser Family and Corporate Foundations, 1970-1973 ........................... 234 6. Board of Trustees of the Weyerhaeuser Foundation, By Nuclear Units in the Third Generation, 1972- 1975 . . . . . . . . . . . . • . • . . . . . . . . . . . . . . • . . . . . . . . . . . . . . . 237 7. Percentage of Weyerhaeuser Family Foundations' Assets in Weyerhaeuser Companies, 1972-1974 243 8. Weyerhaeuser Foundations' CMnership in Five Weyerhaeuser Companies, 1972 . . . . . . . • . . . . . . . . • . . . . 245 9. Jewett Endowment Funds at Harvard College . .......... 259 10. Weyerhaeuser Family 1972 Presidential Campaign Contributions . . • . . . . . . . . . . . . . . . • . . . . . . . . . . . . . . . . . 280 11. Weyerhaeuser Company Di rectors' 1972 Presidential Campaign Contributions ........................... 289 12. Contributions of 12 Prominent Fami 1i es, 1972 295 13. Contributions of 12 Prominent Families, 1956 297 14 . Contributions of 12 Prominent Fami 1 ies, 1960 299 15. Contributions of 12 Prominent Families, 1964 300 xi 16. Weyerhaeuser Family Campaign Contributions, 1968 .... 301 17. Contributions of 12 Prominent Families, 1968 ...•..•. 302 I NTRODlJCTI ON This study is an attempt to increase our understanding of t he class structure and political economy of the United States, through a detailed examination of one extended family of great wealth. Focusing en a single case, the study analyzes the rnulti9enerational and inter-institutional linkages of this kinship group. Through use of a genealogy, kinship ties are traced throuqh five aenerations. The genealogy also provides a medium for identifying the famil y 's links to cor;:iorations, foundations, political processes, and institutions of the upper class. By demonstra - ting how one wealthy family coordinates its activities for t he purpose of maintaining its social and economic position in society, this study suggests that other upper class families may operate in a similar manner. The study v1il 1 show how the family, through a variety of i nsti tut ions, coordinates its activities. The Family Office, the ~nnual Family Meeting, various foundations, and several holding companies are examined as mecha- nisms of internal cohesiveness and of external control over other insti- tutions. The potential for external control and influence aiso extends itself to several large corporations, to trade associations, the candidate selection process, churches, and schools. By focusing on a single family in great depth, new light is thrown upon several important questions in the power structure debates, includ- ing claims about the degree of diffusion of power, about the importance or unimportance of kinship ties, and about the separation of ownership 2 and control. The family (or kinship group) under study traces its origins back to Frederick Weyerhaeuser (1834-1914). A Gennan emigrant, Weyerhaeuser settled in Rock Ishnd, Illinois, where he got his start in the lumber business shortly after the Ci vi 1 War. Fonni ng a partnership with his brother-in-law, he eventually organized all the major lumber families on the upper Mississippi River. Timber operations moved up the Mississippi and its tributaries into Wisconsin and Minnesota, and Weyerhaeuser moved his family to St. Paul. By the time Weyerhaeuser interests bought timber on the West coast, his four sons had joined him in the ownership and management of numerous finns in the timber industry. Frederick died in 1914, but he left a legacy of sons, grandsons, and great-grandsons to carry on. the traditional sociological literature contains a number of quest- ionable propositions about the nature of wealth and power in the United States. Some theorists assume that power in advanced industrial societies is diffuse {Rose, 1967; Dahl, 1961; Polsby, 1963). Others claim that kinship ties and family structure have lost their influence on the larger social structure (Goode, 1963; Parsons, 1964b, 1964c; Parsons and Smelser. 1956). For some, the nuclear family, which is assumed to be the dominant form of kinship unit, is viewed as "isolated" from economic and political affairs (Parsons, 1964a, b, c; Harvey, 1975; Adams, 1970. Adams claims these are the two dominant themes in recent kinship studies.). Some of these theorists a 1s o cl aim that the notion of an hereditary cl ass with any power is no longer valid (Parsons and Smelser, 1956; Parsons, 1960). The separation of management from ownership in the modern corporation is 3 said to be virtually complete (Dahrendorf, 1959; Gordon, 1966; Bell, 1962; Parsons and Smelser, 1956; Larner, 1970). The first chapter of this study discusses these perspectives in re- lation to several continuing debates over the nature of wealth, power, and kinship in the United States. The second chapter discusses the use of the case study method, gene- alogies, and network analysis in exploring sociological propositions. A case study obviously cannot give definit·i ve answers to the questions mentioned above. What I hope to show, however, is how one kinship group operates and how it exerts control and influence over other institutions thought to be differentiated from the kinship group. An in-depth histori- ca 1 case study of one extended family wi11 hopefully shed new 1i ght on claims about the diffusion of power, the importance of kinship ties, and the patterns of family control in the United States. Chapter III examines the internal relations within the kin group itself. It seeks to show the kinds of internal connections and the links that the family has to upper class institutions. The fourth chapter explores the most significant findings of this whole undertaking. It demonstrates how one large kin group has used a Family Office and Annual Family Meeting to maintain cohesiveness and continuity. It suggests that the Office and Meeting may be the new institutional fonn and mechanism for controlling corporations and for family involvement in philanthropic activities and politics. Working outward from these internal connections (i.e., kin linkages and the Office and Meeting), Chapters V and VI trace the family's in- vo 1v ement in numerous timber companies. Stock ownership, di rector inter- 4 locks, and involvement in management are analyzed over time to determine how extensive family control is today, and whether the degree of control has changed. Chapter VII discusses the philanthropic and religious activities of the family. The family's links to politicai processes are examined in Chapter VIII. Both chapters focus on the role of the Family Office and ~eeting in coordinating these activities. It is my contention that the structure of family (more precisely kinship) wealth and power is not only an historical mechanism of class maintenance, but a contemporary form of family capitalism. Rather tha n being broken up, the ownership of corporate stock and, more importantly, the mechanisms of control have taken on new organizational forms. Control, instead of shifting to a new managerial class, remains within the kinship units of the upper class. The kinship group provides the structure and secrecy to perpetuate power relations originating in upper class families and extending to other institutional structures. The fusion of property and kinship, thought to be a vestige of earlier stages of capitalism, may actually be a major contributor to "intergenerational immobility in capi- tal ownership" (Weisskopf, 1972:129). This fusion is represented by shared economic interests and kinship bonds. It is best described as a "kinecon group," a phrase recently coined to replace the notion of family capitalism. In their analysis of kinship groups and the control of large corporations in Chile, Zeitlin, Ewen, and Ratcliff define a "kinecon group" as"~ complex kinship unit J.n. 1-1hich economic interests and kinship bonds are inextricab·1v intertwined" (Zeitlin et 21_., 1974:109; emphasis theirs). Continuing, they spell it 5 out in more detail: The concept of the kinecon group is meant to be class specific: where shares of large corporations have become the typical and decisive form of capital ownership, and the relationship between specific ownership interests and corporate control becomes historically problematic, the concept of the kinecon group applies. The corporation is the legal unit of ownership of large-scale productive property. The set of interrelated kin who control the corporation through their combined ownership interests and strategic representation in management constitute the kinecon group. (Zeitlin et tl-, 1974:110) 6 CHAPTER I: THEORETICAL ISSUES AND LITERATURE SURVEY Several areas in the literature focus on issues having to do with wealth, power, and kinship. The most obvious are the areas of (1) gov- erning class theory, (2) studies of kinship and family structure, and (3) the debate over corporate ownership and control. 1. Pluralism constitutes the dominant view within academic circles as to wlio governs the country and how they do it. This theoretical per- spective contends that a number of co~peting groups--tusiness, labor, fanners, churches, and civic soci eti es--share power in American society. An alternative explanation of the structure of power in U.S. society is provided by governing c1ass theory. A theory indebted to Marxism, it has developed a model which roots classes in an historically specific economic order, characterized by monopoly capitalism. Governing class theory has sought to <:xplain the structure and distribution of power in society by analyzing the continuing conflicts between capital ists and workers. One aspect of these analyses focuses on the social characteristics and proper- t.Y interests of the upper (or capitalist) class. Using empfrical studies the proponents of governing class theory have sought to demonstrate that the major eapitalists are a ruling class which dominates all aspects of American society. 2. Sociologists, for the most part, hav~ net engaged in kinship studies, and when they have, have focused on the lov,er and middle classes. The emphasis in their studies has been on the declining importance of the kinship unit in modern industrial society, and on its lack of interaction with other institutional structures, most notably economic o~es. A survey of the kinship literature in the 1960's, however, suggests that kin groups, especially upper class kin groups, may be more important than has been assumed (Adams, 1970). 3. The decades-old debate over corporate ownership and control has produced a large body of literature in economics, political science, law, and sociology. The major division within the debate appears to be be- . tween those who claim that corporate control has become a function sepa- rate from ownership, and those who argue that control and ownership con- tinue to rest in the same hands. The question is more complex than it first appears. This is especially t,ue when participants in the debate begin relating the issue of control to class, power, and kinship; such is, however, necessary if we are to understand the nature and scope of power in American society. Governi~g Class Theory Governing class theory provides a theoretical model of the .,.;orkings of the UnHed States political economy which is at variance with the traditional sociological literature (summarized briefly in the introduc- tion). The theoretical implications drawn from this model will provide the background for our study of the relationship between kinship and property, of the structure of family wea 1th and power, and of the nature of corporate ownership and control in the case of the Weyerhaeuser kin. The theoretical implications drawn from this model have also con- tributed to a more general theory atout the nature of advanced industrial 8 society (see Miliband, 1969). In the final chapter we will return to examine some of these broader theoretical questions in light of the findings of the present study. For now, it is enough to indicate that the issues under discussion cannot be viewed in total isolation from other theoretical questions about the nature and workings of capitalist society. Although discussion of the broader theoretical issues will come later, a detailed discussion of the roots, and the internal development, of the theory of governing class is in order at this time. C. Wright r-~ills, E. Digby Baltzell, Paul Sweezy, - and G. \..Jilliam Domhoff have been chosen as representative; each has made major contributions to governing class theory. By lumping them together I do not mean to imply that they are in complete agreement on all issues, for tAere are differences in their respective approaches and emphases. The major differences will be illuminated as the discussion proceeds. Governing class theory recognizes that classes are an integral part of an analysis of any society, and it seeks to understand the social, political, and economic structures of any society through an analysis of its class structures. The theory focuses on the dominant classes in society in its search for explanations for the relationships of power and wealth in advanced industrial society. Its central hypothesis is that a social upper class of big business- men and their descendants controls the major institutions in the United States. Through direct occupation of the top positions in these institu- tions, and through careful selection of high-1eve1 employe~s. members of this class are able to exert their influence over major banks, corpora- 9 tions, foundations, universities, trade associations, policy planning groups, and political parties. In other words, this upper social cl~ss functions as a ruling class. The contention is that, in addition to owning a disproportionate amount of society's resources, this upper class also coordinates its activities to maintain control as well as ownership. It does so not in the interests of society in general, but in its own interests . One of the early pioneers in this area was C. Wright Mills. Seeking explanations for the organization of power in society, t,1i1ls drew upon the theoretical works of Marx, Weber, ~osca, and Pareto (Mills, 1959). From the classical elite theorists Pareto and Mosca, Mills adopted the notion that an elite, or "organized minority," consisted of t hose "whe occupy the posts of political command" (Bottomore, 1964:12). The classical elite theorists felt that an "elite," "governing elite," or "ruling class" was inevitable, i.e., that it was universally valid for all societies past, present, and future. Mills however rejected this no- tion. "It is not my thesis," he said, "that for all epochs of human his- tory and in all nat"ions, a creative minority, a ruling class, an omnipo- tent elite, shapes all historical events" (Mills, 1959:20). Statements that attributed "a class that rules and a class that is ruled" to a11 societies, as ~·'osca (1939:50) had done, Mills thought were "mere tautolo- gies" (~ills, 1959:20). He also rejected Pareto's notion of "circulating e 1 i tes. 11 I don't think history is merely a succession of elites which, one after the other, conquer the institutional means of power. That is an omnipotent theory of the elite and an elite theory of history from which I hc.ve been very careful to dissociate my view. The structural mechanisms of institutions must indeed be ~iven due weight. (Mills, 1968:248) 10 In emphasizing "the structural mechanisms of institutions" Mills differs from the nee-Machiavellians by providing a sociological rather than a psychological explanation for elites. The classical elite theor- ists attributed inevitability to elites because they rooted them in the psychological conditions of people. The existence of elites in all so- cieties according to Pareto was the result of sentiments. Sentiments or residues were for him "the fundamental and predominant force in society, the determinino factor of human conduct" (I. Zeitlin, 1968:162; emphasis his). Mosca also reduced his explanation for the dichotomy of rulers and ruled to ''the unchanging psychological nature of man" (I. Zeitlin, 1968: 200). Both Bottomore and Irving Zeitlin point out, however, that Mosca recognized the importance of social forces in accounting for the rise of new elites (Bottomore, 1964:13; I. Zeitlin, 1968:200). But for both Mosca and Pareto, psychological tendencies were more important than in- stitutional, sociological arrangements. The latter factors are th~ ones emphasized by Mills. Bottomore claims that, like other elite theorists who view democratic societies as ruled by an elite, Mills presents us with a pessimistic view. This is not accurate. Because Mills 1 emphasis is on social forces and institutional arrangements rather than on human nature, in determining the make-up of the elite, he does allow for the possibility of change. Mills' analysis of wealth and power in the United States, The Powe.!:_ Elite, stands as one of the classics in ftmerican sociology. It begins with a discussion of three institutional spheres of power, economic, political, and military. The top leadership in each of these "domains cf power--the warlords, the corporate chieftains, and the political direc- 11 torate-:--tend to come together, to form the power e 1i te in America" ( Mi 11 s, 1959:9) , The power that these men have is a structural component, not of their individual personalities, but of the major institutional hierarchies in society. The real basis of power in American society, in other worrls, is institutional. These tightly interlocked institutional structures are the dominant force in the United States. To understand the real basis of wealth, Mills recognized, requires a thorough study of these institutional structures, particularly the large corporation. The linkage between wealth and power has always manifested itself in some form of property. In modern society, the institutional form of property par excellence has become the large corporation. As Mi 11 s says, The pyramid of wealth cannot be understood merely in terms of the very rich; for the great inheriting families ... are now supplemented by the corporate institutions of modern society; every one of the very rich families has been and is closely connected--always legally and fre- quently managerially as well--with one of the multi- million dollar corporations. (1959:9-10) Although he is not explicit about it, Mills is analyzing the economic bases of class. Family wealth is rooted in a specific system of property relations, today the modern corporation. Those families which are shown to be connected to corporate property share a "common position" in the economic order. It is because of their institutional relation to property that these families have power. If Mills is not very explicit about the economic nature of class, he is explicit about its social characteristics. In commenting on the inner circles of "the upper social classes," he says that 12 They form a more or less compact social and psychologica1 entity; they have become self-conscious members of a social class. People are either accepted into this class or they are not, and there is a qualitative split, rather than merely a numerical scale, separating them from those who are not elite. They are more or less aware of themselves as a social class and they behave toward one another differ- ently from the way they do toward members of other classes. (Millsi 1959:11) Mills, however, does not define this 0 inner circle" as the upper class. His focus is rather on the people who fill the top positions in these institutions, not as a class, but as an elite. In defining this group as the power elite, he refers "to those political, economic, and military circles which as an intricate set of overlapping cliques share decisions having at least national consequences. In so far as national events are decided, the power elite are those who decide them" (Mills, 1959:18). There are three major aspects or characteristics of the power elite. (1) Psychological similarities. They share certain common experi- ences, which provide the framework for the "psychological and social basis of their unity." These shared experiences result in a unity and cohesiveness not found in other segments of society. (2) Structural hierarchies. The unity of the elite is rooted in institutional structures. There is an interaction process between the institutions and the people in them, particularly involving the people who occupy t he top positions. (3) Coordination. The cohesiveness of the elite is sometimes the result of 0 a more explicit co-ordination. 0 Working together in both for- mal and i nformal ways the e1He coordinates its activities. Any one of these characteristics provides by itse1f only a partial 13 explanation of the power elite's unity, but together they help us under- stand the cohesive nature of the elite. One of the bases of this unity is found in the life style and train- ing of the upper class. This is especially true of the older upper class families. The core membership of a variety of upper class institutions-- metropolitan men's clubs, country clubs, private schools--consists of these wealthy families. As Mills says, "in these private schools for adolescents, the religious and family and educational tasks of the upper social classes are fused, and in them the major tasks of upholding such standards as prevail in these classes are centered" (Mills, 1959:65). As an "organized extent ion of the family" the private prep school, according to Mills, has replaced the family as the primary socializing agent of the upper class. The upper class family, however, remains a vital element in provid- ing a sense of unity and, more importantly, in solving problems. Accordingly, in the inner circles of the upper classes, the most impersonal problems of the largest and most important institutions are fused with the sentiments and worries of small, closed, intimate groups. This is one very important meaning of the upper-class family and of the upper-class school; "background" is one way in which, on the basis of intimate association, the activities of an upper class may be tacitly co-ordinated. It is also important because in such circles, adolescent boys and girls ·are exposed to the table conversations of decision-makers, and thus have bred into them the infonnal skills and pretensions of decision- makers; in short, they imbibe what is called "judgement." Without conscious effort, they absorb the aspiration to be-- if not the conviction that they are--The Ones Who Decide. (Mills, 1959:69) This points out the equally important role of the family and the power vested in it. But .i..?_ power vested in the family? Or does the 14 family act as a coordinating agent, coordinating other institutional spheres of power? Mills does not always make this explicit, preferring to root power in institutional structures like the corporation and mili- tary. The family, however, clearly represents an important institutional structure, which alone may not have much power but which in relationship with other institutional structures can coordinate its activities in the interests of its class. Families or individuals do not achieve wealth or power in isolation from some form of institutional base. Mi11s 1 emphasis is clearly on the corporate basis of wealth and power. Wealthy families, "in realizing the power of property and in acquiring instruments for its protection," ha ve become linked to 11 the higher corporate world of the twentieth-century American economy." Continuing, Mills says: Not great fortunes, but great corporations are the important units of wealth, to which individuals of property are variously attached. The corporation is the source of wealth, and the basis of the continued power and privilege of wealth. All the men and the families of great weaith are now identified with large corporations in which their property is seated. (Mills, 1959:116) In linking 1t1ea1thy families to corporations, Mills claims to reject both the notion that a "clannish" group of families rules the American economy, and the idea that a group of managers have "expropriated the powers and privileges of such families" {Mills, 1959:147). He argues instead that the propertied class has been reorganized, "along with those of higher salary, into a new corporate world of privilege and prerogative" (Mills , 1959:147). The history of ./imerican capitalism, it is argued, rather than exhibiting the breakup of family wealth or the management of 15 that wealth by a new elite, has exhibited a great deal of continuity on the part of the capitalist class. What is significant about this managerial reorganization of the propertied class is that by means of it the narrow industrial and profit interests of specific finns and industries and families have been translated into the broader economic and political interests of a more oenuinely class type. Now the corporate seatscrf°the rich conta-=rna:Tl the powers and privileges inherent in the institution of private property. (Mills, 1959:147; emphasis his) In a critique of Mills' Power Elite, Paul Sweezy discusses some of the conceptual problems Mills was unable to resolve. Mills, he said, accepted "the substance as well as the terminology of a kind of 'elitist' doctrine which is basically antithetical to the general trend of his thought" (Sweezy, 1968:122). Mills' definition of the power elite is couched in the language of elitist theory, but the theoretica·1 scheme is somewhat ambfvalent, sometimes elitist, other times relying on a class analysis. According to Sweezy, Mills follows two often conflicting approaches as he develops a theoretical scheme in which to locate the elite. On the one hand, he locates the power elite in a national upper class. The hypothesis in this case is that these in the command posts are representatives of a national ruling class. At other times, he lo- cates the elite in the "major institutional orders" of society. Here the hypothesis would test whether the elites are the products of distinct in- stitutional structures. A comprehensive analysis of ft.merican society, Sweezy believes, would have to root the elite in a theoretical framework of social class. Mills 1 cumulative empirical evidence, he argues, supports the interpretation of a class system; enough documentation is presented in The Power Elite to 16 show that the elite is rooted in an upper social class. As Sweezy says. Mills adduces a wealth of material on our class system, showing how the local units of the upper class are made up of propertied families and how these local units are welded together into a wholly self-conscious national class. He shows how the "power elite" is overwhelmingly (and increasingly) recruited from the upper levels of the class system, how the same families contribute in- differently to the economic, military, and political "elites," and how the same individuals move easily and almost imperceptibly back and forth from one to another of these "elites." (Sweezy, 1968:124) In an earlier article, "The American Ruling Class,1' written in 1951, Sweezy sought to develop the concept of the ruling class. His purpose was to clarify and elaborate on the use of this concept in an analysis of the American social structure. There had been many factual studies-- Lundberg's, O'Connor's, and Rochester's are the examples Sweezy gives--of particular aspects of the ruling class in the United States. But the lack of a comprehensive study of the structure of this class remained a major deficiency in the social sciences. Any adequate theory of social class must recognize, Sweezy began, that classes are objective entities, "not the artificial creations of the social scientist" (Sweezy, 1953b:122). After emphasizing that classes cannot be construed as subjective figments of the viewers' imagination, Sweezy defined the key element of class membership. "The fundamental unit of class membership," he said, "is the family and not the individual" (Sweezy, 1953b:123). This is not to say that individuals do not belong to a particular class, but only that the basic unit for assigning member- ship in a particular class is t~e family. "A social class," according to Sweezy, "is made up of free ·ly intennarrying families" (Sweezy, 1953b:124). 17 Family units may be the "fundamental unit of class membership," but the property system is the distinguishing feature of the relationsltip between classes. Property determines the number of classes and the dividing line between them, in any analysis of class. "The upper classes are the property-owning classes; the lower classes are the propertyless classes" (Sweezy, 1953b: 124). After these preliminary remarks on the nature of class systems ir. general, Sweezy warned against a simplified view of class which ascribes a tight homogeneity to classes. Not all members of a particular class play the same role, hold the same values,· or necessarily behave in the same manner. Although he did not use the term, he appeared to be speaking of a differentiation of roles or functions within a class. Using the American class system as his particular example, Sweezy proceeded to analyze the class structure of capitalist society. Classes in a capitalist society are defined by "the very nature of capitalism: the owners of the means of production (the capitalist class), and the wage laborers who set the means of production in motion (the working class)" (Sweezy, 1953b:126). This is the economic criterion of class membership. When coupled with "the criterion on intennarriageability as a test of social class membership," the analysis of classes becomes more complex but more comprehensive. Sweezy concluded: ... the social classes which we observe about us are not identical with the economic classes of capitalist society. They are rather modifications of the latter .... If we keep [this point] firmly ,n mind .,.,e shall be able to appreciate the decisive role of the economic factor in the structure and behavior of social classes while at the same time avoiding an overmechanical (and hence false) econolilic determinism. (Sweezy, 1953b:127; emphasis his) 18 Any analysis of the class structure of the United States, Sweezy said, must recognize "that two social classes, at bottom shaped by the very nature of capitalism, detennine the fonn and content of the system as a whole" (Sweezy, 1953b:128). These two classes are identified as the ruling class and the working class. The national upper class is not merely a group of interrelated families. It is rooted in institutions, particularly economic institu- tions--the dominant fonn being the corporation. This "massive network of institutional relations" (Sweezy, 1953b:132) serves an integrating role in the ruling class. In addition to economic institutions there are a variety of clubs, alumni associations, churches, and fashionable resorts which also play an important part in upper class solidarity. These institutional linkages serve not only as an internal cohesive force bind- ing the ruling class together, but also as a mechanism which perpetuates the ideas and values of this group. The family and educational system are identified as two of the key parts of this process. In transmitting ideas to the rest of society, and in providing an internal reference group, the institutional network provides a continuity of ruling class ideology. Sweezy's work is helpful in clarifying some of the theoretical developments of the notion of ruling class. But it wasn't until publica- tion of E. Digby Baltzell 's study of the upper class in Philadelphia that empirical eyidence was available to support them. Philadelphia Gentlemen represents an explicit attempt to expand Mills' institutional and class analysis of the higher circles. In a manner reminiscent of ~ills, Baltzell traces "the relationship between the upper class and the elite" 19 (Baltzell, 1958:7). By defining class primarily in social tenns Baltzell sought to demonstrate that the upper class is a social class. The ,~per class concept, then, refers to a group of fami 1es, whose members are descendants of successful 1ndiv1duals (elite members) of one, two, three or more generations ago. These families are at the top of the social class hierarchy; they are brought up together, are friends, and are intermarried one with another; and, finally, they maintain a distinctive style of life and a kind of primary group solidarity which sets them apart from the rest of the population. (Baltzell, 1958:7; emphasis his) Mi 11 s had pl aced his emphasis on economic, po 1i ti ca 1, and military institutions. Baltzell's analysis is focused on the family as an institu- tion. The tap root of any upper class, that which nourishes each contemporary generation with a sense of tradition and historical continuity, is a small group of f~milies whose members were born to that class, and whose ancestors have been "to the manor born" for several generations. (Baltzell, 1958:9) The relationship between upper class and ruling class is specified in remarks criticizing Burnham's Managerial Revolution (1941). Baltzell claims Burnham's analysis is confusing because it fails to "determine whether or not these new managers were recruited from the same upper class as the financiers and owners" {Ba1tze11, 1958:34). Baltzell argues that if the "new managers and the 2..!i financiers are recruited from the same upper class," there has been no change in the composition of the ruling class . Defined in this way, the ruling class is that group which "con t ri butes upper class members to the most important, goal-integrating elite positions" (Baltzell, 1958:34). Burnham's problem, accord i ng to 20 Baltzell, was that he confused the ruling class with the ruling elite, which led him to conclude that a new ruling elite of managers had re- placed the old ruling class of financiers. Hypothetically, an upper class is not the same as, or the equivalent of, a ruling class. If it is not, however, it quick1y loses its position. As Baltzell says, If [the upper class] is not a ruling class, it will soon be replaced by a new upper class .... But we emphasize here that an upper class remains the ruling class only as long as its members are in the key executive positions, even though the technical and intellectual positions may be open to achievement. (Baltzell, 1958:34) After defining his concepts, Batlzell specifies the functions the upper class performs in society. These are: (1) to maintain a continuity of control over important positions in the world of affairs; (2) to provide a primary group social organization within which the infonnal aspects of the nonnative order--the folkways and mores--may operate as effective agents of social control; (3) to provide an autonomous power in the community as a protection against totalitarian power; and, finally, (4) to provide a more or less primary group social world within which the younger generation is socialized. (Baltzell, 1958:60) Although his emphasis is on the social aspects of class, there can be no denying that Baltzell also defined the upper class economically. Domhoff has captured this, in summarizing Baltzell 's major thesis that a national upper class of rich businessmen and their descendants came into existence in the last part of the nine- teenth century as a result of the national corporate economy and the national transportation-communication network. (Domhoff, 1968:268) Several empirical studies, since Baltzell and ~ills, have explored 21 the inner workings of the upper class and further refined and developed the theoretical framework of governing class theory. G. William Domhoff, in a manner reminiscent of Baltzell 's, has grounded the social upper class in the dominant economic and political institutions. In this manner he has demonstrated that the upper class continues to act as a ruling class. In Who Rules /l,merica? Domhoff provides a framework for determining upper class membership. Developing indicators and identifying a number of institutional organizations of the upper class, he documents the existence of a national upper class. The indicators show that the upper class is a social class and that individuals in various economic and political organizations are members of this group. For an individual to be considered a member of the upper class, he or she must meet one of the following criteria: (1) be listed in the Social Reoister; (2) have attended an exclusive private preparatory school; (3) be a member of an exclusive gentlemen's club; (4) be the son or daughter of a millionaire entrepreneur or of a $100,000-a-year corporate executive or lawyer, and have attended a private school or belong to an exclusive club; (5) marry a person defined as a member of the upper class by one of the previous criteria. The upper class is comprised of "intermarrying and interacting fami- lies of high social standing" which are distinguished by their "great wealth and unique life style." A.s a social upper class this group can be defined "as people who are listed in certain social registers and blue books, people who attend certain private schools, and people who belong 21 the inner workings of the upper class and further refined and developed the theoretical framework of governing class theory. G. William Domhoff, in a manner reminiscent of Baltzell 1 s, has grounded the social upper class in the dominant economic and political institutions. In this manner he has demonstrated that the upper class continues to act as a ruling class. In Who Rules ft,merica? Domhoff provides a framework for determining upper class membership. Developing indicators and identifying a number of institutional organizations of the upper class, he documents the existence of a national upper class. The indicators show that the upper class is a social class and that individuals in various economic and political organizations are members of this group. For an individual to be considered a member of the upper class, he or she must meet one of the following criteria: (1) be listed in the Social Reaister; (2) have attended an exclusive private preparatory school; (3) be a member of an exclusive gentlemen's club; (4) be the son or daughter of a millionaire entrepreneur or of a $100,000-a-year corporate executive or lawyer, and have attended a private school or belong to an exclusive club; (5) marry a person defined as a member of the upper class by one of the previous criteria. The upper class is comprised of "intermarrying and interacting fami- lies of high social standing" which are distinguished by their "great wealth and unique life style." As a social upper class this group can be defined "as people who are listed in certain social registers and blue books, people who attend certain private schools, and people who belong 22 to certain exclusive social clubs 11 (Domhoff, 1970:32). Pres en ting evidence of "in-group interaction and differenti a 1 1i fe styles," Domhoff argues that the upper class is a cohesive unit and is conscious of itself as something special. "Overlapping social cliques," produced by intennarriage, exclusive schools, clubs and resorts, and shared acquaintances, provide evidence of in-group cohesiveness. Sociali- zation patterns--from an early childhood of governesses to special treat- ment after death--confer special privileges and status on members of the upper class which set them apart from the rest of society. This evidence along with the testimony of "informants" confirms the existence of upper class consciousness (see Domhoff, 1970). Having demonstrated the existence of an upper social class, Domhoff documents its relationship with the national corporate economy. In both Who Rules America? and The Higher Circles, he demonstrates that the upper class overlaps with an economic class. He shows the continued domination of the corporate economy by the upper class through its members' role as directors of the nation's largest banks and corporations. Evidence of a tightly-knit corporate economy comes from studies on the concentration of stock ownership within the upper class and on the patterns of interlock- ing directorates among the larger corporations. As a group characterized by common family position and wealth, a unique life style, and group solidarity, the upper class fulfills the social aspects of class. As a group rooted in the wealth of large cor- porations, this class' economic position becomes clearer. In other words, the American upper class is a socioeconomic class. Having established the existence of a social upper class which over- 23 laps with an economic class, Domhoff 1 s next step is to demonstrate that this class also governs. The upper class does this either directly or through hired employees. These are corporate lawyers and managers who are not members of the upper class but who serve the interests of the upper class. This group of hired employees in 11 command positions in institutions controlled by members of the upper {governing) class, 1' is part of what Domhoff calls the 11 power elite 11 (Domhoff, 1967:10). The power elite also encompasses the active members of the social upper class who operate in these command posts. By grounding the 11 power elite 11 in the upper class, Domhoff has established the existence of a ruling class. The ruling or governing class is a social upper class which owns a disproportionate amount of the country 1 s wealth, receives a dispro- portionate amount of the country 1 s yearly income, and contributes a disproportionate number of its members to positions of leadership. (Domhoff, 1967:9) At the same time, Domhoff has suggested how Mil1s 1 11 power elite" can be seen as the 11 leadership arm 11 or 11 establishment 11 of the ruling class. He concludes that the income, wealth, and institutional leadership of what Baltzell calls the 11 Jl,merican business aristocracy 11 are more than sufficient to earn it the designation 11 ~10verning class. 11 As Sweezy would say, this 11 ruling class 11 is based upon the national corporate economy and the institutions that economy nourishes. It manifests itself through what the late C. Wright Mills called the power e 1i te. (Domhoff, 1967: 156) The empirical evidence accumulated in Who Rules America? demonstrates how this governing class "controls 11 major corporations, foundations, and 24 private universities and 11 dominates 11 governmental processes. In The Hiqher Circles, Domhoff provides more empirical evidence of the domination of the government by members of the upper class and their employees. Here the focus is on the formation of governmenta 1 po 1i cy and on how the govern- ing class transforms its wishes into policy. Through this empirical evidence Domhoff claims to have shown that "a power elite rooted in the social upper class and the major corporations governs the United States" (Domhoff, 1972:29). Literature from two other areas should be integrated into our discussion of governing class theory. Kinship studies and the debate over corporate ownership and control can both contribute theoretical insights and empirical evidence for a further development of governing class theory. In the realm of theory, kinship studies help clarify the conceptualization of class. Their empirical contributions to discussion of the relations between kinship and other concepts, notably class and property, are less useful. Studies on corporate ownership and control, on the other hand, can contribute a wealth of evidence to the governing class thesis. But studies in this area have not reached any kind of consensus on either the theoretical or empirical questions raised. Conflicting empirical evidence has resulted in a great deal of theoretical confusion and often in misleading notions of power in the United States. Kinship analysis in anthropology has been used to develop a variety of theories about human society. Sociologists have tended to place less 25 emphasis on kinship systems, particularly in ' modern society. Some studies, however, have analyzed kinship in relationship to social class, mobility, and other specifying variables. Recently an effort has been made to relate kinship to other institutions in the United States, but, as Eert Adams points out, this process is "barely underway" (A.dams. 1970:589). The majority of kinship studies, descriptive in nature, have focused on the internal relati0nships within the kinship system itself. Adams does provide a good survey of the literature on the role of kinship in the United States. Reviewing the literature written in the 1960 1 s, he discovers that the dominant theme then was actually a carry- over from the work done in the 1930 1 s and 1940 1s. This theme viewed kinship and family structure as relatively unimportant to the functioning of modern society and considered the nuclear family to be isolated from other kin and differentiated from other social systems. It was generally assumed "that industrialization gave rise to the isolated nuclear family" (Adams, 1970:576). Such assertions about the declining importance of kin units received support from many prominent theorists. Talcott Parsons, for one, claimed that the nuclear family had become increasing isolated in the United States. In an article in the early 1940 1 s he called this isolation of "the conjugal family" the "most distinctive feature of the American kinship system" (Parsons, 1964b:l85). The thrust of his argu- ment centered on the family's separation from other kin. "Isolation" also meant, however, that a variety of functions originally fulfilled by the family were now perfonned by other institutions. This differenti- ation of functions further "isolated" the family and reduced the import- ance of ki n groups. 26 On closer examination of Parsons' work on kinship patterns in the United States, we find several "deviations" from this typical pattern. Rural families, upper class "elements," lower class "instability of marriage," and "suburban matriarchy" were listed as four areas where the kinship system did not follow the typical pattern. This left Parsons with the urban middle class family as the "representative" type of American kinship system (Parsons, 1964b:185). The representativeness of thi's group has to be questioned at the very least. It is interesting to note, however, that only in his work on the kinship system (1964b) does Parsons mention these deviations or differ- ences between upper class kin and other kinship units. For example, his discussion of the relationship between kinship and class in his work on stratification (1964a) never mentions any differences. Kinship and class are analyzed in general terms. The assumption is that the kinship patterns of the middle class are representative of all classes in America. Although they would deny the importance of the economic system as a determining factor of class, both Parsons and Daniel Be11 (1962) reccgnize that there is a relationship between the family and property. Parsons says, for example, that "this structural connection between the family unit and the institution of property as the basis for control of the factors of production is the essence of 'capitalism' ... " (Parsons, 1960: 110). This linkage, however, between property and kinship structures is essentially viewed as an historical one by Parsons. Daniel Bell maintains that there was an historical "fusion" between property and family which maintained "a class system." The link which joined these two institutions was the economic system. Yet the factors 27 which "hindered the development of a full system of•family capitalism in the United States" are related to "historical and socio-psychological events" (Bell, 1962:41). The reasons cited as factors leading to the decline of family capitalism are "the decline of the extended family which narrowed the choice of heirs competent to manage the enterprise," and "the increasing importance of professional techniques which placed a high premium on skill rather than blood relationships" (Bell, 1962:41). Parsons' failure (noted above) to analyze the role of the upper class in his kinship studies does not make him unique. In his discussion of social class as a variable, Adams also makes no mention of upper class kinship. Studies he surveyed focused exclusively on the lower or midd"le classes. In discussing the need to relate kinship to other institutions, particularly economic ones, Adams mentions the possible role of kin as links in the corporate structure. The works of Baltzell and Kolko are cited to show how wealth and social position are inherited in U.S. society. This evidence that kinship plays a much larger role in modern society than is usually assumed tends to come from the literature on power and stratification, however, and not from the kinship literature . Adams' review of the kinship literature finds that the debate over the isolation of the nuclear family still continues. His conclusion, drawn from his survey of the literature, is that the family "is not i solated interactionally, attitudinally, or in terms of knowledge" (Adams, 1970:580) . He cites a large number of descriptive studies showing "that 'close, diffuse ties' with kin are maintained by many in the U.S."; yet, as he points out , the "protestations" of scholars continue to assert, with little supporting data, that kin are not as important in 28 modern U.S. society as they were in pre-industrial spciety (Adams, 1970: 580; emphasis his). While the trend (excepting Adams) has been to follow Parsons and those who say that the importance of kin has declined, not all have join- ed the bandwagon. For example, August Hollingshead, in commenting on the importance of kinship systems in maintaining family stability, presents an alternative view. The "established upper class family" is, he says, "basically an extended kin group." Within the kin group the nuclear family is viewed as part of a broader kin group "that included the con- sanguineal descendants of a known ancestral pair, plus kin that have been brought into the group by marriage" (Hollingshead, 1950:41). He also recognizes that economic stability is an important factor in a kin group 1 s ability to maintain family stability from one generation to the next. Bernard Farber also has studied "the role of kinship and family in sustaining class structure" (Farber, 1971 :4). His emphasis is on the different functions perfonned by different classes' kinship groups. Lower class kin units are seen as emphasizing an integrative function. Upper class kin systems "stress the social differentiation function that develops means for keeping the constituent kinship units separate while pennitting them to accumulate and inherit wealth and power" (Farter, 1971:7). In Farber 1 s analysis the concept of kinship is clearly linked to those of social class and stratification. On one hand, kin units insure the continuation of wealth and privilege; on the other, they pro- vide an ideological framework that results in the integration of people into social classes. Viewed in this way, kinship systems reflect class interests in a manner similar to the relationship between religion and 29 class (see Pope, 1948). Farber provides theoretical suggestions about the role of upper class kinship units, but his research data is obtained from a population in a lower class community, thus making it difficult for him to generalize to other socioeconomic classes. There has been a lack of kinship studies of the upper class, and the area remains unexplored. Sociologists have left the study of kinship systems to anthropologists, ,,.,ho have shown little interest in analyzing the top echelon of society. The study of wealthy families, particularly aspects of their wealth and power, has been left to journalists. When compared to kinship studies of the middle and lower classes, our knowledge remains minute. Journalists like Stephen Binningham and Ferdinand Lundberg have provided accounts of the 1ife styles of the rich; sociologists Mills, Baltzell, and Domhoff have helped increase our understanding of the role of the upper class in contemporary society, but their work has provided almost the entirety of our knowledge of the kinship structures of upper class families. There are, however, two approaches which suggest ways of resolving the debate over the importance of kinship and the alleged isolation of the nuclear family. The first calls attention to the need for research on the relation between "kin ties and economic control." There is a need, as Bert Adams suggests, for "studies of the interrelations between kinship and other institutions, studies of the relative importance of kin and other industrial institutions ... " (,lldams, 1970:580; emphasis his). The second comes from Edward Saveth and Tamara Hareven, who both suggest the need for a historical perspective in studying family structure. Saveth ·n particular mentions the potential usefulness of genealogical studies 30 for research into the activities of the upper class. , Ownership and Control The debate over corporate ownership and control has produced a large body cf literature, dating back to the work of Adolf A. Berle and Gardiner C. ~eans in the early 1930's. In The ~odern Corporation ~nd Private Property, Berle and ~eans presented findings that showed a trend toward the concentration of corporate wealth. "The ultimate control of nearly half of industry," they argued, was "actually in the hands of a fev, hundred men" {Berle and ~eans, 1967:46). It has not been their findings showing a trend toward concentration of economic power, however, that have received all the attention. The major focus of the controversy raised by :their work has centered on the question of stock dispersion and managerial control. They asserted that the ownership of corporate stock had become widely dispersed among a large number of people and that the control of the count~y•s top corpora- tions had shifted from a small group of wealthy families to a group of professionally trained managers. This dispersion of stock was interpreted as meaning that "in many large corporations the largest stockholding represents a small propcrtion of the total ownership while the number of stockholders is legion" {Berle and l"eans, 1967: 48). Th.e example they selected as the prototype corpora- tion of the future was the American Telephone and Telegraph Company, which had then assets of five billion dollars, over 454,000 employees, and over half a million stockholders. Berle and reans predicted that in the future a small number of firms like AT&T would come to dominate the 31 country. "One hundred companies of this size," they ~peculated, 11 would [come to] control the whole of American wealth; would employ all of the gainfully employed; and j_f there were~ duplication of stockholders, would be owned by practically every family in the country" (Berle and Means, 1967:5; emphasis added). Even though stock dfspersion had not proceeded that far, there had already taken place, according to Berle and Means, a drastic change in the position of ownership, as well as changes in the control of the large corporation. According to Berle and Means, control had moved away from the owners "ultimat,ely to lie in the hands of the management itself, a management capable of perpetuating its own position" (Berle and Peans, 1967:116). The conclusion of their survey--of the top 200 companies, 106 industrials, 42 railroads, and 52 public utilities for 1930--was that 65% of the companies and 80% of their combined ;,1ealth was controlled by either management or "a legal device involving a small proportion of owner- ship." These figures, when combined with the aggregate data on the diffusion of stock, indicated to Berle and reans 11 the important extent to which ownership and control [had] become separated" (Berle and Means, 1967:110). Shortly after Berle and ~~eans published their results, a Congress- ional committee, the Temporary National Economic Committee (TNEC), under- took its own investigation into the 11 concentration of economic power." Although the findings of the TNEC study, published in 1940, are obviously dated ~ it remains as the only study available which has had "official" data. As a congressional investigation, the committee had the power to 32 subpoena infonnation directly from the corporations under study. Gather- ing lists of the top 20 stockholders in each finn, the committee surveyed the 200 largest nonfinancial corporations. They concluded that "in about 140 of the 200 corporations the blocks in the hands of one interest group were large enough to justify, together with other indicators such as rep- resentation in the management, the classification of these companies as more or less definitely under ownership control" (TNEC, 1940:104). The conclusion that 70% of the top 200 corporations in the country were under ownership control was in sharp contrast to the picture pre- sented by Berle and Means. Their study concluded that 44% of the largest companies were under management control, 21% legal device, 23% minority control, 5% majority, and 6% were privately owned (Berie and Means, 1967: 109). Nearly 70% of the largest finns were found to be under nonovmership control--controlled by either management or some legal device. Several problems stand out in trying to assess corporate control. Researchers engaged in a company-by-company survey have utilized different measures of control and different sources of data to determine stock owner- ship. This makes it extremely difficult to compare findings from different studies. Berle and r1eans, for example, utilized a figure of 20% stock ownership as the dividing line between minority and managerial control. The TNEC study, on the other hand, developed an elaborate scheme which classified stock ownership according to four types: majority (more than 50%), predominant minority (30-50%), substantial minority (10-30%), or a small minori ty (less than 10%). Stock ownership in itself, however, was not viewed as the sole measure of control. In addition to stock, an interest group had to couple i ts o.,.mership with an "active participation 33 in the management, or at least with representation on ,the board of directors," for control to exist (TNEC, 1940:115). The notion that stock ownership has become widely dispersed among a large number of individual citizens has come to be known as "people's capitalism." By 1959 Adolf Berle had adopted this term to describe what he had analyzed two decades earlier (Berle, 1959). He reported in Power Without Property, for example, that six or seven million J\mericans, according to the new York Stock Exchange (which had popularized the term), had become stockholders in 1a rge corporations. J\.ccompanyi ng the notion of dispersion is the allegation that there has taken place a shift in corporate control. Six or seven million people appears to be a large number at first glance, but as Gabriel Kolko (1962) pointed out, this figure only rep- resents a small minority of the U.S. population. Using aggregate data on stockholdings Kolko found that the percentage of people actually owning stock has remained fairly stable for several decades. "In 1937, some 6.6 per cent of the population owned stock; this figure dropped to 5.1 per cent in 1956, and not until 1959 had it increased to 7.9 per cent" (Kolko, 1962:50-51). He goes on to show that not only is stock ownership highly concentrated among relatively few individuals, but, more import- antly, among these individuals stock ownership is highly skewed in favor of the large holders. In 1951, only 2.1% of a11 the people owning stock ) owned 58% of the common stock (Kolko, 1962:51). Thus a relatively small / percentage of the nation's population owns stock, and within this group a relatively small percentage owns the majority of stock. This is hardly an argument , Kolko says, that supports the notion that "the American 34 corporation is owned democratically" (Kolko, 1962:53).' In spite of the TNEC study (which has received little attention) and the relatively small percentage of stock dispersion, both "people's capi- talism" and belief in the "managerial revolution" continue to receive support in the popular press and in academic circles. For instance, one author, although he relies on the TNEC study, interprets the findings differently. Robert Gordon, although in agreement with the findings of the TNEC data that suggest that the concentration of ownership is greater than ordinarily assumed, questions whether control through ownership is as prevalent as the committee implies (~ordon, 1966) . Others have concluded that, as the central feature of the ft merican economy, this separation of ownership from control "has been virtually completed 1tdth the last half-century" (Parsons and Smelser, 1956:252). Citing the work of Berle and Means, Talcott Parsons and Neil Smelser claim that "effective control " has shifted from the owners of capital resources, i.e., "the great industrial magnates and their families," to career managers (Parsons and Smelser, 1956:253,254). These industrial magnates "failed to consolidate their position as the dominant class in society, " primarily because of "high progressive taxation" and the dispersion of stock ownership (Parsons and Smelser, 1956:254; emphasis theirs). The notion that there has taken place a revolution in the economic order that has dissolved "the atom of property," as Berle and Means cla imed, has received support from yet other social scientists. Daniel Bell, in "The Breakup of Family Capitalism," and John Kenneth Galbraith in The New Industri al State, are also representatives of this position, 35 a position which continues to represent the dorninant 'theoretical view within the social sciences. Maurice Zeitlin has captured it well when he writes: The prevailing view is that the diffusion of ownership in the large corporation among numerous stock owners has resulted in the separation of ownership and control, and, by severing the connection between the family and private property in the means of production, has torn up the roots of the old class structure and political economy of capitalism. A new class of functionaries of capital, or a congeries of economic "elites," in control of the new forms of productive property , appear: nonowning corporate managers displace their capitalist predecessors. (M. Zeitlin, 1974:1075) On the other side of the dispute, in addition to the TNEC study and more recent Congressional investigations of the concentration of corporate power (see U.S. Senate, 1974; U.S. House of Representatives, 1964), there are those like Kolko who argue that "to talk of a separation between management and major stockholders in the United States is obviously quite impossible 11 (Kolko, 1962:68). One such critic of the dominant view is Philip Burch, who undertook one of the most comprehensive efforts since the TNEC study to determine the locus of control in the top U.S. corpora- tions. In reassessing the "managerial revolution thesis," he focused his study on t he question of family control. Defining control as "the ultimate overall economic power to decisively affect corporate policy and management," Burch found the "trend" toward managerial control far from complete (Burch, 1972:18). His company-by- company analysis sought to determine "the control status of most of .America's l arge corporations ... as of the mid-1960 1s 11 (Burch, 1972:29). Because of methodological problems associated with using corporate proxy sta tements and Securiti es and Exchange Commission sources (SEC Official 36 Summarv of Security and 10-K Fonns) he was "forced to ' rely" on other sources to detennine who held stock in a given corporation. He searched several business publications, such as Fortune, Forbes, Business Week, Standard & Poor I s, Corpora ti on Record, the New York Times, and ~~oody I s for the years 1950 to 1971, and found a "marked difference in stock owner- ship totals" between the results of his analysis of these business sources and the SEC records. The SEC reports were found to be conservative in their reporting. The SEC requires that only the holdings of the directors and officers and their immediate families be reported, which results in an under-reporting of the holdings of other family members. Burch ' s findings on stockholdings were classified into three cate- gories, according to types of control: (1) probably under family control; (2) possibly under family control; (3) probably under management control. For a corporation to be designated as probably family controlled, two conditions had to be met: (1) "approximately 4-5% or more of the voting stock" had to be held by a family or a family group (Burch, 1972:29); and (2) there had to be "either inside or outside representation ... on the part of a family on the board of di rectors of a company" (Burch, 1972: 30). · /!. corporation was classified as "possibly family controlled" if "some definite signs of family influence" were present. The existence of family influence could be in the fonn of either evidence of a significant amount of stock ownership or representation on the board over a number of years, usually several generations (Burch, 1972:34). \ Surveying the top 300 industrial corporations Burch found that 42.7% / 37 should be classified as probably under family control. Sixteen per cent were classified as possibly under family control and 41 .3% were considered as probably under management control. When the large privately-owned industrial concerns not ranked in the Fortune 500 were added in, family- controlled firms actually outnumbered management concerns (44.7% probably family controlled to 40.0% probably management controlled). Within the group of 13 large privately-owned companies, Burch included such companies as Cargill, Hearst Corp., Hughes Aircraft, Sperry & Hutchinson, and ~ars, Inc. 1-'ost studies fail to include this .group of corporations in their analysis because there is a lack of public information available on them. Yet any study of family ownership and control can hardly exclude these closely held family firms and still make any claim of being comprehens~ve. Like other studies, Burch 1 s analysis failed to take into account the stock held by big institutional investors, banks, and trust companies. Unlike other studies, however, he acknowledged this shortcoming (Burch~ 1972:17). Since his study was published, the Patman reports and the Muskie and Metcalf investigation of corporate ownership have uncovered a significant amount of stock ownership on the part of these institutional investors and trust companies. The question remains, however, of who actually has control of these blocks of stock. Trying to assess the shift in ownership and control patterns over several decades, Burch reranked the top 108 industrial companies by assets to make his data comparable to the 108 industrials studied over three decades ago by the TNEC. He found that 36% of the top 108 corporations in the mid-1960's were probably family controlled, compared to 44 or 49% of the TNEC compa nies i n the late 1930's. His conclusion was that 38 although there has been "a definite trend toward ma~agerial control of big business over the years, the magnitude of this shift in economic authority has generally been overstated" (Burch, 1972:102). "Contrary to most professional and popular opinion,~ interests still play~ fairly prominent role in the conduct of big business affairs in the United States" {Burch, 1972:10; emphasis added). Burch's findings are not only in conflict with those of Berle and Means, but are also in contrast with the findings (from a time period comparable . to Burch's) of Sheehan and Larner. Robert Larner found that in the early 1960 1 s only 14% of the country's top 200 corporations could be considered "dominated by family or other outside interests" (Burch, 1972:5). Sheehan, on the other hand, found 17% of the top 200 corpora- tions under the control of various wealthy families. In spite of the similarity in findings, Larner and Sheehan reached opposite conclusions. Larner, in agreement with Berle and Means, Gordon, and Parsons and Sme15er, concluded that the "managerial revolution" was close to complete. Sheehan, on the other hand, interpreted his findings to mean that family ownership and control was still quite significant. He concluded that his analysis 11 suggests that the demise of the traditional American proprietor has been slightly exaggerated and that the much-advertised triumph of the organization is far from total" (Sheehan, 1970:79). Burch claimed that, compared to his corporate control data and calculations, "Larner and Sheehan failed to identify anywhere from close to 40 to 60 percent, and in one case almost 70 percent (depending on the writer 1 s work and the numerical ranking), of the big family-controlled firms in the mid-s i xties !> (Burch , 1972 :69). 39 Two additional studies that Burch discussed were called jnto question because of their sources of data. The findings of Don Villarejo and Ferdinand Lundberg, although in essential agreement with Burch's own interpretation, re 1i ed primarily on SEC data and the outdated TNEC study, Their findings, however, in contrast with Sheehan's and Larner's pointed toward the continued domination of the large corporation by wealthy families or outside interest groups. Conflicting findings, and contradictory interpretations of similar findings, have led Maurice Zeitlin to question the "alleged" separation of ownership and control and the corresponding notion that kin are unim- portant in modern society (M. Zeitlin, 1974). Surveying the literature, Zeitlin concludes that the evidence does not suggest the kind of "consen- sus" among social scientists on the issue that Ralf Dahrendorf thought existed. The problem, as Zeitlin understands it, is that this "astonish- ing consensus," based primarily on Berle and Means' findings, has achieved the status of general theory. Zeitlin refers to the "explanation" of the U.S. class system provided by the proponents of the managerial revolution thesis as a "pseudofact." Accardi ng to Zeitlin, pseudofacts "serve to deflect attention from critical aspects of social structure, determinant social relations, and basic social processes. They may inspire not merely explanations, but 'inferences' and 'theories' as well, which further confuse and obscure social reality" (M. Zeitlin, 1974:1074). In the same article he makes numerous references to the need to study kin units i n order to determine the locus of corporate ownership. H1ese kin units, he argues, are the central element in a thorough analysis of the re l ationships between class structure and corporate ownership. As 40 he says, It is known that a oreat number of related individuals may participate in the ownership of a family bloc, utilizing a complex holding pattern to keep control concentrated, despite the diffusion of ownership. If control is exercised through entangling interests in several interrelated corporations, rather than limited to one, then such kinship information is vital to an understanding of the control structure. Indeed, the kinship relations between the top officers, directors, and principal shareholders of the large corporations (and banks) are the least studied but may be the most crucial aspect of the control structure. (M. Zeitlin, 1974:1099) Summary: The Family J\.nd Power Structure Debates There is widespread agreement on the central role of the family as a "fundamental unit of class membership." Even theorists from as divergent perspectives as Parsons and Sweezy agree that a strong relation- ship exists between class and family. It is important to emphasize, however, that while kin relations are related to social class, the kin- · ship structure is not the sole determinant of class. It is more appro- priate to describe the relationship by saying that social class is rooted in various kin units; class and kin are also related to a particular form of economic structure or organization, namely property. Because of the close relationship between kinship and class, particularly among the upper classes, th~ social class system has be~n described by Bottomore as operating "largely through the inheritance of property" (Bottomore, 1966:11). It is this economic system, represented by property, which determines one ' s class along kinship lines. The intersection of governing class theory, kinsh i p studies, and the corporate control debate suggests that careful studies of large capi- 41 talist families can contribute to our understanding ' of several long- debated questions. Governing class theory places a great deal of emphasis on intermarriage, clubs, schools, and corporate interlocks in analyzing upper class cohesiveness. Upper class kinship solidarity and economic security resulting from inheritance patterns suggest an alternative to the "isolated" nuclear family thesis. The linkages between kinship and property are important in examining the relationship between upper class families and other institutional structures. The literature on ownership and control suggests that kin groupings may represent centers of control in the large corporations. Exploration of these close relationships among social class, kinship, and property should enable us to develop new evidence with which to support or call into question the work of plural- ists and governing class theorists. Governing class theory, as we have seen, provides an alternative to pluralism. Both have implications for the development of a th2ory of the state, i.e., for analyzing the role of the state in capita1ist society. Pluralism "views the state as a pluralist, aggregating mechanism in which agencies, programs and legislation are substantive responses to the demands and interests of competing groups" (Esping-Anderson et~-, 1976: 186-187). Governing class theory fits into what has been called the instrumentalist approach. "An instrumentalist theory of the state is a theory in which the ties between the ruling class and state are system- atically examined, while the structural context within which those ties occur remains largely theoretically unorganized" (Gold et tl·, 1975:31). David Gold, Clarence Lo, and Erik Wright outline the research agenda of the i nstrumentalist perspective. 42 [It] has focused primarily on studying the nature of the class which rules, the mechanisms which tie this class to the state, and the concrete relationships between the state policies and class interests. The method consists of detailed studies of the sociology of the capitalist class, in the first instance simply to show that it exists; studies of the direct personal links between this class and the state apparatus, and links between the capitalist class and intermediary institutions (such as political parties, research organizations, and universities); specific examples of how government policy is shaped; and reinterpretations of episodes from the annals of history. (Gold et~-, 1975:32-33) Ralph Miliband (1969), G. William Domhoff (1967, 1970), Gabriel Kolko (1963), and James Weinstein (1969) are usually identified as examples of instrumentalist theorists (see Mollenkopf, 1975; Gold et~-, 1975; and Esping-Anderson et~-, 1976). As Gold (1975) and Domhoff (1976b) have both pointed out, much of the empirical work undertaken by the instrumentalists has been an attempt to challenge the conclusions of the pluralists. Instrumentalists have sought to show that the state is not a neutral entity responding to pressure groups, but rather that it functions "in terms of the instrumental exercise of power by people in strategic positions 11 (Gold et~-, 1975:34). The structuralist theory of the state is usually posited as provid- ing an alternative perspective. Nicos Poulantzas, a French Marxist, is identified as the major proponent of this approach to studying the structure of the state. The theory emphasizes the "ability of the state to reproduce class relations and class domination through structural relations that need not be immediately visible 11 (Esping-Anderson ~ .!l_., 1976:189). In their recent Monthly Review article, Gold, Lo, and Wright (1975) present the fundamental thesis of the structuralist approach. 43 [ItJ is that the functions of the state are broadly determined by the structures of the society rather than by the people who occupy positions of state power. Therefore, the starting point of the structuralist analysis is generally an examination of the class structure in the society, particularly the contradictions rooted in the economy. Structuralists then analyze how the state attempts to neutralize or displace these various contradictions. The structuralist theory of the state thus attempts to unravel the functions the state must perform in order to reproduce capitalist society as a whole. (Gold et al., 1976:36) -- While these two approaches, instrumentalist and structuralist, represent alternative theoretical schemes for analyzing the capitalist state there is no reason why the two perspectives could not be merged to provide a comprehensive theory of the state. As Gold (1975:30) points out, "there is no necessary incompatibility among these vari ous strands of thinking." Recent discussions of the two perspectives, however, make the point that "many Marxists have treated them as quite irreconcilable, and much of the recent work on the state has taken the form of a polemic against one or another alternative perspective" (Gold~.!)_., 1975:30). Poulantzas himself seems to be the most insistent on treating them as totally different approaches. In Classes i!:1.. Contemporary Capitalism, for example, he breaks with what he calls "the misunderstanding of the 'functionalist-institutionalist ' tradition, which has always spoken of the role of ' institutions ' in the training and distribution of 'individuals,' particularly under the heading of the 'socialization process' ... " (Poul- antzas, 1975:34). A convergence between the two approaches is suggested by two recent articles referred to above, "~~arxist Theories of the State" (Gold et.!)_., 1975) and "~odes of Class Structure and the Capitalist State" (Esping- .A.nderson et~-, 1976). This view of the state emphasizes the structures 44 of the state as objects of class struggle. Accordtng to this perspective, the capitalist class attempts to create state structures which channel working class political activity in ways that do not threaten capitalist political dominance and objective interests. Working class challenge makes the success of such attempts problematic. A political class struggle perspective on the state tries to locate the state within the dialectical relationship between class dominance and systemic constraints. (Esping-Anderson et al., 1976:190) The political class struggle perspective builds on the work of Claus Offe (1972) and James O'Connor (1973). Their major emphasis is on the capi- talist state rather than on the state in a capitalist society. They analyze the institutional mechanisms utilized by the state to legitimize the dominant ideology and to mediate the relationships between the classes. The convergence of the instrumentalist and structuralist approach has been succinctly summarized as follows: ... capitalists may manipulate the state, but they do not do so just as they please. The instrumental domination of the capitalist class is constrained by the structures of the state fonned out of past class struggles, by the exigencies of current class struggles and by the contra- dictory consequences of state activity for future class struggle. (Esping-Anderson et~-, 1976:198) This study does not directly address these theoretical questions regarding the nature of the state. Our concern is more limited. The emphasis is on one kinship group of the upper class. An empirical in- vestigation of this sort enables us to examine evidence directed at more limited theoretical issues. These issues, having to do with the relation- ship between kinship and property, with ownership and control of corpora- tions, and with the importance of upper class kin groups, are the major 45 concern of the present study. Empirical investigations are necessary to detennine the locus of corporate ownership and control and to detennine the mechanisms of control for both corporations and state apparatuses. It is hoped that this study can contribute to such an investigation. This is not to say that the broader theoretical issues raised by recent debate on the nature of the state, class domination, etc., are not important, but only that it is not the focus of this study. This study, however, may have some implications for the larger debate. For example, if it is true that the dominant fonn of property in contemporary capitalism is the corporation, then analysis of who owns and controls it is important. The corporation does not exist merely at the 1eve1 of abstraction often analyzed by the structuralists. It exists in relationship to classes and in interaction with both those who own t he productive forces, i.e., the capitalists, and those who do not. It seems to me this is what the instrumentalists and structuralists are both talk- ing about. The state may be an instrt.ment of the capitalist ruling class, as the instrumentalists suggest, but the capitalist state is also struc- turally limited (cf . Sardei-Biennann !!_~., 1973). Focusing on one family is not going to prove that "state policies constitute the general interests of the capitalist class" (Sardei-Biermann et~-, 1973:66). It will, however, expose some of the "concealing mechanisms" of the state. Such an empirical investigation could reveal new mechanisms of corporate control and private appropriation used by one kinship group. Additional empirical work of a more comparative nature would be necessary to expand the analysis ; additional theoretical 1-,ork would be necessary to thoroughly analyze the nature of the state. 46 1 CHAPTER II: METHODS: CASE STUDY ANO NETI~ORK ANALYSIS The purpose of this research is to explore several interrelat~d theoretical propositions about class and power in the United States, through a comprehensive and in-depth historical case study of a large and wealthy kin group. A detailed examination of a single case in all its wide ramifications provides the evidence to make generalizations about the validity of several propositions outlined in the previous chapter. The study explores the familial, educational, economic, religious, cultural, and political connections of the descendants and in-laws of Frederick Weyerhaeuser. The inter-institutional linkages of this kin group are analyzed through five generations. A network analysis is em- ployed to trace intra-familial relationships (genealogies or kinship analysis) and external relationships (the individual's links to other institutional structures). The major unit of analysis or major focus is the family, more specifically the kin group. Within this framework the study will present biographical data on individuals within the kin net- work. In the genealogy the individuals are linked together through familial relations. In the other networks it is the various institutional 1conversations and correspondence with G. William Domhoff were in- valuable in formulating the ideas developed in this chapter. The develop- ment of the methodological framework for this study is heavily indebted to him. His recent work (1976a) is the first systematic attempt to devise "The Methodology of Power Structure Research." 47 units that are linked together through individuals. These networks depict the external linkages with and between other institutional structures: corporations, schools, foundations, churches, and various civic and governmental organizations. Although these linkages are represented by individual persons, the relevant links are between institutional entities (cf. Craven and Wellman, 1973). The case study of the \>leyerhaeuser family can help answer several research questions derived from the power structure debate between pluralists and governing class theorists. Do upper ciass families remain a coordinated and cohesive kinship group? If so, how? What are the mechanisms which enable them to do so? How are they able to maintain control over large corporations? How do their activities shape the cultural, reljgious, and intellectual milieu? How do their activities influence the selection of political candidates and shape governmental policy? The literature on kinship and class indicates that the family~ · the~ unit of the class system, but as the last chapter revealed, sociologists have not undertaken empirical studies of upper class kin- ship patterns. Their attention has been focused on the lower and middle classes. The proponents of governing class theory recognize the importance of the family in the upper class, but have centered their analysis on national units (Mills, 1959; Hunter, 1959; Domhoff, 1967), cities (Baltzell, 1958; Hunter, 1963), clubs (Domhoff, 1974a), or 48 political parties (Domhoff, 1972a). 2 Those seekinq to determine the locus of corporate control have turn- ed to surveys of the top 200 or 500 corporations to collect evidence for or against a particular type of control. The only real discussion of family capitalism, Daniel Bell's short essay, is void of any empirical evidence in support of his contention that family capitalism has de- clined (Bell, 1962). The Weyerhaeusers as a kinship unit were chosen for an intensive case study for a variety of reasons related to personal interests, location, manageability and availability of data, and political concerns. The impetus for constructing a complete network of this kin group's linkages to a variety of institutions came from a study of the timber industry in the Pacific Northwest. One aspect of this study sought to determine who owned and controlled the six largest timber corporations operating in Oregon and Washington. This led quite quickly to the realization that at least two of these companies had ties to the Weyerhaeuser family, The Weyerhaeuser Company and Boise Cascade Corpora- tion. George H. Weyerhaeuser and his uncle, C. Davis Weyerhaeuser, are on the board of the Weyerhaeuser Company. C. Davis's nephew or George's cousin, Edward R. Titcomb, is on the board of Boise Cascade (Dunn, 1974). ~t this stage in the research it was sti 11 assumed that the Weyerhaeuser 2The only exception to this generalization is a recent study by Michael Merlie and Edward Silva (1975), which examined the kinship ties of American presidents. They found 21 of 36 presidents to be part of a 219-person kin network {Merlie and Silva, 1975:162), a finding they interpreted as supportive of a governing class theory. 49 company had evolved from a family firm to become a publicly-owned multi- national corporation. This assumption was based on the commonly held belief about the company and a survey of public records. Security pnd Exchanae Commission records and company proxy statements, for example, revealed that the two members of the family who continue to sit on the board of the Weyerhaeuser Company held less than 1% of the outstanding common stock (Dunn, 1974). Later in the present study we will see that this initial evaluation of the family 1 s holdings was a gross under- estimate. ihe point I want to make here is that, in the early stages of the research, the popular notions, often supported by social scientific theories, about the absence of familial involvement in the modern corpo- ration and political processes were thought to hold. Although historians have described the historicai ties between the Weyerhaeusers and various timber corporations, many people in the region where the firms have operations are unaware of links between Weyerhaeuser Company, Potlatch Corporation, and Boise Cascade. even found that middle-level managers at Boise Cascade were unaware (or would not admit) that a member of the Weyerhaeuser family continues to sit on the board (personal communication, ~arch, 1974). Many of these historical linkages are documented by three prominent historians, Ralph Hidy, Frank Hill, and Allan Nevins, in a book entitled Timber and Men: The Weyerhaeuser Story (1963). This book, sponsored and paid for by the Weyerhaeusers, depicts the involvement of the first three generations in numerous timber firms. But in spite of this work the history of these firms and of the family remains veiled. ftnother reason for selecting the Weyerhaeuser family was the question 50 of its political and economic influence in the Pacific Northwest. There is concern in this region over the powerful position of the Weyerhaeuser company. Its ownership of 5.7 million acres of forests, much of it concentrated in Oregon and Washington, makes the Weyerhaeuser Company the largest owner of private timber in the world. In the state of Oregon the company is the second largest private employer, employing nearly 5,000 people. The company's labor practices, its forest management decisions, its environmental practices, and _its log export policies receive a great deal of attention in the mass media in the region. Yet these activities are seldom discussed in relationship to questions of corporate wealth and power. The Weyerhaeuser family is not one of the families at the very top of the national stratification system--although it is perhaps the wealthiest and most powerful family in the Pacific Northwest. It is a / family, as Ferdinand Lundberg (1937) says, that belongs to the "less wealthy family dynasties. 11 Nonetheless, Lundberg claims such families resemble the top echelon in every respect except in the amount of accumulated wealth. This being the case, these families should be expected to match up to the top families, like the Rockefellers or du Ponts, in terms of upper class indicators and inter-institutional link- ages. The advantage of selecting the Weyerhaeusers was that, while the complexity of internal and external family relationships was assumed to be less than that found in some of the better known, wealthier, and probably more powerful families, the forms should follow similar patterns. The family 1 s major activities in the corporate sphere have been limited 51 to one industrial sector--the timber industry. Among other things this avoids the question of bank versus corporate control of the large corporations (cf. Fitch and Oppenheimer, 1970), since the family does not appear to have any bank interests. The only links to banks have been through director interlocks with several banks. This is not to say that other theoretical and empirical questions like the financial control debate are not important, but only that this study deals with the issue of family control. Researchers interested in ascertaining information on the upper class and its institutions have a number of avenues open to them. There are a number of social histories of the "great_family dynasties" (e.g., H. O'Connor, 1933; Abels, 1965; Fosdick, 1956). These reflect both favorable and hostile treatments, but rarely go beyon~ descriptive accounts of the families activities. In many locations the "society page" has made many upper class families household words, but others remain hidden behind the exclusive institutions which train and maintain them. The Weyerhaeusers follow this latter pattern, preferring to remain a "private" family. Historically, they have shied away from the publicity which characterizes families like the Rockefe1lers. If the Weyerhaeusers are known at all it is probably the result of familiarity with the corporate name, the kidnapping of George H. Weyerhaeuser (today the president of the Weyerhaeuser Company) in the late 1930 1 s, or recognition of the name attached to a campus building or church room as the result of philanthropic activity. This reclusiveness is both challenging and problematic for the researcher seeking to determine the extent of the family's involvement 52 in various social spheres. Uncovering the "private" affairs of a major family--its stock ownership, political involvements, philanthropical activities--is a challenge not unlike that of a detective seeking to detennine "who-done-it. 11 Many of the institutions associated with the upper class are more secretive than the families themselves. This is particularly true of corporations, which protect themselves and their largest stockholders behind a number of legal 11 safeguards. 11 (One of the initial reasons for incorporating was to protect wealthy individuals from the legal reper- cussions of their activities.) Those seeking access to this information have often resorted to the courts or to illegal means, such as the methods used to obtain material for documents like the Pentagon Papers. Congressional investigations which lack subpoena power often fail to achieve cooperatio~ from corporate officials; the most recent example is the Metcalf and Muskie investigation (U.S. Senate, 1974) of corporate ownership. Power structure research, however, has utilized a number of · data sources to obtain infonnation and break through this "blockade" of secrecy. This study has used public records, such as probate court proceedings and foundation tax returns, in addition to standard reference works and interviews. A detailed examination and analysis of one social unit, in this case a kinship group, can provide new insights and can deepen our understanding of how a particular unit operates. Exploration and description of the vario us components and concepts which constitute the unit under study are what characterize a case study. The description develops an intimate and thorough knowledge of one particular situation or single case. 53 Inferences and generalizations are based on this compr~hensive kno~iledge of all aspects of the particular case. Like any method utilized by the social scientist, however, the case study approach has both strengths and weaknesses. Many of these are common to any method utilized in social research, others are unique characteristics of the case study. In areas of inquiry where previous studies are scarce the case study approach has proven fruitful (cf. Selltiz et~., 1967). The case study's comprehensive examination of a single case often uncovers unexpected findings. The present study, for example, found no discussion in the literature about the "family office" or the "annual family meetino." Yet the discovery of these two organizations within an upper class family led to the realization that these organizations or institutions could potentially provide a mechanism for family control and could be an explanation for the persistence of family involvement in a variety of institutions. The major limitations of the single-case approach are related to the questions of external validity and the reliability of the findings. As Campbell and Stanley (1970:5-6) point out, the problem of external / validity asks the question of the generalizability or representativeness of the findings. An intensive study of one upper class family obviously does not pennit generalizing to any family regardless of class. It does enable us to show how one upper class kinship group operates. The next chapter attempts to demonstrate how the Weyerhaeuser family is linked to institutions which others have utilized as indicators of upper class standing (cf. Domhoff, 1967, 1970); this is an -attempt to show the representativeness of the Weyerhaeuser family. If it is shown that the 54 family is a representative case, the study could sugg~st that other upper class kin groups may be organized and function in a manner similar to the Weyerhaeusers. This case study of the Weyerhaeusers is also an historical case, which makes it possible to overcome some of the temporal limitations often associated with other methods. Building in an historical dimension-- in this case, studying the family through five generations--will enable us to identify changes and shifts in the role of the Weyerhaeuser kinship group in the United States during the last century. The representativeness and reliability of the sinqle case method are probably the two most significant limitations associated with this method. In addition to the use of historical analysis, three other techniques are utilized to resolve these problems. (1) In order to compare the Weyerhaeusers to other kin groups with similar characteristics it will be necessary to specify units of analysis or concepts. (2) Brief comparisons between the Weyerhaeusers and other kinship groups of the upper class will be made to show that the Weyerhaeusers are not an atypical case. (3) ~ethodological triangulation will be used to reduce problems associated with internal validity (Denzin, 1970:309). In discussing the problems associated with "an empirical analysis of a single case," lipset, Trow, and Coleman (1970) caution against trying to generalize from a case that is merely a descriptive account of some phenomena. _ They stress that if the researcher wants to develop empirical generalizations or theories by usinq this approach, certain conditions must be met, the most important of which is the ability to demonstrate that the unit selected for intensive study follows patterns 55 and shares characteristics of other units in the popula~ion. To generalize beyond the single case it is necessary to delineate "units of analysis" and "characterize the units according to certain general conceots or properties" (Lipset et~-, 1970:129; emphasis theirs). Delineation of comparable units of analysis is carried out through the specification of general concepts. This is a similar strategy to the one used by Baltzell (1958) and Domhoff (1967 and 1970) in their works in developing indicators of criteria for upper class membership . Using the concepts developed by Domhoff the next chapter demonstrates that the Weyerhaeuser family meets these upper class criteria. To show that the Weyerhaeusers share common traits with ot~er upper class families, evidence will be presented which roots individuals of the family in the national upper class. An extensive analysis cf this one kinship group uncovered a variety of mechanisms associated with family control, cohesiveness, and solidarity. Limited infonnation, however, is available on similar patterns among other upper class families. But in areas 1-,here comparable infonnation is known these other upper class kin groups are compared to the Weyerhaeusers. This provides a framework suggestive of the role and activities these groups play in a variety of i nsti tuti ons. In recognizing the need to specify "units of analysis ... according to certain general concepts," the role of the family in a capitalist ruling class needs to be clarified. Recognition that the fusion of the family and property systems maintains a class system, as Bell (1962:40) points out, enables us to define capitalism in class tenns. This is a good starting point, since, as we saw in the last chapter, there is almost 56 universal a9reement in the literature on the familial ba?is of class. In speaking of the family in the present study, the reference is to kinship groups unless otherv,ise stated. When the reference is to the nuclear family, rather than kin group, it will be so specified. Kinship is often defined as a relationship based on common ancestry, but this is too broad for our purposes. In the case of the Weyerhaeuser kin group, for example, Frederick Weyerhaeuser and his original business partner, F.C. Denkmann, married sisters. If the concept of kinship were defined broadly this would place the descendants of both within the same kinship group or family. In the present study these are treated as two separate kinship aroups, although they are linked together in a broader kinship system. 3 Kinship is thus viewed in a manner similar to Firth et~- (1969:3), who defined it as a "set of ties socially recogniz- ed to exist between persons because of their genealogical cpnnections." In the present study the Weyerhaeuser family is defined as the descen- dants of Frederick Weyerhaeuser, including those who have married those descendants. This is the kinship group which constitutes the case study. Sometimes I use the terms "major families" or "the family" to describe the kinship group. The Weyerhaeusers themselves use these terms, and as such give us an insight into their own perception of the import- ance of the family. This is a good example of the way in which the upper 3In the broader kinship group all the Denkmanns are cousins of all the Weyerhaeusers. The definition of kin group used in this study is male oriented rather than female oriented (the latter would be the case if t he Bloedel family was used as the starting point); both the Denkmann and Weyerhaeuser kin continue to define themselves in separate genealo- gies; see for example a privately published book, Weyerhaeuser-Denk!T'ann Centennial (1960), which contains a separate "Family Tree" for each fami 1y. 57 class kinship group views itself as special. The upper class can be defined, in Baltzell 's terms, as a "group of families, whose members are descendants of successful individuals (elite members) of one, two, three or more generations ago" (1958:7; emphasis his). Or as Domhoff says, although the concept of the upper class "points to the great wealth and unique life styles of these ·intermarrying and interacting families of high social standing, the social upper class can be defined as people who are listed in certain social registers and blue books, people who attend certain private schools, and people who belong to certain exclusive social clubs" (1970:32). The term ruling class means that the wealthy families not only m-1n a disproportionate amount of society's wealth but have control over the ownership of that wealth. Wealth normally takes the form of property, and in capitalist society the ownership of property is usually either in the form of land or capital. In Marxist terms property is "the private ownership of the means of production." The ruling class can thus be defined as the group in society which owns and controls the means of production. This relationship to property forms the basis of classes in capitalist society. This brings us to the notion of a capitalist ruling class as consti- tuting propertied families. In reality the concept of upper and r·ulinq class cannot be separated (cf. Baltzell, 1958, and Bottornore, 1966). But for analytical purposes, primarily to explore whether or not the upper class also rules (i.e., is a ruling class), it is important to differen- tiate between the two concepts. The upper class refers to families own- ing property; the term ruling class is reserved for the group having 58 control over the property. The property referred to is pr:oducti ve property, which usually takes the form of capital invested in stocks and/ or bonds. To determine if the upper class constitutes a ruling class, Bottomore (1964:39) says it is necessary to show that the "upper class has been successful in perpetuating its ownership of property." If this can be empirically demonstrated we can then speak of the continuity of a family-dominated capitalism in which the ruling class "owns the major instruments of economic production in a society" (Bottomore, 1964:37) and "also controls the means of political domination" (Bottomore, 1966: 77). An unequal distribution of power also is implied in the concept of a ruling class. But power, as Kadushin (1968) points out, is a notion that is difficult to define and cannot be measured directly. In speaking of such a notion, James Coleman (1970:9) notes that many of the proper- ties or concepts of interest to the social scientist "cannot be directly measured; they are 'underlying' r or 'latent,' and all that can be observ- ed and measured are their manifestations." Others have agreed that because power is not directly observable it must be inferred from certain indicators which point to the manifesta- tions of power (cf. Kadushin, 1968; Barton, 1970; and Domhoff, 1976a). Kadushin (1968) is one of those who refers to the properties of power as a disposition concept. Because direct measurement of a disposition concept like power is not possible, it is necessary to develop indicators from which power can be inferred. Traditionally power has been inferred either (1) by locating a specific decision and ascertaining who deter- mined its outcome, or (2) by locating an elite and observing its use of 59 power (Kadushin, 1968:697). The first method is the decisional method popularized by Robert Dahl (1961). The second incorporates both the positional and reputational methods in seeking to determine power relationships. The central theoretical issue in each case is, Kadushin says, "the extent to which [ the power elite] are interconnected" (1968:697). This conclusion led him to suggest that the "social circle" is the best method for resolving some of the conceptual problems associated with elite studies. This method enables the researcher to move beyond the formal structures of power to study informal power relations and their links to formal organizations of power. Within this scheme power is inferred from institutional positions and other qualitative indicators. The concern is with linkaaes and connections between formal and informal organizations. The social circle is represented by "the major informal mechanism which links power persons and powerful organizations" (Kadushin, 1968:691). In the present study the kinship group is examined as being just such an informal linking mechanism between formal structures of power. The existence of a link between two individuals or organizations does not necessarily mean that there is a relationship of power. For example, kin ties demonstrate a genealogical relationship, but do not necessarily prove that kin interact. As John Porter (1965:526) says, kinship charts do not measure anything, but the "links do serve an important function as one of the means through which elite groups co- ordinate their activities" (cf. Whitley, 1973). The existence of kin connections in and of itself is not sufficient to infer power in the kin 60 aroup. And while they point to possible avenues of interaction and coordination they do not prove that these activities take place. To demonstrate interaction, other testimony must be presented. For example, the existence of an "annual family meeting" and other get-togethers could be presented as evidence of extensive kin interaction. Such inter- action would indicate that the kinship group of upper class families is a cohesive unit. To demonstrate that the family works together to realize common interests, it is necessary to present evidence that they coordi- nate their activities. The role of the "family office," decisions made at the family meetings, the methods of corporate control, and political campaign contributions are discussed to show family coordination. Family control of a particular corporation will be said to exist if three conditions are met. Each is indicative of the possibility of control. When all three are present the likelihood of one family's controlling a corporation is greatly increased. Control of a corporation by a kinship group is said to exist if: (1) members of the family own a combined total of 10% or more of the outstanding common stock; (2) the individual members of the family coordinate their investments through fonnal (holding company or foundations) or informal (annual meeting or family office) means; and (3) family members are represented on the board of directors or in management positions . These requirements are more stringent than the criteria for control that others have utilized. Burch (1972), for example, classifies corpo- rations as under family control if one family owns 5% or more of the stock and has representation on the board. Any assessment of control that uses a percentage of owned stock of less than 50% assumes that the 61 rest of the stock is widely dispersed among a large number ,of small stockholders who are unable to act together. If this assumption is made, the difference between 5% or 10% is relatively unimportant. \4hat is important is to be able to demonstrate how a group of individuals, in this case a kinship group, can coordinate its investments. The existence of family representation on the board or in management over successive generations, in addition to representing evidence of control, provides testimony of family continuity. As a method for studying linkages and connections, social circles and network analysis provide the means for developing "specific propo- sitions" from which pm'ler can be inferred. ( Domhoff [ 1976a] has illustrated the use of networks and matrixes in analyzing power structures.) Network analysis is a way, as Craven and Wellman (1973:58- 59) say, "to elicit infonnation about important 'indirect' ties between people or institutions not directly or obviously related to each other." Networks specify sets of linkages among people and organizations. They "concretely describe the structure of relationships that characterize a group or community or class" (Domhoff, 1976a:162). Network analysis, as conceived for purposes of power structure research has "crucial sociological and mathematical differences" from the approach of conventional sociometry. These differ- ences involve the fact that conventional sociometry focuses primarily on the social relations among the people in a network, conceiving the "points" or "nodes II of the network as people, and the linkages among the points as social relationships (e.g., "linking," "hating," "owing"). This unusual approach does not give full play to the other basic type of social tie discussed by social scientists--membership. We not only "relate" to people and institutions, but we are "members" of groups and organizations, a fact which tends to be put aside, conceptually speaking, in socianetry. (Domhoff, 1976a:164) 62 In the study of one kinship group, the kin network or, genealogy appears as a natural starting point in our analysis. The kin network is an elementary form of net\<1ork analysis, or as Barnes (1972) says, "a paradigm case," linking individuals through affinal (marital) or con- sanguineal (blood) relations (cf. Schusky, 1965). Anthropologists have used genealogical inquiries to explore kinship patterns in modern society (cf. Firth et~-, 1969; Bott, 1971). In sociological research, ho\'1ever, the genealogical technique has not been fully explored. Its usefulness is readily apparent in that it provides not only 'information on kin connections, but more importantly information for tracinci other net\'wrk patterns {cf. Bott, 1971). The case study of the Weyerhaeuser family and its networks utilizes a variety of sources to obtain data on the kinship patterns and activities of individual family members. In the initial stages of the research biographical sources were consulted to ~onstruct a genealogy of the kin- ship group. Reference works like Who's Who ~- .America, National ~1-2..::. paedia 2.f .American Biography (NCAB), Directory of American Biography {D.A.B), Current Bioqraphy (CB), and New York Times Obituaries Index pro- vide information on parental and affinal kinship links {cf. NACLA, 1970). They a 1s o pro vi de i nforma ti on on the i ndi vi dua 1 1 s 1i nks to other networks, particularly corporate, educational, and civic ones. The first genealogy was constructed solely from public sources. It produced only 32 individ- uals in the fifth generation. It did, however, identify all of the family members from the third and fourth generations who are active in corporate affairs. Later in the research the "official" 11Weyerhaeuser Family Tree" became available, which revealed 82 individuals in the 63 fifth and three in the sixth generation. Networks a 1s o can be constructed by i ntervi ewi ng members of the family who provide the kinship and association links to be charted. Thus a combination of bioqraphical sources and interviews can be used to achieve a more comprehensive picture than reliance on only one source could produce. One of the most usefu1 methods of network construction is a technique of sociometrics called "snowball" sampling. This method is usually used to obtain a sample and detennine the relationships amonC! individuals in a given network (cf. Barnes, 1972; Kadushin, 1968). Barnes (1972:23) describes the snowball technique as a way of obtaining a sample "where each successive respondent defines several more who should be taken into the sample." In the present study it also proved to be an excellent way of gain- ing access to individuals who otherwise might have peen uncooperative in granting interviews. Each person who was interviewed was asked if there were other members of the family \<1hom they thought would be wi 11 i ng to be interviewed. These people were then interviewed and asked to supply additional names of family members. Lists of individuals obtained solely by reliance on the snowball technique, as Barnes (1972:23) points out, "cannot be regarded as typical or representative of members of the network as a whole." This limitation reveals one of the primary reasons for triangu1ating data. The triangulation process combines two or more different methods of gathering data to study "the same empirical units" (Denzin, 1970:308). It is most useful as a check on internal validity. This study of the Weyerhaeuser kin draws data from archival records, interviews, and public 64 documents. In this way data from one source can be corrobQrated with information from other sources in order to remove some of the problems associated with internal validity. Data triangulation not only uncovers additional infonnation but also provides a validity check on the evidence by combining data from a variety of sources (cf. Webb et.!!_., 1966). Nonnan Denzin (1970:221) also makes the claim that triangulation "permits generalizations to a broader population by studying a subunit from it." This must be accompanied \1as other than very wealthy was to carry modesty to the point of telling less than the whole story" (Beale, 1964:162). Another historian , Edward N. Saveth, has questioned certain asser- tions made by Parsons and Bell about the internal and external structure of the family in fl,merican society. Focusing in particular on the patri- cian family, Saveth believes the conclusions Parsons and Bell reach are not supported by the evidence. These theoretical studies of the American family, claims Saveth, have been lacking in an historical perspective. "There is every indication," Saveth writes, "that there are additional variables, beyond those stressed by Parsons and !3el1, which condition family structure" (Saveth, 1963:242). In calling for intensive research into family history, Saveth suggests that "oniy a history of the American family could comprehend these variables and provide basis (sic] for an accurate typology of the patriarch at various stages in ,n,merican history as well as a contribution to class theory" (Saveth, 1963:243). He suggests a model of the patrician class which "includes family history with stress upon the factors, material and cultural, which make for family continuity; the structuring of real types of the patrician family centered in factors related to the family's internal and external relationships" [ sic] (Saveth, 1963:252). 72 studying internal relationships, as Saveth points out, enables us to comment on the debate over whether the tl.meri can family should be characterized as "extended" or "isolated nuclear" type. Family history and genealogy emphasize internal structure, but also enable us to uncover many external linkages. The study of a family's external relationships in turn sheds light on the question of family structure and business control. Historians such as Beale and Saveth, and more recently sociologists such as Zeitlin, Ewen, and Ratcliff, have demonstrated the usefulness of the genealogical method. Zeitlin et~. for example have used it to show the way in which shared common economic interests and kinship bonds are often inextricably intertwined (cf. Zeitlin~~., 1974:108). A genealogy can be a particularly useful tool in identifyin9 members of the same kinship group who sit on the same corporate board but who have different surnames. The following genealogy (see Chart I) is a replica of the official "WEYERHAEUSER FAM!LY TREE". It hangs in the Famil,Y Office in St. Paul, Minnesota, and can also be found in the files of the Public Relations Department of the \✓eyerhaeuser Company in Tacoma, Washington. This 9enealogy of the Weyerhaeuser family enables us to trace the family through five successive generations and to study changes in family structure during the past century. In the discussion that follows the focus will be on individuals i n the \foyerhaeuser fandly. It is important to keep in mind, however, t hat the major unit of analysis is the kin group, and that the point in studying one wea 1t hy family is to shed 1i ght on our unders tan ding of the CHART 1 73 WEYERHAEUSER FAMILY TREE II ~-,:·.:.:: 74 class structure and the political economy of the United States. E. Digby Baltzell in his analysis of the class s4ructure of Phila- delphia distinguishes between "two aspects of high class position, an elite and an upper class" (Baltzell, 1958:6; emphasis his). The concept of an elite "refers to those individuals who are the most successful and stand at the top of the functional class hierarchy" (Baltzell, 1958:6; emphasis his). The individuals who Baltzell identifies as the elite are the successful individuals, the ones who have made it, who are the leaders of their "chosen occupations or professions." This is another way of saying that they represent the first generation of the upper class family. For the upper class, in Baltzell 's scheme, is that "group of families, whose members are descendants of successful individuals (elite members) of one, two, three or more generations ago" (Baltzell, 1958:7; emphasis his). Frederick Weyerhaeuser fits Baltzell 's definition of an elite person, a member of what Baltzell called the "business aristocracy." His children, however, were trained and quickly assimilated into the institutions of the upper class. Thus, as we will shortly see, by the second generation the Weyerhaeusers had taken root in the American upper class. By employing Oomhoff's (1967, 1970) social indicators of upper class standing we can show the family's rapid ascent into the upper class. These indicators--social register listing, elite prep school attendance or club membership, and/or whose father was a millionaire entrepreneur and who attended a select school or was a member of an exclusive club-- are the criteria established by Domhoff to ascert1in an individual's 75 ties with the upper class. We use them here to demonstrate the linkages between individual members of the Weyerhaeuser family i,n successive generations and upper class institutions. The starting point for our genealogical excursion into the Weyer- haeuser kin group is Frederick Weyerhaeuser. 1 Born in Germany on November 21, 1834, Frederick Weyerhaeuser emigrated to the United States at the age of eighteen. All histories of the family trace its origins back to Frederick Weyerhaeuser. There is, however, disagreement about his family of orientation. Some accounts--Sarah Sale's Timber Concentra- tion in the Pacific Northwest (1945), for example--claim that Frederick Weyerhaeuser started from "humble beginnings. 11 The more authoritative linformation on Frederick Weyerhaeuser's life comes primarily from the following sources: Fortune, 1934; Weyerhaeuser Family, 1940; Salo, 1945; Haubera, 1957; and Hidy et al., 1963. All of these accounts rely heavily on Pioneer Lumberman (Weyerhaeuser Family, 1940), a book privately published by his children and their spouses. The work by Hidy et al. is usually viewed as the definitive work on the Weyerhaeusers and their companies. The major focus in Hidy et al., however, is on the various business enterprises in which theWeyer- haeusers have been involved. Information on the family itself is rather sketchy. This work is viewed by many as an objective historical account of the Weyerhaeuser enterprises. But whi 1e the dates and other "facts 11 check out with other sources, there is a tendency to romanticize the role of individuals in shaping the political and economic climate of fimeri ca I s development. In describing Frederick Weyerhaeuser and his early partners, for example, Hidy et al. corment that "1ike other enter- prizing young men of the time--ColTTnsP. Huntington, Philip D. Armour, J.P. Morgan, and John D. Rockefeller--they relied upon their own energy, insight, and resourcefulness, and upon the inevitable development of the country" (Hidy et al., 1963:10). There is also some evidence that during the 1940 1 s the Weyerhaeusers were searching for some reputable scholar to write a history of Frederick \