ABSTRACT THE REPUTATIONAL TECHNIQUE IN A CROSS-SOMMUNITY PERSPECTIVE: SELECTED PROBLEMS OF THEORY AND MEASUREMENT by Eugene Curtis Erickson For a number of years researchers in the social sciences have been involved in research on a phenomenon called "community power. " A variety of methods have been employed to study the phenomenon. One of these methods is commonly called the “reputational technique. " It has been suggested that this technique is actually a measure of " However certain theorists have 7 phenomenon other than ”power. suggested that correlations between ”power" and other social phenomena might be expected. Little has been done by either proponents or critics of the technique to specify a more general basis upon which the results of an index of influence might be evaluated. It is this background which led to the specification of the problem, viz., to determine the relation between high evaluation of influence and other evaluative foci or indices of these foci. In this thesis the propositions which are derived from social system theory are related to empirical generalizations drawn from previous research. From these results hypotheses are developed to be tested on "samples" of influentials selected from several community settings. Besides relating the evaluation of influence to a theoretical perSpective, the reputational technique itself is tested for reliability. It is argued that positions in class, power, and prestige hierarchies Abstract Eugene Curtis Erickson [v contribute to the potential for interpersonal influence. lVlore precisely, ' the ”contents' of evaluations relevant to a profile of influence potential embody such general phenomena as social class and prestige, a "history" of power, social background, certain personal characteristics, and the personal control of relevant "objects" which give the influencer an advantage, such as a backlog of skill. At least the broad expectation of a profile of influence potential is derived from theory and many of the specific empirical components are specified. The components of the profile subjected to test are education. occupation, sex, age, marital status. father‘s occupation, religious affiliation, ethnic background, birthplace of influential and length of residence in community, orientation to community, and length of tenure in firm of employment. Hypotheses derived relative to each of these variables are tested against two "universes. The first universe is that ditribution of a specified characteristic among the population of the community from 'intra-community" cas e . V which a group of influentials are draxyn--the. The second universe is the total group of influenctials from whom the specific empirical generalizations were derived—-the "intra—influential" case. Six community influence structures were assessed on the basis of the methodological technique called the reputational technique. Four were communities in the Southwestern United States and two were border communities in Mexico. The six communities varied widely in size, ethnic composition, and industrial base. "Knowledgeables",were selected in each community who in turn selected the basic list of influentials. For each influential interviewed the above listed character- istics were determined. The results were submitted to statistical test for evaluation . Abstract Eugene Curtis Erickson In the intra-comrnunity case the influentials were found to be sharply distinguished from the general community on the variables of education, occupation, sex, age, and marital status for each of the six communities. For the four communities in the United States, two had groups of influentials who were drawn disproportionally from among the Anglo population of the community. For the intra-influential case the specifications drawn from other studies of influentials yielded the following results on comparison with the influentials. Except where noted the results hold for all six com- munities. Education: More had sonie college Sex: More were male Age: More were over 50 years of age Marital Status: Father' 8 occupation: Place of Birth: Length of Residence in community: Orientation to community: Tenure in "firm": Occupation: Religious affiliation: Ethnic bac kg round: (except for one cmnmunity) More were married More were sons of fathers whose occu- pations were white collor or above Varied by community (as expected) More had lived in community over 25 years More had local orientations (except for one community) More had over 2.0 years tenure in firm (except for two communities) Business occupations were dominant In the United States--Protestants were dominant In Mexico--Roman Catholics were dominant (certain reservations are necessary) Anglos dominanted. The reputational technique is held to be an efficient and reliable indicato r of influence . THE REPUTATIONAL TECHNIQUE IN A CROSS-COMMUNIT Y PERSPECTIVE: SELECT ED PROBLEMS OF THEORY AND MEASUREMENT By Eugene Curtis Erickson A THESIS Submitted to klichigan State University in partial fulfillment of the requirements for the degree of DOCT OR OF PHILOSOPH Y Department of Sociology and Anthropology 1961 .\ ‘) ACKNOWLEDGMENTS No statement of indebtedness can adequately fulfill the obligations entailed in the writing of a thesis. The training, learning, and knowledge gained from this experience are generalized and difficult to specify. In general. the support given by the United States Public Health Service and the Carnegie Corporation is most Sincerely appreciated. Specifically, for the direction of these projects and the format of the thesis itself, my major advisor, Dr. Charles P. Loomis, was inspiring. For the more detailed experiences which confront the "learner, " Dr. Archibald O. Haller spent many hours of probing at approaches and ideas which in the -context of professional criticism was appreciated. Also. I should acknowledge Drs. Beegle and Hoffer for their committee work. For numerous other persons it is difficult to itemize contributions. But contributions there were‘. Naming George and Carole Wolff for their hours of editing, Verl Franz for his not-to—gentle needling and he and his wife Margaret for proofing, David Lewis and Edward Pryor, Jr. for Being, succeeds in the specification of but a few. To Robert and Margaret Holloway I extend appreciation for a relationship categorized by Sincereity, for a relationship which had immeasurable effect upon this production. To Ann Marie and Kenyon, Stephen, and Jeffrey I formally extend the kind of informal Thank You which is appropriate only in the context of close family living. I I I , '''''''''''''''''' ii TA BLE OF CONT ENTS CHAPTER I. INTRODUCTION . A. Introduction to the Problem B. The Problem . . . . . . C. Importance of the Problem . II. SOME EXPECTATIONS FROM A THEORY OF INFLUENCE POTENTIAL . A. Introduction 1. Objectives of Chapter. . . . . . . Z. A Note on the Concepts Power, Influence, and Authority 3. Conceptual Framework . B.Discussion. C. Propositions on the Components of Hierarchies of Evaluative Phenomena Correlated with Influence . III. RESEARCH METHODOLOGY ON COMMUNITY POWER STRUCTURES. A. Introduction . . B. Three Methods in Perspective. . . . . . 1. The Reputational Technique for Determin- ation of Influence . Z. The Positional Method for the Determin- ation of Influence 3. The ”Decision" Method for the mDetermiIi- ation of Influence . C. Criticisms of the Reputational Technique . 1. Alternative Phenomena. a.Scope. b "Status" elites and influence. C. "Formal leadership" and influence d. Selected ”personality traits“ and influence. iii we» 11 ll 12 12 13 16 l7 l7 18 19 20 20 TABLE OF CONTENTS - Continued CHAPTER Page 2.TheFactorofTime............ 21 a. Longevity Question . . . . . . . . . 21 b. The ”recency” question. . . . . . 22 D. Summary of Hypotheses and Interrelation with Propositions . . . . . . . . . . . . . . . . . 22 IV. PROFILES OF INFLUENCE POTENTIAL: THE EMPIRICAL GENERALIZATIONS FROM RESEARCH LITERATURE . . . . . . . . . . . . . 26 A. Introduction . . . . 26 B. Elements of the Profile of Influence Potential . 26 1. Introduction: The Classification Model. . 26 2. Social Level: Criteria of Social Class Membership . . . . . . . . . . . . . . 28 a. Education . . . . . . . . . . . . . . 2 b. Occupation . . . . . . . . . . . . 2 3. Division of Labor Variable . . . . . . . . 32 a. Sex characteristics . . . . . . . . . 32 4. Life Cycle: Stages of Social Development 33 a. Age . . . . . . . . . . . . . . . 33 b. Marital status . . . . . . . . . . . 34 ‘5. Family Social Level . . . . . . . . . . . 34 a. Father' 5 occupation . . . . . . . . 34 6. Ethnic Status . . . . . . . . . . . . 36 a. Religious affiliation . . . . . . . . . 36 b. Ethnic background. . . . . . . . . . 39 7. Localism . . . . . ~10 a. Place of birth and length of resi— dence in community . . . . . . . 41 b. Orientation to the community . . . . 42 c. Number of years with firm . . . . . 44 C. Interrelation of Empirical Generalizations with Propositions from Section II.E. . . . . . . . 45 V. RESEARCH METHODOLOGY . . . . . . . . . . . . . . 50 A. Introductiovr- . . . . ‘. . . . . . . . . . . . . . . 50 B. Research Sites . . . . . . . . . . . . . . . . 50 C. The Reputational Techniques in Process . . . . 52 l. Knowledgeables . . . . . . . . . . . . . . 52 2. General Influence . . . . . . . . . . . . . 53 iv TABLE OF CONTENTS - Continued CHAPTER Page 3. Limited Influence . . . . . . . . . . . . . 55 4. The Time Dimension . . . . . . . . . . . 55 D. Research Techniques . . . . . . . . . . . . . . 56 E. Analytical Tools . . . . . . . . . . . . . . . . 56 VI. FINDINGS ON SPECIFIC CRITICISMS OF RESEARCH METHODOLOGY OF REPUTATIONAL TECHNIQUE 58 A. Introduction . . . . . . . . . . . . . . . . . 58 B. The H, potheses . . . . . . . . . 58 l. The "Limited" Scope Question . . . . . . 58 2. The "Experience" Question . . . . . . . . 59 3. The Longevity Question . . . . . . . . . . 64 C. Chapter Summary . . . . . . . . . . . . . . . . 67 VII. FINDINGS ON THE ELEMENTS OF THE PROFILE OF INFLUENCE POTENTIAL . . . . . . . . . . . 69 A. Introduction . . . . 69 B. Hypotheses on the Elements of the Profile of Influence Potential . . . . . . . . . . . . . . 69 1. Social Level. . . . . . . . . . . . . . . . 69 a. Education . . . . . . . . . . . . . . 69 b. Occupation . . . . . . . . . . . . . 73 2. Division of Labor . . . . . . . . . . . . . 77 a. Sex . . . . . . . . . . . . . . . . . 77 3.LifeCycle................ 77 a.Age.... 77 b. Marital status . . . . . . . . . . . 82 4. Family Social Level '. . . . . . . . . . . 82 a. Father's occupation. . . . . . . . . 82 5. Ethnic Status . . . . . . . . . . . . . . 84 a. Religious affiliation. . . . . . . . . 8-1 b. Ethnic background. . . . . . . . . . 91 6. Localism . . . . . . . . . . . . . . 94 a. Birthplace and length of residence in the community . . . . . . . . 94 b. Orientation toward the community . 96 c. Tenure in firm . . . . . . . . . . . 99 C. Summary of Chapter Findings . . . . . . . . . 99 1. The Intra—community Problems . . . . . 102 2. The Intra-influential Problem . . . . . . 103 TABLE OF CONTENTS - Continued CHAPTER Page VIII. SUMMARY AND CONCLUSIONS . . . . . . . . . . . . 105 A. The Propositions Derived from Theory . . . . 105 B. The Criticism of the Methodology . . . . . . . 106 C. The Elements of a Potential for Influence Complex . . . . . . . . . . . . . . . . . . . 109 REFERENCES..........................114 APPENDICES..........................120 vi LIST OF TABLES TABLE 1. \l The Number of Persons with High and Low General or Limited Influence Where Marginals for High and Fixed at Ten for Each of Six Communities. . The Number of A11 Influentials in Executive or Directive Positions, The Number Not in Such Positions and in Agriculture for Each of Six Communities . The Number of Persons with High and Low Influence on the Combined Business—Political Index of 1954-1955 and the General Influence Index of 1958 for El Paso . . Level of Educational Attainment of Top Influentials and Years of School Completed by Persons 25 Years Old and Over Dichotomized Near the Median of School Years Completed for Each of Six Communities. . Number of Influentials With One Year of College Edu- cation or More and With a High School Education or Less for Each of Six Communities . . Number of Influentials and Total Male Experienced Civilian Labor Force by Certain Occupational Cate— gories for All United States Communities and by Total Woik Force by Certain Principal Occupational'Cate- gories for all Mexican Municipios. . The Number of All Influentials Who were Employed in the Occupational Categories of Business, Professions, Government, and Miscellaneous, for Each of Six Communities . . Sex of A11 Persons Who Received at Least One Vote as a Key Influential by All Influentials Interviewed and of All Persons 25 Years of Age and Older in the Commun- ity for Each of Six Communities . . Age of Influentials and Age of All Persons 25 Years of Age and Over Dichotomized Near the Median Age for Each Community. vii Page 60 62 67 7O 74 78 80 TABLE OF CONTENTS — C ntinued TABLE 10. 11. 13. 14. 16. 17. 18. 19. Page Number of Influentials in Various Age Groups for Dan Diego, El Paso, Tucson, Las Cruces, Juarez, andTijuana........................ 81 Number of Influentials Who are Married and Single and Number of 1\«1a1e Persons in Total Community Popula- tion Over 14 Years of Age Who are Married and Single for Each of Six Communities. . Number of Influentials Whose Fathers' Occupations were White Collar, Non-White Collar, or Farm and Number of Persons in the Male Experienced Civilian Labor Force in White Collar, Non-White Collar or Farm Occupations in 1950 for Each of Six Communities 85 Number of Influentials by Various Religious Affiliation Classifications and by No Affiliation for Four Commun- ities in the United States . . . . . . . . . . . . . . . . . 87 Number of Influentials by Various Religious Affiliation Classifications and by No Affiliation for Two Commun- ities in Mexico. . . . . . . . . . . . . . . . . . . . . . . 90 . Ethnicity of Influentials and Number of All Persons With Spanish Surnames 25 Years of Age and Over for Each United States Community: San Diego, El. Paso, Tucson, and Las Cruces . . . . . . . . . . . . . . . . . 93 Place of Birth of Influentials in Each of Six Com- munities..........................95 Number of Influentials Who Have had Residence in the Community for Various Numbers of Years by San Diego, El Paso, Tucson, Las Cruces, Cd. Juarez, andTijuana........................97 Index of Orientation to Community Based on Influentials Willingness to Leave the Community for Each of Six Communities . . . . . . . . . . . . . . . . . . . . . . . 98 Number of Years Working with Present Firm by Influ- entials for Each of Six Communities . . . . . . . . . . 100 viii CHAPTER I INTRODUCTION A. Introduction to the Problem For a number of years researchers in the social sciences have been very much concerned with a phenomenon called "community power. " A variety of methods have been employed to study the phenomenon. One of these methods is commonly called the "reputational technique. " It is a technique which has been the subject of specific criticism by a number of professionals. The reputational technique has generally found a relatively small group of persons within American communities to be "power wielders. " This group of persons may not have been elected to office; they are not subject to recall elections and only in indirect ways to the sanctions of a public. Their actions in decision-niaking processes may further only those public goals which are in congruence with the "private" goals of the group’s members. Generalizations such as these strike at the heart of theories of the ideal democratic society. For whatever reason certain professionals have. purported that the methodological technique upon which these generalizations or research results are based is actually a measure of a phenomenon other than "power. ” Some of the alternative phenomena which the technique is purported to measure are "limited scopes" of influence rather than ”general" influence, a prestige dimension of the persons classified as influentials. the formal leadership of the community, or certain ”personality traits" highly conspicuous in the community. Little has been done by either proponents or critics of the tech- nique to Specify a theory with which the results can be related. Are the critics of the findings of the reputational technique correct in their implication that persons evaluated as possessors of the characteristic ”influence" will not also possess other characteristics such as those mentioned above? Sincehinfluence" is assigned to a given individual as a result of a process of evaluation of some content of action (infra), are there other evaluative foci which might be correlated with a high evalu- ation of influence? B. The Problem The specific problem to which this thesis is directed will be to attempt to derive from social system theory, in however limited a fashion, the relation between the high evaluation of influence and other evaluative foci or indices of these foci. The propositions which contain these relations will then be related to empirical generalizations derived from previous research so that hypotheses may be developed which can be tested on "samples" of influentials selected from several community settings. A corollary problem will be a test, insofar as is possible with available data, the reliability of the methodological approach which is called the reputational technique in the context of the above findings. C. Importance of the Problem The Specific criticisms of the reputational technique should be placed in a theoretical perspective and in turn subjected to a test for the purposes of substantiation or rejection. The focus upon the "universal" correlates which may be. associated with influence is long overdue in the research area of community influence structures. The findings Should substantially increase knowledge in this area. CHAPTER II SOME EXPECTATIONS FROM A THEORY OF INFLUENCE POTENTIAL A. Introduction 1. Objectives of Chapter The objective of the discussion to follow will be to develop the conceptual framework of the study. In addition, it shall be our purpose to use this framework as a device to glean from the theoretical litera- ture certain propositions1 concerning the substantive subject under dis- cussion, namely, the social correlates of influence, or what will be called the components of influence potential. 2. A Note on the Concepts Power, Influence, and Authority. In a thesis concerned with the complex phenomena called "power, " H or "influence, it is appropriate to limit one's scope of interest to a workable Sphere. The conceptual difficulties with these terms have been widely explored (e.g., see Cartwright [ed. ], 1959: 186 for a collection H of definitions). Variously "power” and "influence are used as the generic term by different theorists. Parsons (1951: 95) and Loomis (1960: 20) use the term "power, " while Barber (1957: 232ff) uses ”influence" as the generic term. The matter is relative. Whichever is used, however, some degree of control or effect of an ego over the behavior of alter is implied. Various aspects of this control have been described. For example, the legitimacy of the assignment of 1A proposition will refer to a statement in which something is affirmed or denied of a subject. control over something to a status-role or the illigitimacy of the assumption of control by an ego may be used as a sub-class of the generic phenomena (e.g., Loomis, 9p. Si}: 20). Likewise the application or non-application of Sanctions as distinguishing the relationship between ego and alter has been used as a sub-classification device. Most discussions of power or influence have been concerned with the description of all social relations insofar as there are elements of super- and sub-ordination in them. For the purposes of this paper, the use of these terms will be much more restrictive than this. The sub- stantive context of the social relations involved in community power refers only to those social relations which comprise the action situation in which community-wide goals or objectives are being met. A definition of influence sufficiently applicable to the empirical phenomena of interest to us will be "the capacity of a system-unit (ego) to control other system- units' (alters) action" (i___bi_d.: 20). System-units in the context of this paper are the incumbents of status-roles involved or participating in social relationships the substantive context of which is a community—wide activity. Neither term, power or influence, will be distinguished from the other in this text.1 When the concept authority is used it will mean that the rights and obligations attached to a status-role are legitimate. A status-role properly classified as "influential" may or may not be authoritative. That is, the capacity to control with which the incumbent of a status-role is said to be endowed may or may not result from formal rights and obligations attached to that status-role. 3. Conceptual Framework. Social relations have been defined as "reciprocal activity or inter- action that is repeated and persists" wherein interaction is an event 1Form and Miller (1960: 434) also use the concepts in this way. 4a involving a plurality of actors, communication between actors, a time dimension (and thereby, a history), and an "observable" objective (Loomis, 3p. c__i_t_.: 2-3). Since social relationships are repetitive they can be recognized as social systems, or ”the patterned interaction [reciprocal activity] of members" (ibid.: 4). Social systems have actors including system- units. The member or system-unit has a position or status relative to the other members or System units of the social system, and the system- unit is expected to act in a predictable or repetitive way, and thus has a role (ibid.: 19). The assignment of a position or status implies some form of evaluation of an ego by an alter(s). The evaluation is the process of ascertaining the relative value or importance of the position. Thus an evaluation implies some standard against which the position is rated. One end product of such a process of evaluation is the assignment of a position of influence to a status-role. That is, the status is "rated" on an influence hierarchy. It is analytically possible to duscuss hier- archies--—or systems of things in graded order--of influence. Likewise, it is possible to conceive of hierarchies of esteem--in which the standard of valuation involves a generalized component of estimation, and of prestige--involving a standard which is commonly based on achievement. System-units may also be ”members' of larger aggregates of people wherein similar life-styles and the appropriate symbols associated therewith permit the assignment of class membership. Again, gradation or hierarchies are implied. Positions of authority--with their specified rights and obligations--may also be placed in graded order. Thus, an ego may be evaluated by the alters with whom he interacts on the basis of a variety of standards. It may be said that relative status is assessed on the basis of many possible standards which lead to hierarchies. Individual persons, however, occupy many statuses-~e. g., father, president of "x, " teacher, etc. --and many roles are enacted for each status--e. g., father enacts discipline, affection, to spouse, sons, and daughters. Merton uses the terms "status-set" and "role-set" to describe these pluralities in expectations and positions (1957: 368). Loomis, in turn, also groups status-sets and role—sets in order to discuss a given actor's relative standing or rank. In reality each actor has a "rank" within a system. Analytically, that rank can be subdivided into "singular" status-roles which like a generalized rank are the result of a process of evaluation. An evaluation which has been based upon standards or other expectancies of the system is a process whose object is ego or a group of egos. These are the minimal terms needed to discuss the problem to be developed in the following chapter. That problem is, having been given a process of evaluation the result of which is an assignment of relative influence to a particular status-role, are there any similarly graded statuses or roles which contribute to the individual's influence? That is, are there social elements which may be conceived of as enhancing or creating a potential for influence? This is not a study in influence per se (i. e., in the dynamics of influence). Likewise it is not a study in the variety of forms of influence. Rather, the Correlates of influence are our central focus. It will be necessary to establish an index or measure of influence. To adequately achieve an acceptable measure it will be necessary to discuss some general problems in the contents (i. e. , in the contents of the social relationships which involve influence) of the influence relationship. Once this is done, however, this aspect of influence will be dropped. B. Discussion Merton asserts that hierarchies resulting from the process of evaluation are not, necessarily, concerned with the same phenomena. Specifically, he suggests that hierarchies of class, prestige, esteem, and influence may each be very different (113131.: 419). Nevertheless, he goes on to say ". . . positions in the class, power, and prestige hierarchies contribute to the potential for interpersonal influence, but do not determine the extent to which influence actually occurs" (i__bi_t1. ). The variability in the bases of evaluations (or standards upon which the evaluation is based) is also stressed by Loomis. "Ranking, " he suggests, may be dependent, on the one hand, "upon particularistic characteristics”--that is, standards derived from the status of an object in the relational system--and, on the other hand, "upon universalistic considerations--especia11y technical competence or skill" (92. 5.13.“: 25). However, he adds that both of these dimensions of evaluation may be operating though one or the other may be dominant. Dominance of one dimension over another, as he points out, may vary from society to society and from one community to another. (Community is defined as "a collectivity encompassing a territorial unit within which members pursue most of their every day activities necessary in satisfying common needs"_i_b_i(_i.:118-119, fn. 1).1 These comparisons of the dimensions of ranking are relative and in units such as the community and society may be intertwined. 1For our purposes community will be operationally defined to in- clude the population within the boundaries of large urban centers as these boundaries are defined by census bureau. This means that census definitions of Standard Metropolitan Areas and of an Urban Place as of 1950 will be used for communities in the United States. For communities in Mexico, the boundaries will include the geo-political unit called the Hmunicipio. " ' Parsons confronts the problem of the variety of differential evalu- ations related to the evaluation of influence. To utilize his discussion, the definition of "power" which he uses should be presented. Power is defined as "the realistic capacity of the system-unit to actualize its 'interests' . . . within the context of the system-interaction and in this sense to exert influence on processes in the system" (1953: 95-96). "Interests, " as used above, are equated with goals, the prevention of interference, command of respect, control of possessions, etc.l Except for the contextual limitation of the social relation, the definition embodies the same problems as the one presented for this text. Since "power" is a "capacity"' it is present when certain factors are present (at least power potential is then maximized). These factors involve evaluation. Power, then, is maximized when various forms of evaluation are integrated. Let us present these briefly. First, there ”is the evaluation of the unit in the system according to value standards . . ." (ibid.: 95). "Value standards" refer to the standards of the common value system within the social system; and, they have as their referent those behavioral characteristics which are highly evaluated. The differential evaluations based upon this axis would be relatively diffuse. Prestige and class are examples of such diffuse evaluations. Second, there is "the degree to which the manner in which actors in the system permit deviance from these standards in performance " (i_b_i__d. ), Deviance refers to the extent to which ego is willing to exploit opportunities otherwise forbidden by the norms. Deviance, as Parsons notes, may increase ego's power because he is allowed to "get away with it. ' " Presumably, the relationship of deviance to power depends heavily upon a tradition which a particular ego has established in which 1See following discussion for some of the implications of this definition. he has proven his ability to meet his interests. Thus, if an individual has a tradition or history of using unusual techniques to achieve his objectives, this very fact of unpredictability in means with predict- ability in ends may enhance that individual's power in any given new situation. Third, there is ”the control of possessions" (ibid.) which refers to differentials in advantage. Possessions have a specific relationship to control. Both possessions and what possessions “control” have a system reference. In general possessions are valued objects, and, thus, obtain their "nature" from whatever relationship members of a system assign to them. To rephrase, possessions are objects (situational) over which an actor has control, or they are rights to use or of which to dispose. Possessions are of two orders: goal attainment processes or objects of gratification. As goal attainment processes they are called facilities and are the means by which an actor may achieve certain goals. As objects of gratification, they are rewards. Degrees of technical competence or skill are examples of goal attainment processes. Wealth may be a form of possessions'as a reward. The three factors just discussed each involve elements which are evaluated. In addition, the elements, e.g., skill, wealth, prestige, are components of a status-set (and, in turn, of role-sets). They may be treated as separate or individual evaluative foci. As such, hierarchies may be established wherein many status-roles (occupied by human incumbents) are arranged in graded order. To what extent can one expect these hierarchies to be correlated? As Parsons puts it, "The assumption is that these three sets of factors are interdependent and, hence, that 'position' with respect to any one of them will be correlated with position with respect to each of the others, but they will also be of some degree independent" (ibid.: 96). The argument should be taken one further step. That is, the operational (empirically observable) forms of the generalizations should be developed. It was suggested above that a characteristic (1. e. , result of an evaluation of ego) which implies a close relationship to broad "societal" value standards and which is highly evaluated is a phenomenon such as social class membership. Indices relevant to such a member- ship are education level, occupation, income, ethnic status, family social level, etc. Characteristics or elements which suggest high evaluation of skill and competence are those related to sex, occupational level, or indices of a stage of social development. These latter indices may be subdivided into factors such as age and marital status. Age is very likely a characteristic subject to universal differentiation. Certainly industrialized societies have supplemented age-sex criteria with other elements (see, Loomis, cap: (33.: 26), yet it seems reason- able to expect that age-sex considerations are still relevant to some degree. Persons who are the leader's in the community (by whatever index of leadership or influence one takes) can be assessed as to the character- istics they embody, and it follows that one would expect these character- istics to differentiate this group from the general population in that community. An analysis of this form should yield what Merton called "a profile of influence potential" (see Merton, c_>p_. cit. , and Freeman, e_t 31., 1960). C. Propositions on the Components of Hierarchies of Evaluative Phenomena Correlated with Influence From the above discussion certain propositions can be derived. From Merton's general statement a Specification of some of the evalu- ated phenomena may be made. He states, positions in class, power, and prestige hierarchies contribute to the potential for interpersonal influence (supra). 10 The more specific "content" of evaluations may be in part derived from Loomis‘ generalization, which is that ranking may be dependent upon particularistic characteristics: for example, characteristics inherent in the relational system such as prestige, which may be com- bined with or analytically separate from another extreme, that ranking may be dependent upon universalistic characteristics: for example, technical competence and skill. Parsons also states a specification of the conditions under which power is maximized. Power is maximized when there is (l) a high evaluation of the system-unit on the standards of the common value system: evaluations of_prestige and class, (2) a utilization bythe system-unit of a "history" of activities in which ends have been met: evaluations which imply a backlog of experience, 8. g., age, and (3) a. differential of advantage on the part of the system-unit with regard to the control of possessions, which may be (a) the control of facilities: for example, extraordinary skill or competence; or (b) t_l_1_e_ control of rewards: for example, extraordinary wealth. CHAPT ER III RESEARCH METHODOLOGY ON COMMUNITY POWER STRUCTURES It is now necessary to turn to the problems of the phenomena with which influence may be correlated. A yariety of techniques have been used for studying these problems, and many specific criticisms have been leveled against the technique employed in the present study. However, certain hypotheses may be deduced from these criticisms and these hypotheses may in turn be interrelated with the propositions discussed in Chapter II. A. Introduction Extensive research has been completed on decision-making or ”power structures" in the community. In fact Freeman e_t a}. cite a source which listed 599 such studies (Bell, e_t a_1_l. , 1960, mimeographed, cited in Freeman, e_3_t E_1_1. , 1960: 2). In addition, nearly all of this research has been reported in the last ten years. Over this short period a variety of methodological techniques has appeared for determining, operationally if not actually, the group of persons in each community who may be defined as "influential. ” The field is still at a point wherein much more knowledge is necessary regarding the nature of influence in order to fully specify the Operational devices which best measure the phenomenon. Some of the techniques which have been devised have been criticized ”in substance, ” but few studies have been presented which correlate information obtained from the use of a variety of methods. 11 12 The following section will briefly summarize some examples of the methodological techniques which have been used. The purpose of this presentation is to place within a total research framework one of these techniques which is the focus of this study. There are a variety of classifications which have been presented attempting to place these methodological techniques in perspective.1 The classifications are illustrative for the purposes of this paper; the one used by both Rossi and Freeman, St a_1_1.,wi11 be used. There are three categories into which these techniques have been subsumed: the reputational technique, the social participation technique, and the positional technique. Each of these will be presented in turn. B. Three Methods in Perspective 1. The Reputational Technique for Determination of Influence One group of techniques which is clearly separated from the remainder is that which has used the "reputational technique" for determining influential persons.‘2 The methodological specifications for this technique are, in brief, that a group of informant or knowledge- ables are polled and asked to name influential persons in the community; thus the name, ”reputation for influence. " Hunter‘s study of Regional City in 1953 is a near classic in this approach. From a total list of 175 names drawn from civic, govern- mental, business, and status leaders (1. e., reputed leaders) a six 1E. g., Peter Rossi, “A Theory of Community Power, “ 1960, mimeographed, and Freeman, et a1., op. cit.: 2-3). 2E. g., Specific research reports which have relied on this technique are Hunter, 1953; Schulze and Blumberg. 1957; Miller, 1958a and 1958b; Schulze, 1958, Barth and Abu-Laban, 1959; Agger, 1956; Agger and Ostrom. 1957, Goldrich, 1958, Klapp and Padgett, 1960, Hunter, Schaffer and Sheps, 1956; and Foskett and Hohle, 1957. 13 "judge" panel selected 40 persons as the power leaders. Of these 40 persons 23 ”represented" the occupations of business (commerce, finance, and industry); 5, "society and wealth"; 4, government; 6, the professions; and 2, labor unions, (op. 511': 11-13 and Appendix). Hunter also said these were the people who make the policy decisions. They were integrated, i. e. , acted in concert. They had a pattern of interlocking memberships in community organizations. In short, key decisions were easily made by this handful of persons, according to Hunter (ibid.: 104, 230, 232). Most of the other studies which have used the ”reputational technique" have not been quite as adamant as Hunter on the point that their "leaders" or influentials were an integrated unit. A great deal of emphasis has been placed upon the ”institutional structures" or broad occupational compositions of the persons selected as influentials.I 2. The Positional Method for the Determination of Influence A general method which "defines" influentials (operationally) on the basis of the positions which they occupy in the community structure can be called the positional method for determining influence. Beyond this broad definition, the method has taken a number of forms. ' When "positions” or statuses are defined in economic terms--such as top positions in industry, credit and other business institutions-~it takes a Marxian basis. 2 Stouffe-r, in contrast, used this approach to identify "community elites"—-with the inference that these are influentials--but he used the top political and'civic statuses in the community as the criterion of influential positions. 3 1Especially by Miller, 1958a and 1958b and Form and Miller, 1960. ZSee Schulze and Blumberg, op. cit.: 291, who argue that this is the uriderlying argument which both the Lynds, 1930 and Mills, 1956, follow. 3Stouffer, 1955 cited in Schulze and Blumberg, op. cit.: 291. 14 An example of the technique is Stouffer's selection of 14 public statuses: mayor, president of the Chamber of Commerce, county chairmen of the Republican and Democratic parties, commander of the largest American Legion post, regent of the DAR, president of the women's club, chairmen of the library and school boards, parent- teachers‘ association, the bar association, and the publisher of the largest locally owned newspaper. Schulze duplicated Stouffer's ”elites" in his study of a community-Cibola. Schulze also selected 17 persons as "economic dominants" of Cibola. These were designated as "top formal statuses in the major-local industrial and credit units. ”1 In addition, the more typical group of "knowledgeables" were used to select the individuals by the reputational technique. The knowledgeables used were the formal heads of voluntary associations. These two groups were asked to select, by the reputational technique questions, the "public leaders" or the "power elite, as defined by reputation” for Cibola. In addition, the selected public leaders themselves were asked similar questions. Schulze reports there was no significant difference between the selections of influentials by the economic dominants, the heads of volun- tary associations, and by the ”public leaders” or influentials themselves. All agreed that ”substantially the same set of persons" were ”most influential in the affairs of the community (i_b_i_d_.: 295). " However, when top positions were compared with this list of persons, in no case did more than four of the "positional leaders" correspond with the list 1The specific criterion used selected persons who were ”the heads of all industries employing 75 or more workers, the heads of all banks with-total assets in excess of one million dollars, and . . . persons who were members of the boards of directors of two or more of these industries and/or banks and who thus served in the formal 'interlocking' of the dominant economic units” (op. cit.: 292). 15 of influentials. This held true for the definition of positional leader given by Stouffer (see above) and for other criteria of a political-civic elite (ibid.: 293) . In summary, Schulze found a distinction between persons who occupy formal political and civic offices and those persons assessed as influential by the reputational technique (i_1_)_i_d_. ). Another illustration of this technique is the study of the community power of the city of New Haven (Polsby, 1959b). This study combined the positional method for determining influential persons with the ”issue- area" approach. The latter specifies the substantive issues upon which decisions are made and attempts to determine the persons who are active on these issues. Polsby reports that a "leadership pool“ was drawn from three issue—areas: public education, political nominations, and urban redevelopment. The Specific disadvantage of this technique is the number of persons who are eventually left in the leadership pool and whose actions must be accounted for. For the three issue-areas men- tioned there were 131, 497, and 428 persons involved in each respectively (i_biii.: 798). The positions which defined the leadership pool by issue-area were Public education--"n1embers of the Board of Education, the Superintendent of Schools and his professional staff, school principals, leaders of teachers' and custodians' unions, and PTA presidents. " Political nominations--"all political office holders in the city, both in government and in the party organizations, and delegates to the conventions nominating party candidates for city posts in the 1957 election. . . ” Urban Redevelopment--"the membership of the Mayor's Citizen's Action Committees, a bi-partisan appointive body of 428 citizens. . . " (ibid. ) . The criticism made by Freeman et al. of this technique is appropriate. They state that Hno matter how complex and formal the organization of a community, there are always significant informal personal associations 16 as well as formal ones—-these are ignored by the positional appr‘oach; . this approach allows the investigator to make too many arbitrary judgments in defining which positions are important in shaping the affairs of the community" (23. (_:_'i_t_.: 2-3). 3. The "Decision" Method for the Determination of Influence Different from the above approaches is the concentration on ”issues" or "problems. " However the persons are identified, only those who are "actually" involved in a decision are considered.1 Part of the difference between this technique and those mentioned above is its emphasis upon how decisions are carried out (e. g., see Rossi and Dentler, 1960). We have already cited Dahl's study of urban redevelopment in New Haven (Polsby, pp. cit.) as a study which utilized this "method. " The term "method" is somewhat misleading since the approach only directs the research to specific substantive problems in the influence relationship; the participants may be determined independently. 2‘ A study which takes an issue-oriented approach but which circum- vents the positional method's difficulties is that of Freeman, it (3.1 , (92' £13.). In this study a list of 39 "community-wide" issues were selected. From public documents the formal authorities responsible for making each decision were determined. Each person was interviewed and asked to list all issues in which he actively participated. In addition he was asked to name other individuals who had participated. When two persons had nominated an individual as a participant for the same issue but who 1Appropriately, Rossi calls this method "decision sociometry (1960, mimeo.: 6). 2.E. g., the Polsby report (op. cit.) uses a positional approach in which the positions are authorities for a given issue. 17 was not in the original group of authorities this person was then inter- viewed and asked the same questions. He was dubbed a first-level leader. The process was repeated so that nominations were received two steps removed from the formal authorities for the issue. Inter- views with 628 leaders were completed. The 39 issues were then factor analyzed to determine the patterns of participation in these community affairs. In addition, selected ”social characteristics" of the leaders were factor analyzed to determine the differences between the social characteristics of the leaders and non-leaders in the general community. The specific findings of this study will be related in a following section of which it will be the purpose to develop as many empirical generalizations based on research findings and variables relevant to the problem of this thesis as possible. C. Criticisms of the Reputational Technique Criticisms of the reputational technique, discussed above, have been centered around three questions. They are (1) Is the reputational influence index a measure of something other than generalized com- munity wide influence? (2) How does the dimension of time affect the evaluations of reputed influence? and (3) Are there “serious discrep- ancies between actual and reputed leadership ?" (Freeman, (it 211. , ibid.: 2). 1. Alte rnative Phenomena The alternative phenomena which the reputational technique is said to measure (as selected possibilities) are a series of "limited influence" measures (see below), a form of "status elite, " "formal leadership, " and certain types of personality traits. 18 a. Scope. - As was mentioned above, the general thesis which Polsby (as a single representative of a group of researchers) has developed, stresses the presence (hypothesized) of limited areas or sectors to which a given individual's influence is limited. His thesis does not deny the phenomenon of influence, for he states, "On any single issue, there are only a few influentials: one usually finds that for any particular sector of policy only a small number of persons ever initiate alternatives or veto the proposals of others" (Dahl, in D'Antonio and Erhlich eds.: 7). Thus they suggest an overall thesis that American communities are made up of a number of segmented "power structures, " and that a person's influence is limited to certain sectors (thus limited in m which are defined, presumably, by the "institutional”1 areas in which the individual has skill or with which he is identified. As suggested above, Dahl's thesis is most sharply contrary to that of the reputational researchers who have found that influentials tend to remain the same from issue to issue regardless of the “institutional sector" they represent (See, e. g., Polsby, 1959a: 232 and Wolfinger, 1960: 638). One of the most succinct statements on this problem is that of Wolfinger who states: The term scope is used to refer to those actions of B which are affected by A's exercise of power; for example, the major scope of a school superintendent's power is public education. . . . futhermore, the researcher cannot be sure that his resPondent is [when answering the question on general influence] tacitly 1A brief discussion of this concept is apprOpriate. The most fre- quent use of the concept "institution" has been by Form and Miller, 1960: 19; Miller, 1958a and b; Form and Sauer, 1960; Sauer, 1960; Form and D'Antonio, 1959. Form and Miller trace the concept of Lynd (1945) and Lynd and Lynd (1929 and 1937). Five "institutions" are con- sidered: economic, political, family, religion, and recreation. The operational use of the term has been in terms of occupations (e. g. , D'Antonio, Form, Loomis, and Erickson, 1960). It is in this occu- pational sense that the concept is used here. l9 basing his rankings of community leaders on an implicit scope, with the result that an individual may be given high general power rating because he is perceived to be very influential on a particular issue .which is either currently important to the com- munity or salient to the respondent (o_p. cit.: 638). In summary, the argument has been that the methodology used by the reputational researchers belies reality. (It has already been pointed out that the same argument has been used to discuss the limi- tations of the positional approach to influence which Dahl, Wolfinger and Polsby have used.) Our objective must be to present additional data which will substantiate or refute the general thesis presented here. Following the above argument the hypothesis may be derived: That a given individual's influence measured on a general influence index (by the reputational technique) will vary sharply over measures of influence on particular issue situations. b. ”Status" elites and influence. - Among the alternatives to a measure of influence which certain critics hold is the ”trueH phenomenon measured by the reputational technique is that which Polsby calls "status" elite (1959a: 232, fn. 5). Though the concept is not defined it seems safe to assume that the phenomenon it is intended to involve is high evaluation of social characteristics. Presumably the criterion of this would be elements of prestige and esteem. No argument is presented on the background for this proposition-- either theoretical or empirical. A following section will suggest an argument based on the more generalized interpersonal evaluations from sociological theory which leads us to conclude we should expect little difference in hierarchies of evaluation whose separate criterion is influence or social level. Again however, following the argument, a hypothesis may be derived: That a given individual's influence measured on a general influence index will correspond with his position on an index of "status. " 20 c. "Formal leadership" and influence. - Polsby also (ibid.) suggests the alternative that "the community's formal leadership" supplied the criterion for reputed influence indexes. Again an inference must be supplied as to the meaning of the concept. There are two interpretations which shall be presented. First, the formal leadership refers to these status-roles which are governmental or authoritative in function. Both elective and appointive offices may be involved. 1 A second alternative would involve the above positions and those heads of community committees such as United Fund, Library, etc. This latter case departs from the prior alternatives in that individual skill and identification are not limited to "institu- tional sectors" of interest. The following hypothesis may be derived from the argument: A given individual's influence measured on a general influence index will yield the community's ”formal leadership. " d. Selected "personality traits" as an alternative to influence. - Some very gross characteristics were also noted by Polsby in his criticism (ibid.). For example he lists "old civic war horses, " "letterhead names, “ and ”vocal leaders in the community" without specification as to the empirical phenomenon involved in the generalized characteristics. However, the following hypothesis may be derived: That a general influence index yields a group of persons who are "vocal leaders, " ”01d 'civic warhorses, "' or “letterhead names” in the c ommu nity . 1This criterion seems to be identical to that used by Dahl, Polsby and Wolfinger in their New Haven study (see Polsby, 1959b). This is an odd criticism then since it argues that the reputational technique yields the same measure of influence which the critics use. 21 2. The Factor of Time There are two dimensions to the question of time as a determin- ative factor in these problems presented by critics of the reputational technique. These are problems of "longevity" and “recency. ” a. The longevity question. - This question refers to the assess- ment of influence holders in a given community over time. Olmsted (1954) found substantial differences between panels of "knowledgeable citizens" named in 1943 and 1949. Wolfinger (3p. cit.: 644) seems to imply that a similar lack of persistence would be common under the methodological conditions of the reputational technique. He states: While some individuals might maintain some or all of their power after a change in regime [by election], others would not, and some relatively powerless persons would be placed high in the "power structure." The inclusion of all political actors within a supposed power elite would be neither surprising nor discriminating (ibid.). Wolfinger adds a strange note to this statement. Certainly it is reasonable to suggest that authority positions are seats of power. None of the reputational researchers have argued the negative of this. Thus whether the researcher expected a segmented power structure along the axis of limited scope or, in the extreme, a homogeneous power elite, “positions" of authority must in either case be considered. Election might therefore change the persons who occupy the status- roles but not the status-roles themselves. Even here the discussion is forced back to a definition of "political actors"--which Wolfinger does not supply and to the question of whether this implies ”persons',‘ or "status-roles. " This is a confusion which will be treated shortly. At any rate, a hypothesis on the problem of longevity can be derived from this discussion. Viz. , That the individual's named on a general influence index are so named because of their action on 22 recent issues which have confronted the community and that similar- ities over time would be a chance occurrence. b. The "recency" question. - There are two interrelated facets to this question. One has reference to the recency (in time before the interviews with the respondents occurred) with which specific issues have been before the community. Implied here is the idea that persons named as reputed influentials, since they are each "affiliated" with a single "issue sector, " will vary as issues vary. The second facet focuses on the respondent. The idea states-that whatever issues are particularly salient to the respondent, these issues will have a determinant effect upon his choice of "top community influentials. " The hypothesis derived from this argument is that the individuals named on a general influence index are so named because of certain issues which are especially salient to the respondent. D. Summary of Hypotheses and Interrelation with Propositions The task before us is that of interrelating the propositions con- cerning influence and the specific hypotheses presented in the fore- going section. The proposition has been presented that positions in class, power, and prestige hierarchies (as defined above) contribute to the potential for interpersonal influence. In addition it was pointed out that ranking may be based upon a variety of standards. It is argued, then, that influence may be ”attached" to a status-role (in reference to one position), such as the case involving authoritative status-roles. Or it is possible that the researcher is confronting ranking, involving 23 a complex of statuses and roles (status-sets and role-sets), some (or each) of which are highly evaluated and correlated with the evalu- ation of influence. Under these conditions the carry-over of influence from one substantive-issue situation to another will be expected. The reader is reminded that the "substantive-issue situation" has been restricted to community-wide issues and thereby is a restriction on the forms of influence under consideration. This penetration of influence may be due to two phenomena: the position which is involved in numer- ous issue situations due to its broadly authoritative nature (e. g. , a mayor), or the "composite" evaluation (ranking) of a combination of historically relevant statuses and roles which are attached to a given individual. It may be concluded that if there is a phenomena of general influence, its empirical form will not be different from that relative influence determined in a particular issue situation. If these argu- ments are not true and, if influence, by virtue of relative skills on particular issue situations, is limited to specific issue situations, the comparisons of individuals named on a "general" and "limited" issue influence index will show substantial divergences. Following the first alternative it is held that instead of the hypothesis derived from the arguments of the critics of the reputational approach, viz.: . [C. 1. a:0] That a given individual's influence measured on a general influence index will vary sharply over measures of influence on particular (and limited) issue situations, the converse may be stated for purposes of testing: [C. l. a: 1] That a given individual's influence measured on a general influence index will correlate with influence measured on a particular or limited scope issue situation. In addition, the exceptions to the hypothesis may be specified to be those positions which are influential by virtue of inherent authority apart from the individuals who occupy them. It also follows from the propositions stated that instead of offering the c riticism 24 [C. l.b:0] That a given individual's influence measured on a general influence index will correspond with an index of "status, " the hypothesis may be considered an expectation. This hypothesis will also be treated in more detail in Chapter IV. The argument has relevance for the hypotheses presented in this section on the longevity of persons involved in influential status-roles. Instead of the hypotheses [C. 2. a:0] That the individuals named on a general influence index are so named because of their action on recent issues which have confronted the community and that similarities over time would be a chance occurrence, and [C. 2. b:0] That the individuals named on a general influence index are so named because of certain issues which are especially salient to the respondent, a more general hypothesis may be derived. It would deny both of these as stated above; it would be [C. 2:1]That individuals evaluated as influential at one point in time will also be evaluated as influential at a later point in time. It may also be necessary to revise this hypothesis to account for authoritative status-roles. The problem of the extent of the "time" for which the statement will hold true is one which of necessity must be ignored. Data avail- able dictates that the time period to consider will be approximately four years. For this period it is felt, intuitively, that the statement holds. Obviously, there is a point at which only an ”overlap" might be expected. Two hypotheses presented in the foregoing section are conflict- ing. They were [C. 1. c:O]That a given individual's influence measured on a general influence index will have yielded a group of persons who correspond to the community's ”formal leadership. " (Where "formal leadership” is equated with governmental status-roles.) 25 and, [C. 1.d:0] That a general influence index yields a group of persons who are "vocal leaders, " "old 'civic warhorses, “' or ”letterhead names” in the community. (Where these selected characteristics are not synonymous with or indicies of governmental status-roles. ). To some extent the formal leadership of the large economic bureau- cracies may indeed be expected to be influential. These are positions which are to some degree indicies of skill and, as has been shown, skill may be a phenomenon correlated with influence. However, care must be taken since studies have shown that economic dominance is by no means a predictor of influence in community-wide activities (8. g. , Schulze, gp. c_:_i_t_. ). The general hypothesis may be posited that experiences derived from formal leadership may be an enhancement to community influence. Thus, defining "formal" leadership as a broad range of executive experience, the hypothesis may be restated [C. 1. cd: 1] That a given individual' 5 influence measured on a general influence index will yield a group of persons whose occupational. experience involves executive or directive responsibilities. CHAPTER IV PROFILES OF INFLUENCE POTENTIAL: THE EMPIRICAL GENERALIZATIONS FROM RESEARCH LITERATURE A . Introduction The purpose of this chapter will be to present a summary of the research literature on the basis of certain empirical geheralizations and to interrelate the hypotheses which result with the hypotheses developed in Chapter II. B. Elements of the Profile of Influence Potential 1. Introduction: The Classification Model Of all the research in‘the area of community power to date, Freeman _e_t a}. (i_b_i_d_.) present the most complete data on the inter- relations of the characteristics of influentials. Citation and descrip- tion of the study was given earlier. In addition it must be noted that 15 "social characteristics” were interrelated by the technique of factor analysis. As a result these 15 variables reduced to six major factors: social level, sex elaboration, life cycle, ethnic status, family's social level, and localism. In general, the argument of this section, D in Chapter II, would lead us to expect this very occurrence. The problem now becomes one of specifying what level of education, what occupation(s), and which sex, etc. , operates as the selecting criteria or potential for influence. 26 27 The remainder of this chapter will be addressed to a review of the research literature on community influentials in order to achieve the empirical generalizations of these characteristics. The chapter outline will follow the general scheme presented by Freeman, e_t 8:1. 1 That outline follows. Factor Social Characteristic A Social Level: criterion of social Class membership 1. Educational level of influential 2. Occupation of influential B Sex Elaboration: Division of Labor variables 3. Sex of influential C Life Cycle: Stages of Social Development 4. Age level 5. Marital Status D Family Social level 6. Father's Occupation E Ethnic Status 7. Religious affiliation 8. Ethnic Background F Localism *9. Birthplace of Influential and Length of Resident in Community =2=10. Orientation to Community *11. Length of tenure in Firm with which employed 'IiLength of residence in community, orientation to community, and length of tenure in the firm by which employed, were not included in the list of Freeman's, et a1. characteristics. The reasons for their inclusion hex; 55111 be developed under their respective headings. 1The following social characteristics were also included in Freeman's et a1. analysis: Factor A: Income level of influential, political ideh-tification, and father's educational level; Factor C: Home ownership; Factor D: father's community participation; Factor F: Father's Birthplace. The following characteristics were found inter- correlated under more than one factor. Listed under factors to which they are added: Factor B: Occupation of influential; D: Father's education; and E: Political identification. These items are omitted due to the post factum classification presented in this thesis. 28 2. Social Level: Criteria of Social Class Membership Two characteristics will be reviewed: the educational level of the influential and the influential's occupation. a. Education. - Freeman, St a_l. suggest a college degree is highly valued (or predominant) among community influentials (pp. g_i_1_:.: 16). Hunter reported that 92. 5% of the 40 influentials in his study cite-— Regional City--had a college degree (1953: 39). Sauer reported that 23 of 39 influentials (59%) had graduated from college while 31 (or 84%) had had some college education or more (_c_>_p_. cit.: 64). The studies by Stewart (1947a and b) and Merton (_o_p_. sit.) on interpersonal influence are difficult to interpret (.i_n_f_13). Merton found the following differences. "Localities"l had graduated from high school; eight of the 16 (50%); while ”cosmopolites" had a relatively higher edu- cational background, with 13 out of 14 graduating from high school (93%). Stewart found that of 55 mp influentials, 79% had graduated from college, 88% had had some college or more, and 94% had graduated from high school. As an additional complication, Stewart's sample was primarily "local" (13 persons), or "mixed" (16 persons). Only four were classi- fied as "cosmopolitan" (1947a: 23)."‘ 1The two classes ”localite" and "cosmopolitan" are variables which distinguish the influentials' orientation to the community. They will be discussed and defined in a later section on the subject. ZStewart's use of "local" and "cosmopolitan" follows Merton. He adds, however, a middle class which is termed "mixed. " A "mixed" orientation is held by "one who has interest extending well beyond the home community but who, as shown by participation in civic activities, is an integral and active member of the community (213. £113.. 23). " 29 The following empirical generalizations may be derived. [IV.a:l] That among a given group of persons who are classified as influential in a community, the majority of them will have had at least a year or more of college education. And, [IV.a:2] That a group of influentials will have significantly more education than would have been the case if the group were drawn at random from the population of the c ommunity. (That latter hypothesis attests to the selection of these persons as influentials.) b. Occupation. - Many studies report incomplete or very generalized information on the characteristics of the persons involved in the influential relationship. Though this dictum applies to many specific characteristics which will be discussed, it is especially a problem with the occupational position variable. Freeman, e_t a_l. , in Syracuse, found that among the "leaders" in the community those occupational roles which represented "white collar or above" occupations were "more valued" than other occu- pations (_o_p. cit.: 16). In this very general form, no research study has carried contrary findings. Enumerations of specific categories of occupations are not common. Fourteen of Merton's 16 "local" influentials and nine of 16 "cosmopolitan" influentials were either businessmen or professional (93. c_i_t_.: 502). Likewise, none of 11 top leaders in the Boomtown site reported by Agger and Goldrich were businessmen, while one was the mayor and the last a school board member who was also a brother of the mayor (1958: 386). Schulze and Blumberg report that 11 of Cibola's 16 "public leaders" were businessmen (industrial executives, merchants, professional Chamber of Commerce officials, wholesaler, realtor, salesmen) while four were professionals and one retired (Oi. cit.: 296, fn. 20). 30 Stewart used a number of questions each of which were used to tap "an influence" dimension.1 On the dimension selected for com- parison, he reported a total of 17 persons.2 Of this total two were in government, eight were in business, and seven (of which three were "wives of") were in professions. 1The questions for which Stewart reports detailed information on occupation are (1) When people you know have some personal trouble, to whom are they most likely to go for advice? (1947b: 273). (2) Who do you think would best be able to get people to cooperate in a war bond drive in civilian defense, etc.: (1947a: 26). (3) Who would you say are the important people in town? (1947b: 273). The present writer computed a rank order correlation to test whether the persons named to questions 2 and 3 above would be ranked with relatively the same magnitude to each variable. The result was an rs = . 33--A Spearman Rank Order Correlation,--which indicates little correspondence in the rankings. However, when gross groupings of Stewart's lists were compared, the relationship between persons named on these two questions were much closer. For example, six of the top ten on both lists are the same persons. These three questions illustrate the problem of assessing the nature of influence relationships. The phenomenon of extraneous factors slipping into the general evaluation of influence is illustrated by the first question listed above. In this question, Stewart reports that it is the clergy to whom people go for advice on personal difficul- ties. The meaning of this particular relationship is so ambiguous it is difficult to place it within the framework of influence. Because of this ambiguity question one is omitted from further consideration. Though the gross comparison between lists of persons named in response to questions 2 and 3 indicated significant overlap, it can be argued that question 3--important people in town--is closer to the index of influence used in the study to be reported in this thesis than question 2. Therefore, results of that question will be detailed in the text. 2Of the 17 persons named on the question, 13 were males. In the case of women named, they are classified as to the occupation of their husband (therefore, are "wife of" a man in . . .) with one exception-- a woman executive in a business. .- 31 Miller has compared the occupational background of the influ- entials in two cities which he studied with results reported by Hunter (1958b). The same data was later compared with eight other com- munities with the result that in all United States communities similar occupational groups were found to dominate the influential structure (D'Antonio, et a1. ). Persons who dominate the community power structure generally “represent" the business institution of the com- munity. Likewise, "high" government officials and occasional pro- fessionals are represented, though with less frequency than business. There are at least two reasons why one should be cautious in making generalizations about the dominance by the business and pro- fessional persons. First, business dominance has varied among communities. For example, Miller has shown that one English City manifested a high dominance of labor leaders (the only case of such representation reported in the literature) (1958b). A Mexican com- munity was reported to have a high representation from the sphere of government (Form and D'Antonio, o_p_. 213). We conclude that there must be confounding circumstances which lead to the production of influentials in particular communities. Second, despite the fact that business institutions may furnish most of the persons who are reputed to be influentials, not all business- men have an equal chance of becoming influential. Thus, other characteristics must also be assessed. It is recognized that categories such as "business, " " govern- ment, " and "professions" may involve individuals whose identity to one or the other of these categories is obscure: for example, a "businessman turned mayor for a term, " or a "corporation lawyer. " Though chances of ambiguous classifications exist, it is tentatively held that experiences of persons who have been educated in a profession , turned from a business to politics, etc. will be different enough to warrant their separate classification in these studies. 32 In short, the literature leads us to the generalization [IV. bzl] That for communities in the United States, the dominant occupational background of influentials will be "business" occupations, followed by government positions and the professions. For Mexico, on the other hand, we may only tentatively generalize [IV.b:2] That for communities in Mexico the dominant occupational background of influentials will be both business and governmental status-roles. 3. Division of Labor Variable Two variables are involved in the discussion of this section: sex and occupational prestige. The second of these, occupation, has already been discussed. a. Sex Characteristics. - From all the specific studies which have been completed on influence structures one generalization seems secure. It is that males dominate the structure. We have already noted that Stewart found on his index of influ- entials that 13 of 17 named were males. He also suggests that over all dimensions of influence on which he had information he found "male dominance, even when women were expressly asked for (1947b: 273). " Sauer found only males in positions of influence in the com- munity of Wheelsburg (pp. gi_t.: 55-56). Freeman c_i_t a_l. , found general male preference in Syracuse (3p. c_:_it.: 16), and Hunter lists but one female among 40 males in Regional City (92. c_it_.: 38ff). It should be noted that a host of additional studies do not mention the sex characteristic of influentials. However, it seems safe to assume from the context of their reports that the groups may have been composed entirely of males. Occupational characteristics and the explicit neglect of the characteristic of fenialeness are the basis of .‘I'k’ '; 33 this generalization. Studies on which it is based are Schulze and Blumberg (2p. cit.), Smith (1960), Fanelli (1955, 56), Agger and Goldrich (2p. git), Agger (<_)_p_. c_i_t.), and Miller (1958a and b).1 The generalization may be deve10ped [IV. czl] That a sig- nificant proportion (a (majority) of a‘given group of influentials will be male. 4. Life Cycle: Stages of Social Development Two Characteristics will be discussed under the above topic: age and marital status. a. Age. - Most studies have included information on age character- istics of the influentials. However, the data are generally given in the form of gross findings such as range, median age, or mean. Freeman, e_t a_l. , (22. <_:it.: 16) found an age range from 35 to 64 years was "most valued" among their group of influentials in Syracuse. Stewart reported a median age of 48. 7 years for all of his influentials (e_p. C.i_t.). Hunter notes an "average" age of 50 years (pp. cit.: 38ff). Schulze and Blumberg found a median age of 53 years (3p. cit.: 294), while Sauer reports a median age of 57 years for the influentials in Wheels- burg (o__p. _C_i_t.: 64). Merton found that in dividing his sample of influentials by "local" and"cosmopolitan" criteria, the former was "older, " and the latter 1One study is a “companion" study to that of Sauer (p_p_. 23") It concentrated on the women who were active both socially-~devoting their time to a few select associations with their peers--and civically—- participating in civic, social, and religious organizations. Wolff found support for a hypothesis that "women whose husbands are either economic dominants or top influentials or whose family is in the top twenty socially choose social means . . . to express their social position; whereas, women whose husbands have not achieved community prom- inence [as any of the above specifications] . . . Choose civic means . . . of expression (1961: 71)." 34 were "younger" (32' _c_it_.: 395), However, when the variable "length of residence in the community" was considered, the older—local group had lived in the community longer than the older cosmopolitan group. DeSpite Merton's findings, it seems sound to argue that age is itself a selecting criterion. The experiences and "history of effective- ness" which can only be attached to individuals presuppose an element of time. Time, in turn, presupposes aging. From the above studies the following generalization was developed: [IV. d: 1] That the mean age of a given group of influentials will be over 50 years of age. In addition, it may be suggested [IV.dz2] That a given group of influentials will differ significantly in their age Characteristics from the general population of the community from which they come. b. Marital Status. - Only one study to this writer's knowledge gives data on the marital characteristics of the group of influentials which were described. The study is the Syracuse study by Freeman, e_t £11. , in which it is reported that the characteristic "married" is preferred in the community. Influentials in Syracuse are more likely to be married than unmarried. Basing an empirical generalization on this one study it may be stated [IV. e: 1] That a majority of the influentials will be married iather than single in marital status. Also, [IV.ez2] That a given group of influentials will differ significantly in their marital status (iiore will be married) from the-general population of the community 30m which they come. 5. Family Social Level a. Father‘s Occupation. - Occupational mobility deserves special consideration. Schulze and Smith have both reported, in studies of two 35 "1 are no widely separated communities, that "economic dominants longer evaluated as part of a community's influence structure. Indeed, "economic dominants" are largely a group distinct from the group of influentials. It might be assumed that economic dominants consist primarily of persons who have inherited wealth, since they are defined, at least in part, by their unusual control of property. Contrary to the Schulze and Smith thesis is that of Hunter, who reports that wealth and influ- ence are largely coterminous. Apparently 14 out of 40 influential persons in Southern City had inherited a business or a "leadership position" (_o_p_. (21.: 29). Against this economic foundation of influence is the thesis that influence is founded in a broad class hierarchy. Thus, Freeman, e_t ail. (tip. C__it.:16) report that most of their respondents' fathers were members of white-collar occupations. Schulze and Smith both con- clude, on the basis of historical evidence, that the economic foundation of influence has been decreasing in recent years. However, this is most directly measured in terms of official governmental positions which economic dominants are reported to have held in the past. 2 Certainly these data suggest the following generalization. [IV.f: 1] That a group of influentials in a community will be drawn 30m a population whose fathers were engaged in white-collar or highe r occupations . m 1See footnote page 14 for description. ZQualification may be necessary on this and other variables when the conditions of community variability are considered. For example, does "community age" have a determining effect on the type of father's occupations which predominate for a given group of influentials? Though recognized, these possibilities are not properly considered here. 36 6 . Ethnic Status Two characteristics of ethnic status are to be considered: religious affiliation and ethnic background. a. Religious affiliation. - Two studies have reported the dis- tribution of influentials among the various religious affiliations. Freeman, e_t a_._l. report that "Protestantism" is highly valued in Syracuse (2p. <_:_i_t.: 16). Smith reports that 14 of 16 t0p influentials in the community of Northville were of the Mormon religion.1 Religion as a social phenomenon may be viewed in quite diverse ways when the implications of its linkage with the influential structure are considered. First, it can be considered as indicative of a sincere value or belief commitment on the part of its adherents. Therefore, if religious affiliation is a community-centered phenomenon--as Smith suggests is the case in Northville--commitment by influentials to a specified belief system is a community-oriented phenomenon. Smith writes, "The integration of these influentials . . . appears to depend much more on the kind of stake they have in the community . . . [with] Mormon religious sentiment . . . relatively strong in the community . . . there exists at present a relatively high degree of residence commitment to the community . . . " (iliitla 88). The nature of the pattern differentiation to which this order of the phenomenon refers is "internal. " Here the dominance is the local belief system. A pattern such as this might be expected in those communities which have a largely homogeneous religious foundation. The expectation on the level of the influentials comes not so much from the relative frequency with which a given faith is held in the 1The community of Northville which Smith studied is reported to be highly dominated by Mormons. This affiliation made up approxi- mately 80% of the 18, 000 inhabitants in 1959. community as it does from the adherence by these influentials to the dominant value pattern in the community. Presumably, insofar as religion is concerned, this might also vary as the nature of this religious commitment varies. For example, since the Mormon religious sect might be described as an autocratic, dogmatic religion, the sect probably requires more of a personal commitment than some other religious affiliations. A second view of the meaning of religious affiliation in the con- text of community influence is quite distinct from the first. This view suggests that the pattern differentiation for religious affiliation is primarily an "external" affair. That is, that the dominant value pat- terns in the society are determinants of intracommunity religious affiliation. Much research suggests a dominance of Protestantism, which is only one characteristic in a syndrome (e. g. , Warner and Srole, 195 ). Presumably, as in the case of Syracuse, the dominance of Protestantism reflects this broader, more diffuse characteristic. Church affiliation can also be considered completely apart from the internal-external value components. The View would bring into focus an interactional'point of view. Unfortunately, this writer knows of no study of influentials which reports their membership in a particu- lar church community. (This is more specific than a particular denomination since in large urban centers, at least, many denomina- tions have more than one church community within their boundaries.) It is true, of course, that even from an interactional point of view homogeneity in belief commitment is possible. However, the "content" of an expectation on this level would, in all likelihood, be quite dif- ferent when built upon the basis of sheer interaction as against a commit- ment to a religious ideal. We have no reason to believe any of the four United States' com- munities under study represent homogeneous religious commitment 38 such as Smith reports in Northville. Thus, we should hypothesize on the basis of this that there will be differential affiliation in the various denominations. Reasoning from the Syracuse study, it is possible to expect that the predominant type of religious affiliation would be Protestant. Some Southwestern communities in the United States have at least a majority of their population affiliated with Roman Catholic churches. This fact is due to the heavy concentration of persons of Mexican heritage. If the pattern which we suggest does not hold up, it will surely be on this ground that the explanation of the exceptions must be made. The situation in Mexico offers a different problem. Here Roman Catholicism is the dominant religious affiliation. However, whether or not it represents a dominant commitment to a religious ideal or to a dominant societal value pattern is another matter. The history of Mexico suggests a segmentation of religious and political commitment (Davis, 1958: 16 and 238). This relationship between church and state and between the church and the individual (on the level of personal commitment to a belief system) is one that deserves a complete study in itself. In general it would seem unwarranted to expect religion to play more than a nominal .role in the context of community decision- making. Exactly what form this would take in the context of a question "what is your religious affiliation?" is difficult to predict. In extreme cases, the question would probably result in an answer indicating no commitment whatsoever. Precisely what prOportion of persons would represent this situation is almost impossible to tell. As a contrasting development it must also be mentioned that the Northern states of Mexico are among the major centers of a growing minority party in Mexico. This is the Partido Accion Nacional (PAN), 39 the Party of National Action. The majority party, Partido Revolucionario Institucional (PRI), has had a traditional anti-clerical constitution. PAN, on the other hand, has developed a firm alignment with the Church-- Roman Catholic. Thus, the "content" of an individual response to a question on religion might be very different in Mexico. Nevertheless, if respond he does, it would probably be to indicate Roman Catholic affiliation. Generalizations developed in this section are [IV. g: 1] That a group of influentials in a United States community will have religious affiliations characteristic of those highly evaluated in the general society, viz. , protestant, regardless of proportions of non-protestant affiliations in the community. And, [IV. g:2] That a group of influ- entials in a Mexican community will have religious affiliations characteristic of the dominant religion in Mexico, viz. , Roman C atholicism . b. Ethnic background. - Few studies have been aimed at the relationship of ethnic identities to the compositions of influence structures. Freeman e_t a_1. report that Syracuse influentials were predominantly North-western EurOpean in national background (32' cit.: 16). They do not state the national backgrounds or minorities which are non-North-western European. Hunter found no Negroes among his top influentials in Southern City (93. cit.). Likewise, if Barth and Abu-Laban's study (tip: c_i_t_.) is a counterpart to Miller's (1958a and b) study for Pacific City, no members of the Negro subgroup are represented in that influential structure. It is generally Clear that the status-role in influence structures may not be divided proportionately among the various ethnic groups (or nationality groups) in the community. It is quite plausible to expect 40 representativeness of a ethnic group among top influentials to vary with such characteristics as size (proportion of the population) and relative prestige. Ethnic identities, much like religious identities, are subject to value standards within the wider society, and, therefore, to patterns external to the particular community. The place that an ethnic (or religious) group whose status is low has in community decision-making would very likely depend upon the explicit effects the group can bring to bear on decisions by way of formal, legal activity. For example, their voting or economic power in a community may be important. For intra-community comparative purposes it will be necessary to compare the ethnic composition of the top influential hierarchy with the proportions of like-ethnic groups within the total population. Briefly, one can expect that a larger number of "representatives" from various ethnic "minorities" will be found among top influentials as the size of that "minority" increases in proportion to the total popu- lation of the community. Two generalizations can be stated. [IV.h:1] That a Siginificant proportion of influentials will be members of ethnic groups highly evaluated within the general society, viz. , anglo. And, [IV.hz2] That representation of an ethnic "minority" (subordinate group) among the ggneral influentials will be less than proportional to that minority's number in the total population of the community. 7 . Loc ali sm Four variables are to be summarized in this section. They are place of birth and length of residence in community, orientation toward the community, and number of years with firm of employment. 41 a. Place of birth and length of residence in community. - Freeman e_t a}. found that birth within the local community was more highly valued than birth in some other place (92' c_i_t_.: 16). Schulze and Blumber, on the other hand, found that five of their 18 "public leaders" (28%) were native-born Cibolians (tip: c_i_t_.: 295). More important than place of birth as a selective factor for influentials is the length of time that influentials have spent in the community. _ Schulze and Blumberg found that the median number of years residence of all 18 public leaders was.30 years (ibitii). Stewart does not give specific information as to the place of birth or the length of residence of each influential in the community (_O_p_. Cit. ). He does, however, report the general information that the length of residence in years increases as the number of mentions on various influence scales or indexes increases. Hunter reports that 16 of 27 persons interviewed in his study were born in the state; and that 12 of these 16 persons were born in Southern City itself (pp. gi_t.: 33). Hunter does not report length of residence. Merton reports sharp differences between the "localites" and "cosmopolites" on these variables. The local-oriented influentials tended to have been born in the vicinity. Also, 14 of the 16 had lived in the community over 25 years. The cosmOpolitan-oriented influentials had a more mobile background. "Fewer than half of the [14] cosmopolitan" influentials had lived in the community of Rovere for over 25 years (op. cit.: 395). Finally, Sauer found 34 of 39 influentials (88%) in Wheelsburg were born in the "midwest" as contrasted to those born in the West, South, East, or some country other than the United States. Two generalizations, one of which is not selective of influentials, may be stated. [IV. i: 1] That a group of influentials will have varyirtg characteristics as to place of birth. (Is is important that this be selected for testing since there has been disagreement on its iinport- ance.) And, [IV.i22] That a majority of influentials will tend to have lived in the community for. 25 years or more. b. Orientation to the community. - The primary distributing characteristic used by Merton to distinguish the influentials in Rovere was that of their orientation toward the community. He developed two 1 types: the "localite" and the "cosmOpolite. " The localite "largely confined his interest to this community. . . . he is preoccupied with local problems. . . . [and is] parochial"(o_& c_i_t_.: 393). The cosmo- politan type, on the other hand, "is oriented significantly to the world outside Rovere, and regards himself as an integral part of that world. . . . [he] is ecumenical" (1%.).2 Merton found differences between these two groups in their participation in community organizations-- the localites participated more--and in the spheres of influence within which each person developed his relations--the localite "is typically concerned with knowing as many people as possible " (ibid.: 396) 1The name types were adopted by Merton from Carle C. Zimmer- man, who, in turn,- "used them as translations of Toennies' well-known distinction between Gemeinschaft (localistic) and Gesellschaft (cosmo- politan). . . . here applied to empirical materials on types of influ- ential persons" (Merton, 1957: 393, fn. 7). Zimmerman's work was based on Ferdinand Toennies' Fundamental Concepts of Sociology (New York, 1940), a translation by C. P. Loomis. ZMerton is not, in this article, concerned with the "objective determinants" of differences in orientation. As he states, "A vaguely formulated question enables each respondent to project his basic orien- tations into his replies" (Merton, 1957: 394). He also indicates that more formal criteria of these types might create the necessity of development of intermediate types ”which approaches neither the local nor the cosmopolitan poll" (ibid.: 393, fn. 8). Cf. discussion of Stewart (1947a and b). —_— 43 while the cosmopolite is concerned with "the kinds of people they know. . . " (ibid.: 397). Stewart also divided his group of influentials on the basis of these orientations. He added, however, a middle type which he called "mixed. " In full, these types are defined as Loca1--"one whose interests and activities are confined just about exclusively to his home community, " Mixed—~"one who has interests extending well beyond the home community but who, as shown by participation in civic activities, is an integral and active member of the community, " _ Cosmopolitan--"one whose interests and activities lie principally outside the local scenes. " (22. cit.: 23). One other study used the variable "orientation. " This was Smith (22. cit.) who reports that all 16 persons identified by reputation and called "determinative influentials" were "local-oriented. " He contrasted this with a group variously called "economic elites, " and "absentee-unit elites" (ibid.: 87) all of whom were "cosmopolitan- oriented. ” Smith "defines" local orientation as describing those persons who (a) were more interested in local community newspaper, (b) evaluated events as to effect on community, (c) were more ethno~ centric about community, and (d) measured social popularity in terms of numbers (quantitatively rather than qualitatively) (_i_bi_<_i_.: 86). Both Smith and Merton agree that localites are "activists" in the community, i. e. , that they belong to organizations which allow them . broad contact with other community members. It is argued that broader social contact implies broader issue-interest and broader influence base. Smith's findings are entirely consistent.l . 1At this point the reader should be reminded of a qualification to generalizations on the nature of influence. Merton maintains that interpersonal relations with which he was' concerned comprise a differ- ent phenomenon than influence relations on the community level. This is exceedingly difficult to assess since Stewart, who used Merton's approach, is concerned with the phenomenon of influence identical to 44 Due to seemingly contradictory data the generalizations made on the basis of these studies must be tentative. However, it would be consistent with arguments presented earlier, i. e. , that influence in this context has a community-wide objective and that evaluations are important in giving an individual a potential for influence, to expect influentials to be oriented to the local community. Thus, the following generalization is :prOposed: [IV. j: 1] That a majority of a group of influentials will have an orientation to the community. c. Number of years with firm. - Sauer reports that 27 of 39 influentials (69%) had been with their present firm more than 20 years. None of the influentials had been with their firm for less than five years (32. cit.: 67). Though there may be a minimum number of years experience which "qualify" an individual in his evaluation by others, Sauer‘ 5 data hardly justify a firm conclusion. It is still quite problematic, in other words, as to what long years of experience with one firm might mean. Since it may be but one variable among a large number and since Sauer's data permit a statement, however tentative, the problem will be explored. (cont'd) that with which we are concerned, viz. , who in a community exerts influence on projects which are of community-wide significance? The differences in the influence phenomena notwithstanding it should be reported that Merton found his locally oriented influentials to be "monomorphic" in their influence-~i. e. , they exerted influence in a narrowly defined area, e. g. , politics or as "experts in limited fields. The cosmopolitan influentials whom he studied, on the other hand, tended to be ”polymorphic"--i. e. , they exerted influence in a variety of spheres, expecially those with the most influence (1957: 414). Smith does not give an indication of the specific areas to which his "determinative influentials" limited their influence. From his very general discussion on this point, however, it may be assumed that this influence is "general, ” i. e. , it comprises many different issue- Situations. 45 Thus, the empirical generalization is [IV. k: 1] That a majority of influentials will have had over 20 years experience with their firm of employment. C. Interrelation of Empirical Generalizations with Propositions from Section II. E. The section just completed has summarized a number of Specific studies which have been reported in the area of community influence structures. Only selected data has been abstracted from these studies. The selection of data was based upon a classificatory model drawn from the literature itself. From this selected data empirical generaliza- tions1 have been stated. Then the generalizations are interrelated with framework derived from theory, they may be considered hypotheses subject to verification or denial. It is this step which will be taken in Chapter VII. Justification for this move is simple; theory suggests , as has been shown, that the elements of a profile of influence potential may be drawn. Research, on the other hand, has verified for a single case the justification for the expectation of such a profile. It will be the object of this research to further verify the validity of these expectations. It was stated that positions in class, power, and prestige hier- archies contribute to the potential for interpersonal influence. In ad— dition, it was stated that the "contents" of evaluations relevant to the l"Empirical generalizations" is defined by Merton as "an isolated proposition summarizing observed uniformities of relationships between two or more variables" (1957: 95). In the form of isolated statements they are nothing more than a miscellany of propositions. With the order imposed upon them by generalization (such as Freeman's et a1. model, 1950) the set of statements is still an empirical generalizatio—r-i however more removed from reality. 46 profile-—which in effect give a differential advantage to one partner in a social relationship, ergo, influence--embody such general phenomena as social class and prestige,a ”history" of power which would involve components of the life cycle, social background, personal characteristics, and the personal control of relevant "objects" which give the influencer an advantage, such as a backlog of skill (related to the division of labor in the community) and wealth. Thus, at least the broad expectation of a "profile of influence potential" is substantiated from theory and, indeed, many of the specific empirical components also. The following hypotheses have been presented in this chapter. [IV. a: 1] That among a given group of persons who are classified as influential in a community, the majority of them will have had at least a year or more of college education. [IV. a:2] That a group of influentials will have significantly more education than would have been the case if the group were drawn at random from the population of the community. [IV. b: 1] That for communities in the United States, the dominant occupational background of influentials will be "business" occu- pations, followed by government positions and the professions. [IV. b22] That for communities in Mexico, the dominant occupational background of influentials will be both "business" and govern- mental status-roles. [IV. c: 1] That a significant proportion (a majority) of a given group of influentials will be male. [IV. cz2] That a group of influentials will have significantly more males among them than would have been the case had the group been drawn at random from the general population of the community. [IV. d: 1] That the mean age of a given group of influentials will be over 50 years of age. [IV.d22] That a given group ofinfluentials will differ significantly in their age characteristics from the general population of the com- munity from which they come. [IV. [IV. [IV. [IV. [1v. 47 . e: 1] That a majority of influentials will be married rather than single in marital status. e:2] That a given group of influentials will differ significantly in their marital status (more will be married) from the general population of the community from which they come. f: 1] That a group of influentials in a community will be drawn from a population whose fathers were engaged in white-collar or higher occupations . . g: 1] That a group of influentials in a United States community will have religious affiliations characteristic of those highly evaluated in the general society, viz. , Protestant, regardless of propor- tions of non-Protestant affiliations in the community. . g. 2] That a group of influentials in a Mexican community will have religious affiliations characteristic of the dominant religion of Mexico, viz., Roman Catholicism. ‘ h: 1] That a significant proportion of influentials will be members of ethnic groups highly evaluated within the general society, viz. , Anglo, for the United States communities. h:2] That representation of an Ethnic "minority" (subordinate group) among the general influentials will be less than propor- tional to that minority numbers in the total population of the community. .i: 1] That a group of influentials will have varying characteristics as to place of birth. . iz2] That a majority of influentials will tend to have lived in the community for 25 years or more. .j: 1] That a majority of a group of influentials will have an orien- tation to the community. k: 1] That a majority of influentials will have had over 20 years experience with the firm of employment. The hypotheses listed above will be grouped in two ways by which their reference will be specified. The first type refers to hypotheses about the relation between the influentials in a community and the 48 specified characteristics of the "non-influentials" in the community. This will be called the "intra-community" problem. It asks whether the influentials are affected by these populations from which they come with regard to specific characteristics. The hypotheses which are of this form are [IV. a: 2] [IV. b: 3] [IV. c:2] [IV. d:2] [IV. e:2] [IV.hz2] The second type of problem will be called "intra-influential. " It is concerned with generalized characteristics which can be attributed to influentials no matter what the characteristics were of the community within which their influential social relations took place. The questions regarding the components of the problem are specific. The hypotheses relevant to this problem are [IV.azl] [IV.czl] [IV.d:1] [IV.ezl] [IV.le] [IV.izl] [IV.i:2] [IV.jzl] [IV.k:l] Finally, there are a group of hypotheses of the intra-influential type which Specify inter—nation uniquenesses. The five hypotheses of this type are [IV. b: 1] 49 The hypotheses will be retained within the classification of social characteristics as they have been presented in the foregoing section. The grouping which has just been presented will not be used until the summary of the results have been presented. CHAPTER V RESEARCH METHODOLOGY A. Introduction Data gathered and reported in this thesis were part of a larger project at Michigan State University for which funds were made avail- able by the Division of Hospital and Medical Facilities of the United State Public Health Service for project W— 108, "Anglo-Latino Relations in HOSpital and Communities, " and the Carnegie Corporation for a project dealing with the United States-Mexican Border. This fact in large part dictates the selection of sites and a number of questions in the interview schedule.- That schedule is attached in Appendix I. The purpose of this chapter will be to present the descriptive data on the communities selected, the technique used to determine influential persons, relevant aSpects of the schedule, the respondents interviewed, and other details of analysis. B. Research Sites Six communities were selected representing a wide range in the variables of size, industrial composition, and ethnic composition. Details on these variables for each of the six communities are sum- marized in Table A, Appendix II. Brief descriptions of the six communities follow. 50 51 San Diego is a commercial and financial center in southern California with especially large proportions of its population employed in public administration and personal service. Its industries include fishing, fish packing and aircraft. It is an important transportation center with an ocean port. One major college is located there. El Paso is a transportation and communication center and a major tourist port of the southwestern United States. Its industries include clothing manufacture, metal and oil refining, and meat packing. It houses two major military installations and a small college. Tucson is a financial and commercial center of southern Arizona. A high proportion of its population is employed in personal and professional service. Its favorable climate attracts considerable tourist trade and a number of state, federal, and private hospitals. It has a major state university. Las Cruces serves as a commercial center of a district rich in cotton, corn, fruit, alfalfa, truck and dairy products. Ahigh proportion of its population is employed in public administration. It houses a state university. Ciudad Juarez, Chihuahua is a commercial and distributive center of Nor,th-Centra1 Mexico. It is also a minor manufacturing center. As a transportation center, it probably reflects the extensive tourist trade which comes to Mexico through El Paso and Juarez. In addition, it is the largest of the Mexican border cities. Tijuana, Baja California is also a commercial and distributive center for the Northwesternmost corner of Mexico. It is a minor manufacturing center. It is also a tranSportation center and attracts tourist trade from the western states, particularly from the San Diego environs. (Descriptions from D'Antonio, et al. , 32' c_i_t_.: 442). — _ 52 C. The Reputational Technique in Process 1. Knowledgeables A group of knowledgeables were selected1 in each of the six communities. 2 Each knowledgeable was asked to name as many 1Since this was contract research, professional sociologists were employed to select the influentials and conduct the necessary inter- viewing in the community with which they were familiar. These persons were under the direction of a field director who helped in all phases of the research. Zln each case there was an attempt to get as wide a representation of the various institutional sectors of the community as possible in the persons of the knowledgeables. However, the same status-roles are not appropriate for each community. Tucson offers an example of the status-roles used as knowledgeables in the cities of Tijuana, San Diego, and Las Cruces. Twenty-one individuals were interviewed. They were the radio station manager, the Chief of the Associated Press bureau, the City Hall reporter for one of the newspapers, the Managing Editor of a newspaper, a "prominent" representative of the Democratic Central Committee, the President of the Tucson Council of Churches, the Secretary to the Bishop (Roman Catholic), Assistant Superintendent of Schools, the County Superintendent of Schools, the Advertising Manager of the Chamber of Commerce, a representative of the Better Business Bureau, a member of the largest Public Relations firm in the city, the Director of the Council of Social Agencies, a Director of a large labor union, the County Health Officer, the Secretary of the Chairman of the Board of a Bank, the Director of the YMCA, the Director of the Red Cross, an Executive Vice-President and Public Relations Director of a major bank, and the Director of the United Fund. (Source: information from the field researcher in Tucson.) For two cities, El Paso and Juarez, a modified technique was used. Both of these cities had been studied in detail in 1954-1955 (the reader is referred to the report of this study for relevant details including methodology, D'Antonio, 1958: 47-51). Consequently, extensive lists of influential persons were already avail- able for these two communities. Therefore, a small group of knowledge- ables who had been particularly good informants in the earlier study were asked to review these lists for their current accuracy. These persons were an active member of the Medical Society auxillary, a Vice-President and public relations officer of the largest bank, an active political reporter, and the City Clerk all in El Paso. 53 influential persons in the community as possible.1 The criterion which each knowledgable was asked to use was that these persons be the "most influential" in community affairs in general. That is, they were to name persons who could and did initiate, block or significantly help resolve issues which were of community-wide interest. From these lists a combined general list was complied. Only the most frequently named persons were selected for the final list of TOp Influentials.2 These lists varied from about 22 to 45 persons for each of the six c ommuniti e s . 2 . General Influenc e Interviews were conducted with the group of Top Influentials in each community. As many persons as possible were interviewed with a schedule of questions. Among the questions was the following (see question 36, Interview schedule, Appendix I). Listed above are the names of people who are con- sidered to be influential in this community. We have talked about some of them already. Now we would like you to consider the whole list and rank the ten you consider to be the most influ- ential. If a name has been omitted which you believe should be here, please feel free to add it. They are listed in alphabetical order. The respondent was encouraged to add names which should have ap- peared among the Top Influentials (in his judgment), to vote for that 1A list of about 50 persons was requested from each knowledge- able. However, not all knowledgeables could name this many persons. The lists of completed names from knowledgeables varied from 20 to 50 names. 2Researchers were instructed to include about 40 persons on the final list (see footnote 2, page 12 for references). Smaller lists indicate that the next lower number of mentions by knowledgeables would have included far over this number. 54 person as one of the ten most influential if he wished, to remove any names which he felt should not have been left among the Top Influentials, and to vote for himself if he chose. The question asked respondents to rank the tOp 10. This was not required in later interviews. Many respondents, though perfectly willing to name the ten most influential, were very much troubled by the request to rank them. It is the result of this selection together with the Knowledgeables list which is the final list of Top Influentials (T.I.) used in each com- munity. Since each respondent, who is a T.I. , has cast 10 votes for the 10 most influential, there is an additional subdivision possible. As all the votes are tabulated a list of the ten most frequently chosen influential may be selected. These are the Key Influential (K. I . ). Where possible data 6%? obtained on all of these persons named by both knowledgeables and respondent Top Influentials. Obviously, some TI were unavailable for interviews.1 Certain information was available on nearly all influentials named in each community. For example, for each person named there was an attempt to get his firm of employment and his position within that firm. This information did not require an interview. Since there was a substantial difference between the number of persons named and the number of persons inter- viewed in some communities, it was deemed advisable to compare these two groupsu-interviewed and not interviewed influentials--as to the possible bias in selection. When complete information is not avail- able a complete comparison is impossible. However, the groups can be compared as to the various occupational categories from which they 1There were, however, very few refusals. In El Paso there was but one formal refusal among the 33 for whom an interview had been scheduled. There are other shortcomings in that some respondents refused to answer certain questions. The number varied from question to question. No attempt has been made to classify the various types of "no answers" to specific questions. 55 come. The data are presented in Table B, Appendix II. There was no significant statistical difference between these two groups for any of the six communities when they were compared as to the occupational categories among which they were distributed.1 3 . Limited Influenc e "Limited" scope of influence was determined by another specific question (question 28, Appendix I), which reads Suppose that a major hospital project were before the community, one that required decision by a group of leaders whom nearly everyone would accept. If you were completely free to choose, which people would you choose to make up this group--regardless of whether or not you know them personally? This question asks the top influentials to choose about eight persons. The question appears early in the interview schedule of questions. It is the first question which requests the mention of specific persons . . 2 1n the community . 4. The Time Dimension Finally, the dimension of time will be covered for one community. It has been mentioned that an earlier study in El Paso yielded a'group of influentials as of 1954-1955. This group is compared to the list determined in the fall of 1958. IThe Chi—square test was used to test the null hypothesis that there is no difference between the two groups as to the proportion of persons in various categories (Siegel, 1956: 101). This hypothesis is not rejected in any case. 2It is not the last question before the respondent is shown the pre- arranged list. There are two scope questions in which the respondent is asked the details of the resolvement of two issues. Only after the respondent has confronted this variety of scope questions is he asked to make the choice of the ten most influential persons from the pre- arranged list. 56 D. Research Techniques Interviews were conducted with from 22 to 42 influentials in each community. The interview schedule took about one and one-half hours to complete.1 The schedule was divided--temporally--as follows. A section on the personal background information of the respondent (questions 1 through 14), was followed by questions on the respondent's-- and his firm's--activity in various community affairs (15 through 27), the question on "limited" influence (28), the role of the medical doctors in community health issues (32 and 33), two questions on issue resolve- ment (34 and 35), the list of general influentials and respondents inter- action with them (36 and 37), questions on perception of issues (38 and 39), and a final section on the respondent's image of persons of Mexican descent in the community (40 through 47). E. Analytical Tools The statistical techniques and tests used were Chi-square (Siegel, 1956: 104-111), t-test (Dixson and Massey, 1957: 115-119), a corre- lation coefficient and a multiple correlation coefficient (Walker and Lev, 1953: 230-242 and 315-322), and a binomial test (Siegel, pp. 311.: 36- 42). Chi-square is used to test such hypotheses as "there is no dif- ference between two groups in their proportions as they are distributed among Specified variables. " The t-test tests a hypothesis “that a uni- verse mean is not less than a specified constant. "2 The binomial tests 1Not all interviews took this long since there were refusals to answer certain phases of the questions. On the other hand, many inter- views took as long as three or more hours. In general, the cooperation of the respondents was superb. zThe t—test may, of course, test the hypothesis in which the con- verse to this statement is tested or in which one is interested in the equality of means. None of these forms of the hypothesis are used herein. 57 the hypothesis "that there is no difference between the probability that influentials are drawn from a population in which a specified variable is distributed according to a given proportion. " The correlation co- efficient is used for descriptive purposes while the multiple correlation is used to determine the correlation between observed scores on two indexes of influence (as determined in the 1954-1955 study) and the index of general influence from the current study. CHA PT ER VI FINDINGS ON SPECIFIC CRITICISMS OF RESEARCH METHODOLOGY OF REPUTATIONAL TECHNIQUE A . Introduction The Specific hypotheses which-are to be tested in this chapter have been presented in Chapter III. Following a presentation of the results a brief comment on the general significance will be stated. B . The Hypothes es 1. The "Limited" Scope Question In hypothesis [C. 1. a: 1] it was stated That a given individual's influence measured on a general influence index will correlate with influence measured on a particular or limited scope situation. Two indexes of influence have been described in Chapter V: a general influence index on which a given individual receives a number of "votes" (question 36, Appendix I), and a limited influence index on which a given individual receives a number of "mentions" (question 28, Appendix I). The general influence index is derived from a question in which the respondent is asked to name persons who have general influence in the community, i. e. , are influential on community-wide issues. The limited influence index is derived from a question solicit- ing those influential when the issue area is perscribed as health. For comparison purposes the indexes will be dichotomized into "high" and "low" number of votes or mentions. The "high-low" 58 59 dividing point will be based upon an arbitrary ten persons (or, under the condition of ties, the nearest number) who received the most votes-mentions for "high" and the remainder for "low. " Comparisons are presented in Table l. Chi-squares were com- 1 The Chi-squares for each puted for each community distribution. community were: San Diego, 20.73; El Paso, 17.94; Tucson, 8.45; Las Cruces, 20.76; Ciudad Juarez, 17.68; and Tijuana, 24.47. Each Chi-square is significant at a probability level greater than . 005. Therefore, the null hypothesis that the two indexes are unrelated or independent is rejected for each community. Therefore, the hypothesis as stated above is confirmed. 2. The "Experience" Question In hypothesis [C. 1. cd: 1] it was stated That a given individual's influence measured on a general influence index yields a group of persons whose occupational experience involves executive or directive responsibilities . Executive or directive responsibilities will be defined to include major executives in business, 2 executives in government either elected lChi-square is an appropriate statistic since marginals are fixed and the test must be one of the independence between the two indexes. The null hypothesis which is tested is "that there is no difference be- tween the proportions of persons given High and Low general influence in the proportion given high or low limited influence. Alpha will be . 05 for which one degree of freedom requires a x2 > 2.71 in order to reject the null hypothesis. _ ZMajor executives in business include hotel and restaurant owners, retail store owners when specified as executive, manufacturers including officials, owners, managers and superintendents, bankers and bank officials, real estate agents and officials, builders and building con- tractors, owners of two or more drug stores, radio and TV station executives, and Generals or Admirals in the military. 60 Table 1. The Number of Persons with High and Low General or Limited Influence Where Marginals for High and Fixed at Ten for Each of Six Communities. General Influence Chi- Community and Low High Totals Square Limited Influence (n) (n) (n) (X2) P. San Diego 11 or less-12 or more votes High (8 or more 1 10 Mentions) 9 Low (7 or less 2 Mentions) 6 Z . 28 Totals 27 ll 38 20. 73 .0005 El Paso 10 or less-ll or more votes High (4 or more 5 6 11 Mentions) Low (3 or less 7 4 Mentions) 9 5 8 Totals 84 ll 95 17.94 .0005 Tucson 13 or less—14 or more votes High (8 or more 6 4 10 Mentions) Low (7 or less 0 Mentions) 9 6 96 Totals 96 10 106 8.45 . 005 Las Cruces 12 or less-13 or more votes High (6 or more 4 6 10 Mentions) Low (5 or less 2 4 Mentions) 7 76 Totals 76 10 86 20.76 .0005 Ciudad Juarez 8 or less-9 or more votes High (5 or more 6 7 13 Mentions) Low (4 or less 2 Mentions) 7 5 77 Totals 78 12 90 17.68 .0005 Tijuana 3 or less-4 or more votes High (3 or more . 5 l Mentions) 8 3 Low (2 or less 2 Mentions) 6 3 65 ¥Totals 67 11 78 24.47 . 0005 P is one-tailed. 61 or appointed,1 and salaried professionals.2 All other positions, including farming, are classified as non-executive or directive positions. Table 2 is a summary of this data. No distributions of the com— munity population are available against which these "sample" results could be compared. However, it is argued here that the proportions listed in Table 4 are significantly in favor of executive or directive experience. The extremes in this category range from 89% for San Diego to 459E in Las Cruces.3 A qualification of the hypothesis is necessary. Namely, that though executive experience may be helpful in the achievement of influence, there is a variation from community to community. It is doubtful that presence of businessmen as the majority of executives is a distinguishing variable since executives are a very limited number of all businessmen and since buinessmen also appear among the non-executives. lGovernment executives either elected or appointed include mayors, County Judge or other Chairmen of County Boards, and County commissioners. ZProfessionals who are salaried include college presidents and professors, clergyrnen, Superintendents of Schools. 3There is a statistical test which may be computed on these data also. The test is the binomial which asks, in the form of the null hypothesis, whether there is no difference between the probability of an influential being employed in an executive position or not being so employed. The alternative hypothesis is that the probability of being employed in an executive position is greater than not being so employed. It is necessary to assume that the probability of employment in either position is equal to . 5. The decision to reject the null hypothesis is based on p being < . 05. Inspection of the binomial probabilities on Table 4 show H0 is rejected in the case of four of six communities. 62 N a. o o o o caCSSom I o m o m 0 85% O O m fl N O Owflgmmsom H E N m N EfimcofimmoGOanufiEom po.»oaeCOu:em NE MN 0 m o A chofimmeMOHnH 12894 0 o q o m o ”Hemgchg/oO o>SmHmwwoA m o o o o o ”HCmECHmi/oO m _ w E m N o>33uoxm HOCEZ .3330 can 0 N d o o o hoped descmz 2 o>3oopfifl NEE mm LVN mm Om Em mN mm oq Om mm. AE w .HO ®>SSUQXH uOZ Hmcoflmmowounfi m. m a. m w E Uofipmfimw o>35oexm o o . M v A uC®ECpo>oO v>flu Om 1:. m: 5N OM OM uduoxm Home: o>fioefifl SN mm mm 3 a... me 3. E X S 3 mm 2 .8 QEBSQ 9: e; E as E e: E E E s: E e; 3 828m 1308 mcmsflh noumsh moode mad COmudH Ommm am— owofiQ cmm .mofificsEEoU xfim mo seem pew ousfidopw/w a: pen mcoflfimonw £05m cw “OZ ponESZ org. .mcofifimonw o>fluopfifl .HO o>350oxm a: mdmficvdch :< mo 533832 0&8 .m. 3nt .mddOpm 03s. 05 200330 NGLMMZEG mo mfimvfitomens 08. pooflms 3:03 mesmsqmuflfiO 05. mo 352 be; u ~X 33:3 so“ .2 ”HA: pew .N. ”MIH .2 "m2- .3 ”m: .mcmsfle ea... .30. n Nx 5E3 .80 3 “as. .s ”MA .3 ”mm- .m: ”m: .Nfimsm 4:0. n MK £0E3 HOW ms ”HA: Use .3 “HA .Nd ”m3- .Nd ”m: .900de $me .23. n ax £053 MOM A ”m4- 28 .o. ”m; .2 ”m7 .2 "E 683:. .eom. ”Nx €33 80 2 ”m4- 98 .2 ”m; .2 “mm- ON um: .0mmnm mm .omJ n NX m were?» mCOmCmmEOO .p0303u0ucw Hos 0.303 pew max/35088 00: 0.53 0£3 N «Ed .373 woe/330003 ”.0: “:03 0:3 U030fi>umucfi 0903 0&3 N .33; U030fi>soucfl 00: 9:03 0£3 $3390on 0303 0:3 N. Am: met/330003 0.303 033 00301302“ mfimficoddcfl mN 0.303 0.3:“ owowQ 5mm pom 0033\5002: 0.203 033 805. wcoam 0.0.02? 09. Umumeroo 03 $55 mCOmueQ emech. 533535500 02m 2.3 Go Lowe cw Hmflcmsdcw cm mm 0085: uo>0 ramped some so coflmasofii movies“ 038. wit. so 930 3 IO wwo. ONO. as. 000. N00. 000. V Q 525303 fiasco £23 .3500 3500 mcofiwmssooo Esau pew 0>3500xm|coz so e>3500xm as“: “moo. Hmfigocfim owm on: moo 00L cm on: mm ooh we ooh No on: cm 25.0% 0 .OH N. N N o o o o o o o 0 03933333094 qu a. o o o o a a o o o o o o Sagas 64 If it had been assumed that the critics of the reputational tech- nique defined "formal leadership" as "governmental executive positions" the assumption would clearly be questioned by the interpretation of these data. It is held then, that to some extent the hypothesis that executive or directive experience is conducive to influence status-roles is true. 3. The Longevity Question The hypothesis on longevity [C. 2: 1] was stated so That insofar as individuals are evaluated as influentials at one point in time, the same individuals may be expected to be evaluated as influentials at a later point in time . In Chapter V it was mentioned that the data on influentials obtained in a 1954-1955 study consisted of two indexes of influence (D'Antonio, _(_)__p_. (_:_i_t. ). The two indexes were determined on the basis of a frequency count of mentions of given individuals on each of the following questions. 1. Would you please give us a list of names of people whom you believe are and have been the most influential in El Paso government and politics in recent years ? (D'Antonio, op. cit.: Appendix A, 241). _ _ 2. Will you please give me the names of the most influential businessmen in El Paso? (ibid.: 237). There were 51 persons who qualified as business influentials and 30 who qualified as political influentials as a result of these questions. Eleven persons qualified on both lists thus gave a total of 70 different persons who were reputed to be influential in business and/or politics in El Paso in 1954-1955. When the tabulation of votes on the 1958 list was completed, there were 48 names. Eight of these names had been added in 1958, 65 i.e. , they did not appear on the 1954-19551ists.1 In order to find some single index of influence for 1954-1955 by which relative 1958 influence may be assessed, the inter—correlation of the three indexes must be completed. For purposes of a corre- lation, there are certain exclusions which must be made. For example, the presence of too many persons with a ”0" score or vote may yield a spurious correlation since, then, the assumptions of correlation are not precisely met:2 Therefore, the following criterion were devised. (1) Eliminate all persons who have retired from active public life in the years between the two studies. (2) Set up the requirement that an individual must have been named in 1955 on at lease one of the two indexes--business or politics--and that he must have been named in 1958.3 Forty-five persons are thus eliminated from the total list of 79 persons.4 The 34 persons who remain were thus named on either or both of the indexes used in 1954-1955 and on the general index in 1958. lHaving received a vote is important here since the fact of having a vote will be used as a criterion for inclusion on the lists to be corre- lated (see below). 2At the same time the exclusions must be made on criteria general anough so the results are not purposely biased. One cannot select out those who have a few votes on one index and a large number of votes on another. 3Criterion (1) follows in that new names may be added when older influentials are replaced. Criterion (2) follows since some persons no longer take part in influence relationships (for whatever reasons) and therefore pass from the scene. Of interest to the problem pre- sented here is: Do the same persons recur on influence lists and at what relative magnitude? 40£ these 45 persons, none had been added in 1958, 16 had with- drawn from public life or were among the 9 who were named as political influentials only or the 20 who had been named as business influentials only. Of this group which received mentions on only one index, (31.1.2 two had received more than seven votes on one list to which they had been named. One was the County Judge who had since "retired, " i. e. , been defeated in election. 66 Correlation coefficients may now be computed on these indexes.1 Simple product moment correlations computed on each pair of indexes follow: Index 1954-55 1958 Index Political Business General (Z) (Y) (X) 1955 Political (2) 1.000 -.098 .296 1955 Business(y) -.098 1.000 .581 1958 General (x) .296 .581 1.000 A multiple correlation2 was computed on these indexes. Of interest is the correlation of the general index on the political and business indexes. Rx. yz = . 683. Since RZX. yz = proportion of variation of the general influence index which can be ascribed to variation in the business and political indexes (Walker and Lev, op; Eit': 322), this variation may be calculated to be 47%. Considering the state of knowl- edge concerning the components of influence, this supports a statement that a relationship between these indexes exists. The correlation also gives us a justification for combining the two indexes used in 1955. Table 3 presents the data with the comparisons between the com- bined 1954-1955 business-political index and the 1958 general influence index. A Chi-square computed on these data lead the investigator to reject the hypothesis that the indexes are independent, the probability level being < . 02. The hypothesis presented above is therefore confirmed. 1Computation of both correlations follow Walker and Lev (pp. Sit”) The correlation coefficient of all persons named on either the business or political index in 1954-55 was r = . 028. That relationship is not notably different from the one presented in this text based upon an N of 34 rather than 70. ZComputation for multiple correlation follows Walker and Lev (ibid.: 315-322). Table 3. The Number of Persons with High and Low Influence on the Combined Business-Political Index of 1954-1955 and the General Influence Index of 1958 for El Paso.a 1954-1955 Business 1958 General Influenceb and Political IndexC Low High Total High 5 7 12 Low 18 4 22 Totals 23 ' 11 34 x2 = 5.72 8LThe marginals have been fixed so that Key Influentials are compared with Top Influentials. High scores include those who received 11 votes or more and Low refers to those who received 10 votes or less. c . . . ° High scores include those who received 19 votes or more and Low refers to those who received 18 votes or less. C . Chapter Summ ary Data have been presented by which it has been argued that certain specific criticisms of the "reputational" technique have been invalid. These criticisms concerned the problem of "scope" of influence, i. e., when "scope" is "narrow” an individual will be limited in the contents of the social relationships in which he has influence. It does not follow that a general influence index yields only individuals of limited scope when a limited scope index also yields substantially the same individuals. Since this relationship held up over six widely different communities, it is much more reasonable to suggest that general influence permeates many different contents of social relationships. Though the data used to confront the problem of the continuity of a group of influentials over time was not discussed in this context, it 68 is also relevant to the problem of scope of influence. Since the two indexes determined in 1954-1955 were essentially scope indexes-- albeit a very different classification of scope, i. e. , business and politics--the fact that each of the indexes of this "special" influence contributed to general influence supports the thesis that a general influence does indeed exist. It is not argued here that there are not individuals within a com- munity who indeed do limit (or are limited in) the social relationships in which they exert influence to certain content areas. However, neither was that presented as an alternative thesis by the critics of this particular technique. On the problem of whether there exists a group of influentials who are reCOgnizable over time it has also been argued that such is true. Nothing was implied in the ”longevity" hypothesis concerning the inter- relationships of the influential group. That is, nothing concerning whether the "group” of influentials either acted in concert or acted as individuals . CHAPTER VII FINDINGS ON THE ELEMENTS OF THE PROFILE OF INFLUENCE POTENTIAL A. Introduction Hypotheses and data relevant thereto will be presented for each of the items listed in the profile of influence potential. A summary section will follow the presentation of data. B. Hypotheses on the Elements of the Profile of Influence Potential 1. Social Level a. Education. - In hypothesis [IV. a:2] it was stated That a group of influentials will have significantly more education than would have been the case if the group were drawn at random from the general population of the community. This hypothesis must be restated in the 1 form of a working hypothesis. The working hypothesis is That a group of influentials will have proportionally more education than all persons twenty-five years old and over in the community when the distribution of persons is divided at median number of school years completed for all persons . Table 4 presents the data on the level of educational attainment for six communities. The distribution of the number of school years completed for all persons 25 years old and over in each community has 1Following Zetterberg (1954: 31) a working hypothesis is the hypothesis that is submitted to empirical test. 69 70 Table 4. Level of Educational Attainment of Top Influentials and Years of School Completed by Persons 25 Years Old and Over Dichotomized Near the Median of School Years Completed for Each of Six Communities. a f - Community and All Persons 25 Years of School Years of Age Chi- Completed and Overb Top Influentials Total Square pC (n) (if) (n) W) (n) (X3) San Diego, SMA (Median = 12.0 years) 12 years or more 161,975 51 25 96 162,000 11 years or less 158, 309 49 1 4 158,310 Totals 320, 284 100 26 100 320,310 19. 39 < . 0005 El Paso, SMA (Median = 9. 3 years) 9 years or more 47, 885 51 30 97 47, 915 8 years or less 45, 799 49 1 3 45, 800 Totals 93,684 100 31 100 93,715 25.32 < .0005 Tuscon Urban Area (Median = 11. 8 years) 12 years or more 13, 230 51 40 95 13,270 11 years or less 12, 738 49 2 5 12,740 Totals 25, 968 100 42 100 26, 011 34.44 < . 0005 Las Cruces Urban Area 9 years or more 2,674 49 26 100 2, 700 8 years or less 2, 740 51 i O O 2, 740 Totals 5, 414 100 26 100 5, 440 26.12 < . 0005 Municipio of Juarez (Median = 3. 9 years) 7 years or more 5,772 11 25 100 5, 797 6 years or less 47, 018 89 0 0 47,018 Totals 52,790 100 25 100 52,815 183,43 <.0005 Municipio of Tijuana (Median = 4.0 years) 7 years or more 3, 678 13 19 90 3, 697 6 years or less 23,776 87 ' 2 10 23, 778 Totals 27,454 100 21 100 27,475 99.64 <.0005 aCensus data is from Table 34, "1950 United States Census of Population: Texas, California, Arizona, and New Mexico, General Characteristics, " and Cuadro 8, "Septimo Censo General de Poblacion, 6 de Junio de 1950, Estado de Chihuahua and Baja California Territorio Norte. " Persons for whom school years completed are not reported are omitted. CA one-tailed test is used. «J p—o been dichotomized at the median. The influentials are compared with this distribution. Chi-square is computed for each community. The results were: San Diego, 19.39; El Paso, 25. 32; Tucson, 34.44; Las Cruces, 26.12; Juarez, 205.42; and Tijuana, 99.63. Each chi-square is significant at the probability level of . 0005. The null hypothesis that the influentials' level of education is distributed as that of the general population is rejected. The hypothesis as stated above is confirmed. More specifically, the hypothesis [IV. a: 1] was also stated That among a group of influentials in a community, the majority of them will have had at least a year or more of college education. Table 5 presents the number of influentials with some college education as against those with a High School education or less for each of six communities. In addition, the mean number of years of education is presented. The mean number of years education for influentials in each community were for: San Diego, 14.69; El Paso, 14. 76; Tucson, 15.07; Las Cruces, 16.23; Juarez, 14.00; and Tijuana, 13.05. It will be necessary to determine, statistically, if these means differ from the specified mean, viz. , 13 years, significantly. Inspection indicates all these communities have groups of influentials whose mean number of years of education are equal to or greater than. 13 years.1 1This is a case in which the statistical computation is a manipu- lation because of the particular hypothesis which is to be tested. The null hypothesis would be that the universe mean (taken from the sample mean) is equal to or greater than the constant, 13 years. A t-test for difference between a universe mean and a specified con-stant has been computed. H0, is to be rejected, with an alpha of . 05, if t < tl-alpha' The computed t score and the necessary rejection region for each size N are shown on Table 5. In no case can HO be rejected. Therefore, the hypothesis is confirmed. 72 3:- :57 woeg- 23.7 30.7 wows- 3038mm mo.o+ ww.z+ oo.o+ mw.m+ mw.m+ no.m+ ”a 3;... 36 Ed $6 34 add u .o.m 3.2 8.3 3;: 8.2 3.3 3.3 n m $2 o2 2 OS mu 2: cm 2: 2. 2: mm of 3. 230m. om 3 o 3 o S m m: m E. w 2 w 984 8 Serum swam wfi em 2 A: E S S :w E on mm 3 f 6.82 no wwoZOO 2Com a: :2 E :3 3 :3 E Q; E :3 E :3 x5 $308 mcmsflflrfl nonmsh moude mwiH 500de. Ommnm Hm owofiQ cam soflmuSUH .mmfiZCdEEOO 02m mo gumm no“. mmod .Ho COSNUSUM doonum swam a £935 Use «.502 .Ho COSMUSUM owQSOU mo .33? 320 new? mHmSCoSCCH mo ponESZ .m 3an 73 Therefore, the hypothesis as stated above is confirmed. b. Occupation. - Before confronting the hypotheses that predicted which occupations would predominate among influentials, it is necessary to meet the problem of the general distribution of occupations in the community and among the influentials. By following arguments pre- sented earlier a general expectation may be stated. It if [IV.bz3] That a group of influentials will have significantly different occupations than would have been the case if the group were drawn at random from the general population of the community. Table 6 presents data on the number of influentials and the total male experienced civilian labor force by certain occupational cate- gories for all communities in the United States in the sample. Data are also presented on the number of influentials and the total work force by certain principal occupational categories for all Mexican communities. 1 Chi-squares were computed on each of these classifications. The chi-squares were 145. 21 for San Diego; El Paso, 191. 30; Tucson, 74.04; Las Cruces, 54.09; Juarez, 1042. 37; and Tijuana, 567. 50. These are all significant beyond the probability level of . 0005. The hypothesis that the influentials are drawn proportionally from occu- pational categories as they are distributed in the community is rejected. Since it has now been established that influentials are not randomly selected from the population of the community in which they reside as to some broad occupational categories, the specific hypotheses which specify the content of the alternative categories of occupation may be presented. The relevant hypotheses were [IV.h:1] That for communities in the United States, the dominant occupational background of influentials 1The N for influentials for Table 8 is based on all influentials named. This N is taken to utilize all the information possible on the group. Reference to Table 2 indicates there was no significant differ- ence between the influentials interviewed and the total group of influentials on the variable of occupation. The census classifications between the United States and Mexico are not perfectly comparable. However, it is doubtful that erroneous comparisons would affect the results since the groups are substantially different. 74 Table 6. Number of Influentials and Total Male Experienced Civilian Labor Force by Certain Occupational Categories for all United States Communities and by Total Work Force by Certain Principal Occupational Categories for all Mexican Municipios. a Community and Total Experienced Chi Occupational Civilian Labor Force Top Influentials Totals Square p Category (n) (‘7) (n) (9E (n) (X2) San Diego, SMA Managers. . .b 15, 963 14 3o 83 15, 993 All Others 101,534 86 6 17 101,540 Totals 117,497 100 36 100 117,533 145.21 <.0005 El Paso, SMAb Managers... 5,453 14 38 61 5,491 ProfessionalC 2, 986 7 19 31 3,005 All Others 31, 779 79 8 8 31, 784 Totals 40,218 100 62 100 40,280 191.30 <.0005 Tucson, Urba Area Managers.. . 1,662 17 26 59 1,688 ProfessionalC 1, 226 13 14 32 l, 240 All Others 6,686 70 4 9 6,690 Totals 9, 574 100 44 100 9, 618 74.04 <.0005 Las Cruces, Urban Area Managers. ..b 417 16 22 45 439 ProfessionalsC 308 12 15 31 323 All Others 1, 948 72 12 24 1, 960 Totals 2,673 100 49 100 2, 722 54.09 <.0005 Municipio of Juarezd Directing. . .f 566 1 51 59 617 Professional 2, 166 5 24 28 2, 190 All Others 39,159 94 11 13 39,170 Totals 41,891 100 86 100 41,977 887.06 <.0005 Municipio of Tijuanad Directing. . . e 277 1 43 69 320 Professionalf 1, 359 6 13 21 1, 372 All Others 19,898 93 6 10 19,904 Totals 21,534 100 62 100 21,596 567.50 <.0005 aData from census' is from Table 35, "1950 United States Census of Population: California, Texas, New Mexico, and Arizona, General Characteristics, ” and from Cuadro 11, "Septimo Censo General de Poblacion, 6 de Junio de 1950, Estado de Chihuahua and Baja California Territo-f'io Norte. " — — _ 15Managers, officials, and prgprietors, except farm. CProfessional, technical, and kindred workers. Total work force data for the Mexican municipios includes both men and women. Also, it includes those persons who are "economically active: employed or unemployed for not more than 12 weeks. " . . eDirecting personnel except in agriculture, cattle, forestry and fishing. fProfessionals and technicians in every branch of activity. 75 will be "business" occupation, followed by government positions and the professions, and [IV.bz2] That for communities in Mexico, the dominant occupational background of influentials will be "business" and govern- mental positions . Table 7 shows the number of influentials classified in each of four occupational categories: business, professions, government, and mis- 1 . . . . . cellaneous. For each of the 81X communities, buSiness occupations are the most frequent. And, for each of the six communities, professionals are the next most frequent. However, comparing all four United States communities with the two Mexican communities, there are nine govern- mental status—roles in each of the Mexican communities and two, seven, two and two in the United States communities, respectively. Thus, in sheer numbers, the Mexican communities offer a greater number of governmental status-roles even though the proportions of these roles are smaller because of the larger number of total influentials. Therefore, support is indicated for the hypothesis which specifies that business occupations will predominate among influentials in United States communities. Though this seems to be true there is a qualifica— tion which must be stated. The variation within communities is sub- stantial. Where 78% of the influentials in San Diego are in business occupations, only 4192, of those in Las Cruces are so involved. Also, 24%. of the Las Cruces influentials are involved in occupations $135? than business, professions or government. lBusiness occupations included banking and finance, insurance and real estate, retail merchants, wholesale merchants, manufactur- ing and utilities, construction and transportation. The professional category included religion, welfare, education, medicine, law, and communications. The government category includes local, state and federal government while miscellaneous includes agriculture, labor unions (officials), households, executive secretaries and officers of voluntary associations, personal service occupations, retired persons, and active politicians (a classification used only in Mexico for persons who claimed no other occupation). Spomofiwo Some ca «626365 mcoSmQSouo mo 2323533 p0w .mn omen J euocabOH eem . m. 76 0mm 2: No 02 cm OOH we :2 we 00‘ No :2 pm 34305 mm on 0 ma Z «.N NH 0 w w m m H msoecmdeomflz Em m: 0 0‘ o v m. m N 2 N. o N “coECHEVOO om EN ma mm «am am ma mm a; am m: I m mfimcofimeMOAnm 0: mm «4m 04. N4. :4 ON mm 4N om am we mm mmecfimdm Ev 6.63 c: 3; A5 E $3 .5 A5 3w: A5 6.6.3.: A5 mneasomfino 130R. 95:38 nonmsh 966de med COmoSH. OmmnH mm owofiQ cam fimcoEmQSUUO 603358500 62m mo zomm pow .mdoecmzoumflz pew #:6822360 .mcoammowonnm .mmecfimdm mo meComeumO HmcoflmasooO me: E po>oHQEH one? 0:3 mEmCCoSGCH :34. Co #69532 65H. .5 2an 77 2. Division of Labor a. Sex. - By hypothesis [IV. d:2] it was stated That a group of influentials will have significantly more males among them than would have been.the case had the group been drawn at random from the general population of the community. The sex of all persons who received at least one vote as a key influential by all influentials interviewed and of all persons who are 25 years of age and older in the community is given in Table 8 for each of the six communities. Chi-squares computed for each community were: San Diego, 25.53; El Paso, 40. 36; Tucson, 41.89; Las Cruces, 30.40; Juarez, 68.65, and Tijuana, 66.16. Each chi—square is sig- nificant at the probability level of .0005. The hypothesis that influ- entials are drawn proportionally from the males and females in each community is rejected. The hypothesis as stated above is confirmed. The hypothesis was also stated [IV.c: 1] That a significant propor- tion of a given group of influentials will be male. If a significant pro- protion is assumed to be a majority, the hypothesis is also confirmed by data presented in Table 8. The highest percentage of females occurred in Las Cruces in which four of 43 persons who received one vote as a top-ten influential were women. 3. Life Cycle a. Age. - In hypothesis [IV. d:2] it was stated That a given group of influentials will differ significantly in their age characteristics from the general population of the community from which they are drawn. Stated in the form of a working hypothesis this would read Influentials in a community will be proportionally older than the median age of all persons twenty-five years of age and over in the community. 78 Table 8 . Sex of All Persons Who Received at Least One Vote as a Key Influential by All Influentials Interviewed and of All Persons 25 Years of Age and Older in the Community for Each of Six Communitiesf“ Community and All Persons Top Influentials Totals Chi-Square p Sex (n) ((73) (n) (‘51) (n) (X?) San Diego, SMA Males 162,109 49 31 100 162,140 Females 166, 855 51 0 0 166,855 Totals 328,964 100 31 100 328,995 25.53 <.0005 El Paso, SMA Males 48,054 50 46 96 48,100 Females 47,938 50 2 4 47,940 Totals 95, 992 100 48 100 96,040 40.36 '< .0005 Tucson Urban Area Males 12,084 45 41 95 12,125 Females 14,498 55 2 5 14, 500 Totals 26,582 100 43 100 26,625 41.89 <.0005 Las Cruces Urban Area Males 2, 701 49 39 91 2, 740 Females 2,816 51 4 9 2,820 Totals 5,517 100 43 100 5,560 30.40 <.0005 Municipio of Juarez Males 25, 358 47 60 100 25,418 Females 28,435 53 0 0 28,435 Totals 53,793 100 60 100 53,853 68.65 < .0005 Municipio of Tijuana Males 14,368 51 66 100 14,434 Females 14,029 49 0 0 14,029 Totals 28,397 100 66 100 28, 463 66.16 < . 0005 :‘zCensus data is from Table 34, ”1950 United States Census of Population: Texas, California, Arizona, and New Mexico, General Characteristics, " and Cuadro 8, "Septimo Censo General de Poblacion, 6 de Junio de 1950, Estado de Chihuahua and Baja California Territorio Norte. " 79 Relevant data is presented in Table 9 . The chi-squares com- puted for each community were: San Diego, 14. 22; E1Paso, 17. 28; Tucson, 38.16; Las Cruces, 18.66; Juarez, 11.56; and Tijuana, 22.00. Each chi-square is significant at the probability level of . 001. The. null hypothesis that influentials are distributed evenly around the median age of the population of the community is rejected. Therefore, the hypothesis [IV. d:2] is confirmed. Data relavant to the more specific hypothesis, [IV. d:2] That the mean age of a given group of influentials will be 50 years of age or 913:, is presented in Table 10. The number of influentials appearing in various age groups and the mean age of influentials is given for each of six communities. The mean ages were for: San Diego, 55. 38; El Paso, 54.23; Tucson, 58.52; Las Cruces, 49.42; Juarez, 46.40; and Tijuana, 51.64. The decision must be reached as to whether the means are equal to or greater than 50.00. A statistical test is re- quired. The t-test is appropriate.1 A t-test is computed against the null hypothesis that the mean of the universe (estimated by the sample) is equal to or greater than 50. 00. Based on an alpha of . 05, the decision to reject HO will be made if t < -1.708 (for Las Cruces N = 26) or t < -1.711 (for Juarez' N = 25). Since t : —0. 35 for Las Cruces and -1.98 for Juarez, the decision is to reject the null hypothesis for the community of Juarez and not to reject the null hypothesis for Las Cruces.2 lActually, only the means for Juarez and Las Cruces, 46.40 and 49.42, are of significance to us since a test of the hypothesis stated must confirm the statement that all other means are equal to or greater than 50. 00. ZThe t scores for the other communities are also given on Table 10. They do not, of course, reject the null hypothesis. 80 Table 9. Age of Influentials and Age of All Persons 25 Years of Age and Over Dichotomized Near the Median Age for Each Community. 7 Age All Persons 25 Community and Years of Age Chi- Age and Older TOp Influentials Total Square P (n) (92') (n) (‘4) (n) (X2) San Diego SMA (Median 2‘ 42. 7) Older than 164, 356 50 22 88 164,378 42.7 years Youngerthan . 164,424 50 3 12 164,427 42.7 years Txnals 328,780 100 25 100 328,805 14.22 < .0005 El Paso, SMA (Median age = 40. 3 years) Older than 50 25 80 47, 981 40.3 years 47’ 956 Younger than 48,016 50 3 11 48,019 40. 3 years Totals 95,972 100 28 100 96, 000 17.88 < .0005 Tucson Urban Area (Median age = 45. 7 years) Older than 13,296 50 41 98 13, 337 45.7 years lounger than 13, 389 50 1 2 13,390 45.7 years Totals 26,685 100 42 - 100 26,727 38.16 < .0005 Las Cruces Urban Area (Median age = 40. 9 years) Older than 2, 925 50 24 92 Z, 949 40.9 years ‘ Younger than 2,942 50 2 8 2,944 40.9 years Totals 5, 867 100 26 100 5, 893 18. 66 < .0005 Municipio of Juarez (Median age '= 39.8 years) Older than , 7 26 776 50 21 84 ’,797 39.8 years ’ O Youngerthan 26,816 50 4 16 26,820 39.8 years Tinals 53,592 100 25 100 53,617 11.56 <~.0005 Municipio of Tijuana (Median age = 38.8 years) Older than 14,175 50 22 100 14,197 38.8 years Younger than 14,166 50 0 0 14,166 38.8 years Tinals 28,341 100 22 100 28,363 22.00 <:.0005 2’: 'Census data is from Table 33, "1950 United States Census of Population: Texas, California, Arizona, and New Mexico, General Characteristics, " and Cuadro 2— B, "Septimo Censo General de Poblacion, 6 de J_u__nio d__e_ 1950, E____stado n1 -1 _ _ ___.1 n-:.s r: 1' (fivn‘ld anr1fnr1ONOrIe- 81 .mo . H manm 666:3 .234- :64- @064. $8- 36.2- woes- v 3:66.432 3:: we; - mm .o- 8.6+ 3.4+ 3.3 a $8 2.0 main men 3.2 24.2 on 46.3 04.64 44.84 8.? £14m. we .3 64.4 532 an: 2: NM E: 3 on: on on: 2. 2: am on: 3 fleece an m z 3 .2 S n n a f m 2 a. 69.: so are?» 2. am 14 a .2 a as m S m S m 2 m 286» we. 6:4 am. 3 m m n z w E 2 ON a 3 m 6686.» mm 8 as. a.» mm 6 mm m 2. e 3 S. m4 3 E E 688 .8 man?» 3. .3 C... E 3.; x5 3: A5 a: 3 3.: 3 a“: E 1305 mania. nopmdh meoDnO mmiH COmUSH Ommnm Hm owofiQ cam mQSOHO ow< 682:0 6:81.338 pcm .Nmumsh .mmuSuO mmJ .COmoSP .Ommnm mm .owofifl com pom mmsouO amass msozm> cw 38356332: Go nonESZ .0d @388 82 Therefore, for all communities except Juarez the hypothesis [IV. d: l] is confirmed. For the community of Juarez the hypothesis is not confirmed. b. Marital status. - The hypothesis [IV. e:2] was stated such That a group of influentials will differ significantly in their marital status (more will be married) from the marital status of persons in the general population of the community. The corrollary may be stated specifically, [IV. e: 1]Thata majority of influentials will be married rather than single. To test the first hypothesis the influentials are distributed as to the categories married and single. This distribution is then compared with the distribution of the community from which they come. The data is presented in Table 11. Chi-squares computed for each community are for: San Diego, 11. 56; El Paso, 9.96; Tucson, 15.87; Las Cruces, 5.60; Juarez, 14.79; and Tijuana, 20.02. With an alpha of .05 the rejection region (against the null hypothesis that the proportions of the two distributions are the same) includes all X2 _>_ Z. 71. The null hypothe- sis is therefore rejected for each community. The hypothesis [IV. 3:2] is confirmed. Inspection indicates the lowest proportion of married influentials was 93% (in El Paso). This fact0 permits a statement of confirmation of the second, more specific hypothesis [IV. ezl]. 4. Family Social Level a. Father's occupation. - In hypothesis [IV. f: 1] it was stated That a group of influentials in a community will be drawn from a popu- lation a majority of whose fathers were engaged in white-collar occu- pations. To complete a test on this hypothesis it was decided to use census data on each community. The proportions of persons employed in white-collar, non-white-collar, and farm occupations as of 1950 are 83 Table 11. Number of Influentials Who are Married and Single and Number of Male Persons in Total Community Population Over 14 Years of Age Who are Married and Single for Each of Six Communities. ::: Community and All Persons 14 Marital Status”: Years of Age . Chi- and Over Top Influentials Totals Square p (n) (9?) (n) (9'5 (n) (X2) San Diego, SMA Married 145,015 69 26 100 145,041 Single 65,176 31 0 0 65,176 Total 210,191 100 26 100 210,217 11.56 <.0005 El Paso, SMA Married 43, 549 64 26 93 43, 575 Single 24,137 36 2 7 24,139 Totals 67,686 100 28 100 67,714 9.96 <.005 Tucson Urban Area Married 10, 293 65 38 95 10, 331 Single 5, 446 35 2 2 5, 448 Totals 15,739 100 40 100 15,779 15.87 <.0005 Las Cruces Urban Area Married 2, 736 73 23 96 2,759 Single 1,005 27 1 4 1,006 Totals 3,741 100 24 100 3,765 5.60 <.01 Municipio of Juarez Married 24,000 59 23 100 24,023 Single 16,894 41 0 0 16,894 Totals 40,894 100 23 100 40,917 14.79 <.0005 Municipio of Tijuana Married 11,508 52 20 100 11,528 Single 10,894 48 0 0 10, 552 Totals 22,060 100 20 100 22,080 20.02 <.0005 M :izCensus data is from Table 34, "1950 United States Census of Population: Texas, California, Arizona, and New Mexico, General Characteristics, " and Cuadro 15, "Septimo Censo General de Poblacion, 6 de Junio de 1950, Estado Chihuahua and Baja California Territorio Norte. " “The United States Classification of marital status includes "Males, 14 years old and over: Single, Married." The Mexican census has two relevant classifi- cations: "Heads of Family and Persons married or Living in Free Union. " The Mexican census does not, however, divide single persons as to sex and the classifi- cations include all age categories. Consequently, all persons under 14 years of age were eliminated from the totals of male population once the ”Married" cate- gory had been extracted. The remaining number was taken to be the "Single” class. This difficulty accounts in part for the larger proportions in this class in 84 to be compared with the distribution of influentials' fathers occupations.1 The data are presented in Table 12. Chi-square computed for each.community is for: San Diego, 26.17; El Paso, 20.02; Tucson, 7.77; Las Cruces, 34.49; Juarez, 57.57; and Tijuana, 39. 10. Each chi-square is significant at the probability level of . 005. Thus the null hypothesis that the influentials' fathers' occu- pations are distributed as the occupations of the community are dis- tributed as to white-collar or non-white collar is rejected for each of the six communities. By inference it is possible to state that since fathers' occupations were significantly white—collar in number (from a low of 58% in Las Cruces to as high as 87% in Juarez), influentials will tend to be sons of fathers who have been so employed. The hypothesis as stated above is therefore confirmed. 5. Ethnic Status a. Religious affiliation. - In hypothesis [IV. g: 1] it was stated That a group of influentials in a community in the United States will have religious affiliations characteristic of those highly valued in the general society, viz. , protestant, regardless of the proportions of non-protestant affiliations in the community. Data on the religious affiliations of influentials from the four communities in the United States is shown in Table 13. A chi-square 1This is actually a very conservative measure of the relationship. It is known that the proportion of persons engaged in agriculture has changed from 32.4% in 1910 to 12. 192' in 1950. Those engaged in White- collar occupations have changed from 11ftt in 1910 to 1792 in 1950 while non-white collar occupations have changed from 57972. in 1910 to 7092. in 1950. Had comparisons been made with an earlier distribution, the differences in the proportions would have been even more extreme. It may also be argued that no specific census year would be an appro- priate measure since there is a variation in age of influentials. (Data is selected from Loomis, 1960: 87). 85 Table 12. Number of Influentials Whose Fathers’ Occupations were White Collar, Non-White Collar, or Farm and Number of Persons in the Male Experienced Civilian Labor Force in White Collar, Non-White Collar or Farm Occupations in 1950 for Each of Six Communities. :1: Occupational Total Male Category 30': Experienced Top Influentials' Chi- and Community Labor Force Father's Occupa. Totals Square p (n) we (n) (%) (n) (x2) San Diego, SMA White Collar 45, 366 39 21 81 45,387 Non—White Collar 63,611 54 1 4 63,612 Farm 8,530 7 4 15 8,534 Total 117,507 100 26 100 117,533 26.17 < .0005 El Paso, SMA White Collar 14,895 37 22 76 14,917 Non-White Collar 22, 523 55 4 14 22, 527 Farm 3,290 8 3 10 3,293 Total 40, 708 100 29 100 40, 737 20. 02 < . 0005 Tucson Urban Area White Collar 4, 566 48 29 69 4, 595 Non-White Collar 4, 920 51 9 21 4, 929 Farm 90 1 4 10 94 Total 9,576 100 42 100 9,618 7.77 < .005 Las Cruces Urban Area White Collar 1,053 39 15 58 1,068 Non-White Collar 1, 539 57 5 19 1, 544 Farm 104 4 6 23 110 Total 2, 696 100 26 100 2, 722 34.49 < . 0005 Municipio of Juarez White Collar 9, 099 22 20 87 9,119 Non-White Collar 25, 700 61 2 9 25, 702 Farm 7,155 17 1 4 7,156 Totals 41,954 100 23 100 41,977 57.57 < .0005 Municipio of Tijuana White Collar 5, 426 25 19 86 5, 445 Non-White Collar 11, 452 53 1 5 11, 453 Farm 4,696 22 2 9 4,698 Total 21, 574 100 22 100 21,596 39.10 < .0005 "‘Census data from Table 35, ”1950 United States Census of Population: California, Texas, New Mexico, Arizona, General Characteristics, " and Cuadro 11, "Septimo Censo General de Poblacion, 6 de Junio de 1950, Estado de Chihuahua and Baja California Territorio-Norte. " — — — 86 Table 12 - Continued ~~~~~~ WWhite Collar, Non-White Collar and Farm include the following occupational categories in the United States and Mexican census' . UNITED STATES MEXICAN White Collar White Collar Managers, Officials, and Pro- Directing personnel except in agri- prietors, except farm culture, cattle, forestry and fishing Professional, technical, and Professionals and technicians in kindred workers every branch of activity Clerical and kindred workers Sellers in every branch of activity Sales workers Non-White Collar Non-White Collar Craftsmen, foremen, and Skilled workers in every branch kindred workers of activity Operatives and kindred workers Laborers, artisans (craftsmen) and Private household workers day laborers in extractive industries Service workers, except private Laborers, artisans and day laborers household in processes of the production of Laborers, except farm and mine property and services: Directors and those driving vehicles and not directors Occupations (those employed), with compensation, who borrow service personnel in homes or in institu- tions or management of service personnel, recreation or social. Farm Farm Farmers and farm managers Occupations (those employed) in Farm laborers, unpaid family agriculture, cattle, forestry and workers fishing including directing personnel Farm laborers, exc. unpaid, and farm foremen 87 fly: on: ow on: :u on: om OOH hm mama—0H. ON ma w om w mm N. w A cofimflflwm OZ m w A n m m a o o 30h m: NA 4. ON w ON 0 w a oflosumU :mEom ow mo N; em NM. mm A: mo mN pnequHOHnw A5 3% A5 Q: A5 3% A5 3% EV HmpoH meoduO med cemoDH Ommnm Hm owefiQ cam coflmoflfimmmfiu segue: msozmzom mmumum $0”?ch 9.3 C“ mofiflCSEEOO .Hsoh new Coflmfiflmqfl 02 >9 pew mcoflmfifimmmflu coflmfimwd. mdowwfldm msoimsr >3 mfimflcgzwca mo Hefiadz . 3 3nt L8 88 was computed on the proportions of persons affiliated as Protestant or Non-Protestant among the four communities. That chi-square was 13. 19. It is significant at the . 01 level; a fact which leads the investi- gator to the conclusion that the communities are indeed different from one another as to the proportions of persons committed or not to Protestantism. As to the proportions of persons involved in each community a further step is necessary. This involves a test of a hypothesis which is stated as follows: that influentials are drawn from a population (or universe) which is either Protestant or non-Protestant with equal probability. The proper test for this hypothesis is the binomial.1 Let us assume that non-Protestant includes _o_n_l_y_ those persons who state an affiliation with Roman Catholicism or Judaism. The decision toreject a null hypothesis (of equality in proportions) will be based on an alpha level of .05. That is, the null hypothesis will be rejected if p _<_ . 05. The probabilities associated with the occurrence of the frequencies specified are for: San Diego, <.000; El Paso, <.047; Tucson, <.041; and Las Cruces, <.008. Thus the statement is accepted that among those influentials who stated a religious preference of Protestant, Roman Catholic, or Jew the number of Protestants was significantly greater than would have occurred by chance. But, it may justifiably be asked, what about those persons who declared they had no affiliation? It is proper to submit this question to a test also. Thus the non-Protestant category should be redefined to include Roman Catholic, Jew, and No Affiliation. Under a similar test and the same null hypothesis (same the redefinition of non- Protestant) the probabilities associated with the occurrence of the 1The null hypothesis is symbolized as HO: p1 = p; = . 5 where p, = Protestant and p2 2' non—Protestant. The alternative hypothesis is H? p1 > P2- 89 frequencies specified are for: San Diego, <.000; El Paso, <.429; Tucson, <. 378; and Las Cruces, <.081. Thus the statement that among all alternative types of religious commitment or non-commitment the number of influentials that were Protestant in affiliation is significant for only the community of San Diego. Before commenting further on the meaning of these findings the data on the problem of religious affiliation in Mexican communities should be confronted. The hypothesis [IV. g:2] has been stated That} group of influentials in a Mexican community will have religious affilia- fi tions characteristic of the dominant religion of Mexico, viz. , Roman Catholicism. Table 14 presents the data on the religious affiliations of influ- entials from the two communities in Mexico. A chi-square was com- puted on the proportions of persons affiliated as Roman Catholic or Not Roman Catholic between the two communities. That chi-square was 9.82. It is significant at the . 005 level, a fact which leads to the conclusion that the communities are indeed different from one another as to the proportions of persons committed or not to Roman Catholicism. Again the question was asked, are the commitments to Roman Catholicism, as against the commitments to Protestantism and Judaism, proportionately more than one would have expected by chance? The binomial test indicates that the probabilities associated with the occurrence of the frequencies Specified are for Juarez, <.000 and for Tijuana, <.008. ‘ However, when Not Roman Catholic is defined to include No Affiliation a strikingly different result occurs. The probabilities associated with the occurrence of the frequencies thus specified are for C. Juarez, <.005 (still within the rejection region indicating dominance of Roman Catholicism as a commitment), and for Tijuana, <. 974. The latter level of probability occurs because of the comparison 90 Table 14. Number of Influentials by Various Religious Affiliation Classifications and by No Affiliation for Two Communities in Mexico. Religious Affiliation Juarez Tijuana Total Classification (n) (%) (n) (if) (n) Protestant 0 00 0 00 0 Roman Catholic 18 78 7 32 25 Jew 1 4 0 0 1 No Affiliation 4 17 15 68 19 Totals 23 100 22 100 45 91 involved. Seven influentials in Tijuana claim affiliation with the Roman Catholic church. Fifteen, on the other hand, claim no affili- ation at all. The number "seven" is significant; but it is significant because it is oso small and not because it is so large. The fact of the "dominance" of the religious affiliations in the directions hypothesized does not fully explain the variations noted when No Affiliation is considered. Unfortunately, no data were gathered which could attest to the significance of the affiliation which an individual had with his church.1 For the present, it will be held that the hypotheses are ”supported” albeit with certain reservations. b. Ethnic background. - It was held that (hypothesis [IV.h:2]) Representation of an ethnic "minority" among the general influentials will be less than proportional to the number of that minority in the total population of the community. The data of persons with Spanish surnames are taken from a special census report and are shown in Table 15.2 The Chi-squares lOne further bit of information is enticing which relates to an earlier discussion of the meaning of religious affiliations (supra). In the community of El Paso it was discovered that 11 of the 48 influentials frequently named were members of one Presbyterian Church. And in addition four other influentials were either "nominal" members or had a relative in the church--a wife, son-in-law, and son. Thus 15 (or 31%) had direct connection with one another by way of this membership. 2The differences in percentages of Spanish surname population within communities‘ between Table l and Table 15 are due to the re- striction of the latter to those persons 25 years of age and over. It was decided to restrict these comparisons to those parts of Tables 8 and 34 in the United States Census reports (identified in Table 15) which were based on a sample of total returns. Under this condition persons 25 years of age and over are listed directly and do not require error-laden abstraction from other distributions such as age. All persons--male and female—~were included in Table 15. This was done to permit use of the statistical sample of the population for both categories, Anglo and Spanish surname. Table 8 of the Census does not give a statistical 92 computed for each community are for: San Diego, 1. 370; El Paso, 13.715; Tucson, 2.475, and for Las Cruces, 7.898. With alpha at .05 the null hypothesis is to be rejected for cases in which X‘2 > 2.71. Therefore, the statistical hypothesis is rejected for two communities: El Paso and Las Cruces. In both of these communities, the hypothesis, as stated above, is held to be confirmed.1 The rejection of the hypothesis under consideration is thus im- proper for the two communities, San Diego and Tucson, until a finer measure of the “intent" in the restriction of ethnic groups is devised. breakdown by sex. Besides, simply using males (or females) would not change the proportions of Anglos and Spanish surnamed persons since sexes are evenly distributed. (A Chi-square was computed on males only, 25 years of age and over, for El Paso with the result that the chi-square is larger (and therefore more stringent) and that no difference results in the decision to reject or not reject the statistical hypothesis.) The ”Anglo" classification might more properly be called the ”Non Spanish surname" population since it includes the Total popu- lation minus the Spanish surname persons. Finally, a note on the reliability of the Census classification, Spanish surname, is appropriate. In an Albuquerque survey, the 1950 census enumeration district results were systematically compared with results from more detailed report- ing methods and interviews (Winnie, 1960). The conclusion of this study indicated "that no single criterion can be selected as 'best' for identify- ing Hispanos in population statistics in general or in censuses in particular. . . . [how-ever, ] the surname criterion seems to offer as good a general-purpose classification as any from the technical standpoint. " Its disadvantage is that it understates the size of the population (appar- ently by about ten percent) (ibid.: 365-366). 1In the case of Tucson and San Diego, a cautious comment is required. Tucson has among its total population of persons 25 years of age and over 18% who have Spanish surnames. Ten percent of the influential in Tucson are persons of Mexican descent. The fact that these proportions are not statistically different does not obviate the fact that the proportion of influentials of Mexican descent is not equal to or greater than the proportion of their numbers in the total popu- lation. Likewise for San Diego. Here only four percent of the specified portion of the population have Spanish surnames. Yet none of the influentials are of Mexican descent. 93 Table 15. Ethnicity of Influentials and Number of All Persons With Spanish Surnames 25 Years of Age and Over for Each United States Community: San Diego, El Paso, Tucson, and Las Cruces. All Persons Community and 25 Years Old Chi- Ethnic Categoryb and Older '1 Op Influentials Totals square p (n) ("7a) (11) (if) (n) (X2) San Diego, SMA Anglo 315, 947 96 28 100 315, 975 Spanish- surname 13, 020 4 0 0 13, 020 Totals 328,967 100 28 100 328,995 1.37 <. 15 El Paso, SMA Anglo 58,417 61 43 96 58,460 Spanish-surname 37, 578 39 2 4 37, 580 Totals 95,995 100 45 100 96,040 13.72 <.0005 Tucson Urban Place Anglo 21, 742 82 38 90 21, 780 Spanish—surname 4, 841 18 4 10 4, 845 Totals 26, 583 100 42 100 26, 625 2.48 <. 10 Las Cruces Urban Place Anglo 3,148 57 22 85 3,170 Spanish-surname 2,386 43 4 15 2, 390 Totals 5, 534 100 26 100 5, 560 7.898 <.005 aCensus data from Table 8, ”1950 United States Census of Pepulation: Persons of Spanish Surname, " and Table 34, "1950 United States Census of Population: Texas, California, Arizona, and New Mexico, General Character- istics. " "Anglo" category includes all persons over 25 years of age not classified as "Spanish-surname. " 94 6 . Localism a. Birthplace and length of residence in the community. - The hypothesis [IV. i: 1] was stated That influentials will have varying characteristics as to place of birth. (See supra for discussion of rationale.) Data on the place of birth of influentials interviewed in each of the six communities are presented in Table 16 . As was hypothesized, the communities divide or vary on the basis of this variable. However, the form of the variance is interesting. San Diego and El Paso had a higher percentage of their influentials who were natives of the com- munity than any of the others. Notably, Tucson and Las Cruces had the highest proportion of influentials who had not been born in the state within which the community was located, 69 and 58 percent respectively. Finally, the two Mexican communities showed a larger proportion of influentials who were not natives of the city and who were not out-of- state migrants but who were natives of the state in which the community is located.1 The three results were found to include internally similar communities, but each pair was different from any other pair in com- binations involving their central tendency. It is possible that more information concerning the history of each community would suggest reasons for the specific differences which are noted here. For our purposes, however, it may be concluded that the hypothesis as stated above was confirmed. The other hypothesis stated in this section was [IV. i:2] That a majority of influentials will tend to have lived in the community for 25 years or more. lChi-squares were computed on the categories grouped as "City- Not City” and "State-Not State" with the result that the pairs were not different. The Chi-square results for the various combinations or communities and categories were: 95 N: 02 MN OOH mN oofl cm. do me cod 3 on: 0N H308 N. mu m. o o m N N H o o m N .2330 v o M m N o o o o o o o o 9.7er ”chem Cwfiouoh oo o o o 0 mm 2 do on NM on ow NE enmpmnHOuudO me it S N» 3 ma v N. m. A: m m N Baum mo @352 3. m 2 cm m 2 m S. o 3 8 mm 2 35 mo 2:82 A7: wt: 3 .on A5 33 A5 E A5 32 A5 33 A5 130% 9338 nenmsh. meoan med semesrfl Ommnw mm owefiQ 9mm afflm mo momfinm ©3630 .mmfiwcsgaoo Xfim mo Loam a: mHNSCeDCcH mo 23.3mm mo momfinm . A: «3an 96 Data for each of the six communities on the number of years in which the influentials interviewed had resided in the community are given in Table 17. The mean number of years residence is also given. The results were: for San Diego the mean number of years residence was 39.88; for El Paso, 40. 32; Tucson, 33.67; Las Cruces, 25.46; Juarez, 26.00; and Tijuana, 28.14. Since none of the means are less than 25 years, the hypothesis as stated above is confirmed.1 b. Orientation toward the community. - The hypothesis was stated [IV. jzl] That a majority of influentials will show orientation to the com- munity as against not showing orientation to the community. The index of orientation toward the community will be taken as the respondent's answer to the question (number 5, Appendix 1) ”Under what conditions, if any, would you consider leaving this community?" Data on orientation toward the community are presented in Table 18. In only one community--Las Cruces--were as many as one-third of the influentials willing to state a willingness to leave. In general, this is interpreted as confirmation of the hypothesis stated above. A slightly more stringent test may be completed by submitting these frequencies to a binomial test. In this test the null hypothesis is Category Community Comparisons City-Not City State-Not State San Diego by (X) El Paso . 99 2. 70 Tucson X Las Cruces . 05 . 27 Juarez X Tijuana 2.51 1.09 San Diego and El Paso X Tucson and Las Cruces .89 9.22 San Diego and El Paso X Juarez and Tijuana ‘ 13.05 10.80 Tucson and Las Cruces X Juarez and Tijuana 1.18 35.52 (Rejection region for alpha 2 .05 with one d.f. all X2 > 3.84.) 1A t-test was computed on these data. The. t's were for San Diego, +3.65; El Paso, +4.91; Tucson, +3.44; Las Cruces, +0.16; Juarez, +0. 28; and Tijuana, +1. 25. For relevant rejection regions see Table 12. None of the t-scores are smaller than those necessary to reject Ho' 97 mNJ+ wN.o+ c~.o+ ww.m+ :84: mo.m+ u 8.: 40.: 3.3 3;: ~92 Show. on E .wN 8.8 34K $5M 3.04 3.3 28> 582 m: ooh, MN oi fl 2: fl. 9: 3, oo 2 2: on axon. S o o E. a S m m a o N m: 4 $2 .8 38> 2 Om om w om m mm o m: w m H m N . wheel» om 0» 3 a o o w M w M. m. n o m 4 2 2a.; 4N 8 Z 4: E .2 we 2 Om 2 E Om am on C E 9.8an 38> 3 A5 $3 A5 WC A5 S3 A5 33 A5 33 A5 Aw; A5 H308 acmsflh nonmsh. meosuu mmA .COmUSH. Omenm Hm owefifl new KSEDEEOO CM @3630 mocepfimem mo newcmd .MCMSPB new .ueumdh .UO .meoan mnA .COmosrfi 6me HM .owefiQ 5mm >3 nude? mo mMeQESZ mdoCm> new 3358850 one 5 mocepfimem pm: 95$ 05$ mHmEGeSGCH we neJESZ .m~ 3an So. :0. $0. ooo. So. moo. a pm: 002 ON on: f cod mN ooz ow 002 cm oofi mm. 330B om; mm 5 2w 2 we : mm mm on .N ow on 0593 on wcflfiwfi “OZ Om 2 m m: m mm m m N Om 0 on m e.,/m3 on mmmcucflzg EV Q: Ev Aw; E 3.: A5 3: Ev 33 A5 :3 A5 390E... 955.39 Nonmsh. meoDnO mme cemosfi Omenm Hm omefifl cam coflmucofluO pmpdfiU mo xech .meEflCSEEoU xfim mo Summ n0w 3.2258800 ext 92va o» mmocmcflfig mfimflcodficH co pemmm KSEDEEOO op coflmuceCO we xech .m: 3an. 99 set up which states that the probability of Willing to Leave and Not Willing to Leave will occur with equal frequency. Results were that the probability of these frequencies would occur for San Diego with a chance < .002; for El Paso, <.022; Tucson, <.000; Las Cruces, <.054; Juarez, <. 011; and Tijuana, <.001. With the exception of Las Cruces, this would be accepted as significant at an alpha level of . 05. c. Tenure in firm. - The final hypothesis to be tested was stated in the form [IV. k: 1] That a majority of influentials will have had over 20 years experience with the firm with which they are employed. Data on the number of years the respondent had worked with his present firm is given in Table 19. The mean number of years of employment with the firm were for San Diego, 29.04; El Paso, 23. 23; Tucson, 24.02; Las Cruces, 13.73; Juarez, 14.40; and Tijuana, 16. 33. A t-test is the appropriate test of the significance of these results. The t scores for each community were for: San Diego, +2.87; El Paso, +1.25; Tucson, +1.87; Las Cruces, -3.32; Juarez, -2.45; and Tijuana, - 1. 47. The region of rejection where alpha is . 05 will include all t scores less than a specified number. The region of rejection is shown on Table 19. The result is to reject the null hypothesis that the universe mean (of which the sample is an estimate) is 20 years for the communities of Las Cruces and Juarez. The variability of these findings do not permit confirmation of the hypothe sis as stated . C. Summary of Chapter Findings A simple tabular summary of the state of the hypotheses will be used. The first column will list the hypotheses which were of interest as elements or parts of the Profile of Influence Potential. Each com- munity will follow in a column form and an "X" will indicate that the 100 857 2:4- 857 88.7 88.7 woes- a: cofimen :ofiuemem $4.. SiN- N88 5.2+ 8.: >w.N+ a 3.: 3.: 8.8 88.2 N13 8;: .o.m 3.8. 2...: 8.2 No.8 NN.NN 8.8 882 t: 88 2N 02 mN OS 8N 02 N4 :: ON 2: 8N 888.2. ON 3 m wN > MN 8 m N 2 N m N 82 8 88> m 8 2 N. f a >N > > N 2 8 m N . 88> 2 8 8 4m 2 4 Nm w 2 m «N Nu m: 8 2 m 88> ON 8 2 m» mm N. 4N 8 S w 3 mN am 2 m8 2 808 .8 88> TN 3 cs 3 3.: E :3 3 § 3 S; E 3: c: amuofi mcmdflflrfi nenmdh 85an main COmosrfi Omanm hm owefifl Gem Shah 4.33, 93:58 ©3380 mummy? mo Hengsz . m 03838800 82m mo £08m new mHQSCvSGCH >0. an?“ Haemehnm e33, wcflxpoa whee? mo .85552 A: 95MB lOl hypothesis was confirmed, an ”(X)" will indicate substantial confirm- ation. A final column will indicate, by the same sign, that the hypothesis was confirmed for all six communities. San E1 Las Ciudad All Hypotheses1 Diego Paso Tucson Cruces Juarez Tijuana Communities [IV. a:Z] X X X X X X X [IV. bz3] X X X X X X X [IV. c:2] X X X X X X X [IV. d:2] X X X X X X X [IV. e:2] X X X X X X X [IV.h:Z] X X - - [IV. a: l] X X X X X X X [IV. C21] X X X X X X X [IV. d: l] X X X X X [IV. e: l] X X X X X X X [IV. le] X X X X X X X [IV. i: l] X X X X X X (X) [IV. i:2] X X X X X X X [IV.j: l] X X X. X X [IV. k: l] X X X X [IV. b: l] X X X (X) - - (X) [IV. b: 2] - - - - (X) (X) (X) [IV. g: l] (X) (X) (X) (X) - - (X) [IV. g:2] - - - - (X) (X) (X) [IV. h: l] (X) X (X) X — - (X) To this point there has been no attempt to differentiate two very different reference points from which and to which inferences can be made. In summarizing these results this shall be done. The first reference point is "intra-community. ” The question is, do these groups of influentials actually differ in selected character- istics from the general community from which they are drawn? 1The reader is referred to Chapter IV, Section C (pages 46 and 47) for the summary of all the hypotheses discussed in this section. The symbols remain constant with that group. (See page 73 for pre- sentation of [IV. bz3]. lOZ Whether the fact is assumed or is known, communities vary widely as to age distributions, religious affiliations, etc. Do these differences affect the characteristics of influentials ? The second reference point is "intra-influential. " That is, are there generalized characteristics which one could expect of these persons labeled as "influential" no matter what the characteristics are of the community within which their relevant social relations take place? Does the influential have a specified age characteristic, for example? This ”intra-influential" reference point has been derived from the specific empirical generalizations made from other studies of influentials. Also, generalized national differences have been accounted for in some instances. It was felt that a foundation could be laid for expecting differences between communities within different national circumstances on certain problems. Three problems included expectations which were specified for the different national contexts in which the community was laid: occupation, religious affiliation, and ethnic identity. 1. The Intra-community Problems The following hypotheses were of this type: [IV. a:2], [IV. bz3], [IV. c22], [IV. d:2], [IV. e:2], and [IV.h:Z] dealing with education, occupational structure, sex, age, marital status, and ethnic groups respectively. The results were that for all problems except ethnic group representation, the hypotheses were confirmed. That is, for each community, the influential groups were significantly different from the community population. The exception was the ethnic group representation problem wherein the trend was established that Spanish surnamed persons were underrepresented among the influentials in most communities . 103 Z. The Intra-influential Problem The following hypotheses were built upon a generalized reference to influentials in western, capitalistic societies. They were [IV. a: 1], [IV. c: 1], [IV. d: l ], [IV. e: 1], [IV. f: 1], [IV. i: 1], [IV. i:2], [IV. j: l], and [IV.k: 1]. These were concerned with educational level, maleness, age level, marital status, father's occupation, the place of birth, length of residence in the community, orientation to the community, and the tenure of the influential in the firm of employment. With the exception of age, orientation to the community and tenure in the firm, the hypotheses were confirmed for all communities. (Place of birth of the influential was expected to vary. It was held that this was so, but some unexplained patterns developed suggesting certain similari- ties between pairs of cities.) The age level specified for influentials did not hold for the smallest city in the study, Las Cruces. Orientation to the community, likewise, did not hold for the smallest city. For all other communities, these elements held as specified. The Specified length of tenure in the firm did not hold for both Las Cruces and Juarez, and, in addition, showed wide variation for the remainder of the c ommunitie s . Finally, there were five hypotheses which had specified outcomes to be different across societal boundaries. These were [IV.bzl] and [IV. b:Z]—-expecting different institutional areas to be predominant in the occupational categories within which influentials were employed, [IV. g: l] and [IV. g:2]--eXpecting different configurations of religious affiliation, and [IV. h: l]--expecting a generalized ethnic dominance of Anglos over persons of Mexican descent in the United States communi- ties. Only the last was held to be confirmed with little qualification. The "business" institutions were generally dominant in all communities regardless of societal foundation. The professions were also repre- sented in relatively large numbers. The qualification necessary on 104 these findings are a result of (l) the variation in these representations in the different communities, and (2) the lesser place that governmental status-roles held in the Mexican communities, proportionally. The question of religious affiliation was upheld so long as the reference point included flspecified commitments to religious orientations. However, when the qualification was made of allowing for the alternative of No Affiliation with a religious group, the thesis was upheld only for San Diego and Juarez. In short, the Profile of Influence potential was maintained for the characteristics of education, sex, marital status, father's occu— pation, and residence in the community. Other characteristics which were variable by the community situation were age, ethnic background, religion, and occupation. CHA PT ER VIII SU MMARY AND CONC LUSIONS The problem to which the discussion and research in this thesis has been directed was (I) to derive from selected social system theory certain propositions concerning the interrelation of various foci of evaluations one of which was influence, (.2) to interrelate these proposi- tions with (a) statements derived from the specific criticisms of the findings of a methodology called the reputational technique, and with (b) empirical generalizations concerning certain indices of differential evaluation. These empirical generalizations are taken from the research literature. A corollary problem is to test the reliability of the reputational technique as a method for the determination of influenc e . A. The Propositions Derived from Theory From the selected theory it was possible to deduce propositions with regard to the expected interdependence of certain foci of evalu- ation. For example, Positions in class, power, and prestige hierarchies contribute to the potential for interpersonal influence. The statement implies that three general standards of evaluation of actors are interrelated. Thus, positions on a class hierarchy are related to positions on a prestige hierarchy both of which are related to positions on a power hierarchy. Hierarchy refers to a graded order of things where, in this case, the grade is high or low evaluation. 105 106 It was also stated That ranking may be dependent upon evaluations which range from particularistic characteristics to universalistic character- istics such as competence or skill. Extreme differences in the standards of evaluation may be attributed to a relational situation by the observer. The standard upon which he is drawing may vary from the limited relational system to a very broad, generalized base. It is possible that both of these extremes are operating for the individual evaluating others within the relational system. Thus the range in phenomena which are the foundation of differential evaluation are great. Finally, it was suggested that power is maximized (in the status- role of a given individual) under certain conditions. These conditions were: (1) when there is a high evaluation on the foci of generalized phenomena such as class and prestige, (2) when there is high evaluation of a history of achievement, and (3) when there is high evaluation of certain possessions such as skill or competence or wealth. In short, some degree of intercorrelation of evaluative hierarchies one of which is influence is to be expected. B. The Criticism of the Methodology Moving now to the assertions of the critics of the reputational technique, it is possible to state their objections in the framework of these expectations just presented. It will be remembered that the critics have suggested that the reputational technique does not measure influence. Instead, it is said to measure some other phenomena. What are the alternative phenomena which it measures? These are the elements "status, " “formal leadership, " and "personality traits” indicating long activity in the community. 107 In addition, the critics suggest that the "scope" of influence is narrow and not general. That is, they assume that influence is determined by the relative skill or interest (presumably learned or proved over time) which the individual exhibits in particular, sub- stantive areas. The criteria which specify these substantive areas is not given. By inferences from the areas chosen in the research of the critics, one might assume that examples of these areas are ”public education, " "urban renewal, " or ”politics. " It is rather difficult to imagine the skill boundaries for issues such as these without substantial exploratory research. The bases of relative skill have not been spelled out by the critics. Let us reason from two assumptions drawn from this work. The first is that influence varies by the skill or interest of the influential. And, therefore, individuals will be prominent in those issues which confront the community at a particular time that happen to conform with whatever of their skills or interest. The second is that the respondent who is evaluating relative influence in others is basing his evaluations not on influence but on a multitude of other factors. The conclusion follows that the same group of individuals would not be named as influentials over time. When the propositions derived above are combined with these arguments it may be said (1) that contrary to the critics the high evaluation of influence should correlate with high evaluation of prestige (which is what is assumed to be equal to their "status"), of skill, and of ”experience. ” In addition, two arguments must be developed further. (2) The question of "general" vs. ”limited" scopes of influence must be tested independently, and (3) the longevity of influentials should be confronted directly. Let us now look at the possible outcomes of their theses on these latter points. 108 If the "general" vs. ”limited" influence argument is accepted as it was stated above--i. e. , that influence is a matter of skill or interest and that each occupant of an influential status-role is limited to specific substantive areas--the statement is not different from that derived from theory, viz., that influence is related to skill. The result would, however, limit the usefulness of the reputational tech- nique since the technique would be masking more accurate indicators of influence or at least more specific indicators. If, however, the argument is in fact rejected--i. e. , that general influence is not to be distinguished from limited influence--substantia1 reliability may be placed on the reputational technique as an efficient indicator of influence. In addition, the factor of skill as related to influence is not denied nor is it held to be an exclusive indicator of influence. On the question of alternative phenomena as events independent from influence evaluation, more will be commented below. Finally, the longevity question is concerned with the problem of whether the same incumbents of status-roles which are assessed as influential appear over time. If different incumbents appear over time, serious reservations would need to be placed on the technique. This result would indicate that the evaluators of influence were judging the status of influence independently and that the status of influence is likely to be separate (independent) from other statuses which the influential individual occupies. On the other hand, if substantially the same incumbents are reported as influentials over time, the reliability of the reputational technique is well established. It would be inferred that influence is but one focus of a series of evaluations. The findings of this research on these matters were that "general" influence cannot be distinguished from "limited" influence. General influence was operationalized by a question which did not specify is sue- areas to which the influence was restricted. Limited influence was 109 determined by naming only persons influential on a specific hospital project in the community. It has also been shown that the reputational technique, when determined at two time periods separated by four years, reports substantially the same incumbents in the influential status-roles assessed. No more difference between incumbents in the two time periods was found than one would have expected from retirements and deaths in the interim. There is one exception to this statement. The exception involves those status-roles which are primarily authoritative in form. In the context of community-wide activity, these authoritative status-roles are primarily elective political offices. For these status-roles the relative influence seemed to be independent of the incumbents of the positions. The loss or gain of political office largely meant the loss or gain of influence. The fact that even in these situations the loss may not be ”total, " attests to a multiple foundation for influence. In general, then, respondents who are asked to evaluate others as to influence tend to report the same incumbents ofinfluential status-roles over time. The interpretation given here is that these findings largely refute the arguments presented by the critics of the reputational technique as outlined here. There is: however, one further problem. What if the component being evaluated is not influence but some other phenomena? Would the same consistency over time and over subject area permit the above conclusion ? C. The Elements of a Potential for Influence Complex To confront the question just stated the set of variables dealing with alternative evaluative hierarchies must be presented. It will be remembered that the selected theorists supported the generalization 110 that certain parallels would be exPected between various evaluative hierarchies. The critics of the findings in the area of community power have suggested that the measure used here finds nothing but some alternative evaluative bases, i. e. , bases other than influence. Partly to meet these objections and test the propositions and partly to extend research generalizations concerning the influential person in the community, an extensive review of the literature was used to develop a series of empirical generalizations. These generali- zations were ordered on a model which itself was generalized in such a way that it fit within the concept areas in which we have been interested. The specific components or elements which were sum- marized were justified in part from findings of past research and in part because of dictates from the propositions derived from theory. The proliferation of research hypotheses develop from this base. There are, however, two questions which must be directed to as many of these elements or indices of alternative foci of evaluation as possible. These two questions are: is the ”sample" of influentials in a, given community different from that which could have been drawn at random from the community as to particular elements? and, is the characteristic of the influential such that a "universe“ of influ- entials--regardless of community variation-—might be a reality? Once the first question has been answered--viz. , that the in- fluentials are different from the general community--it may be assumed that for that element some independent selection is entering into the choice of these individuals. Then the second question is permitted wherein it is asked if differences overlap community boundaries. The reference for evaluation must be reiterated at this point. The influence question asked the evaluators to assess the status of influence by naming incumbents of status-roles which were influential. There must be an assumption that these persons are knowledgeable as to the nature of influence. The personal characteristics of the influ- entials are not evaluated by the observers. The individual character- istics are therefore indices of more general factors which have been shown in other studies to be highly evaluated. The detailed findings have just been summarized in Chapter VII. The general finding was that a given group of influentials was indeed different from the population of the community in which their action took place. Thus the group is selected from the population of that community according to specified criteria. There is sufficient simi- larity within communities to argue that this selection is generally substantiated. In addition, when these samples are compared to expected results which are based on a wide range of other communities, very little difference was found. Therefore, statements such as these resulted: influentials tend to be different from the population of the community in which they reside as to their age distribution and, more specifically, they tend to have a mean age of at least 50 years. The result over a series of elements is a remarkable consistency over the boundaries of the various communities. This is all the more important since the communities themselves differed widely. Their populations ranged from 25, 000 to one-half million. The size of the ”minority" group dominant in the Southwestern United States-—persons of Mexican descent--varied from less than five percent of the total population to over half of the population. The com- munities were culturally separated by political allegiance to two dif- ferent countries--the United States and Mexico. (This latter point may be of lesser importance since the two Mexican communities are con- tiguous to the United States boundary.) The conclusion is that elements such as age, sex, occupation, father's occupation, ethnic background, religion, education and marital status are in total or in parts indices of evaluative hierarchies other than influence and that there is substantial interdependence among these hierarchies. The specific objection that "status" is the element measured by the reputational technique is substantiated (not rejected) in the sense that individuals named tended to be high on indices of social level. It seems unlikely however, that "status" or prestige is the gill}; phenomena being assessed by the general influence measure since nearly all of the elements were related in the expected direction. A view much more reasonable is that the interrelation of these indices of evaluative hierarchies with the evaluation of influence is true.1 It has been noted that indications of a potential for influence are life experiences such as those implied by an advanced state of edu- cation, age, and skill. Likewise, generalized life styles indicated by social level, ethnic status, religion would seem to be among the items which create a condition permitting or giving potential to influence. D. Evaluation Problems arise in research which cannot be adequately handled within the limits of time and ‘budgets available. Ideally, for example, 1Appendix III reports the results of an extention of this argument wherein not only influence but relative influence within the influential group is compared to the possession of the specified characteristics. The results are inconclusive in that only two communities show a very high positive correlation between the two indexes while the other four communities showed no correlation at all. An analysis of the items involved in the multiple social characteristics index reveals that the four communities which showed no difference were substantially homo- geneous on these items and that the items did not, therefore, adequately discriminate between relative influence and relative score on the multiple social characteristics index. As has been shown, the individual items did discriminate between influentials and non-influentials. The data in Appendix III do not affect the conclusions of this thesis. a class, prestige, and possibly authority structure should have been determined for each community so that each influential could have been accurately placed therein. So doing would not only permit more accurate placement of influentials within other evaluative hierarchies but would have placed the entire analysis within the same referent system. That is, a given Ego would always have been the focus of the evaluation. A less involved alternative would have been achieved had samples of the community been drawn with which the evaluations of influentials could have been compared. Another phase which in future research will yield more precise information on relative influence among influentials involves increas- ing the size of the sample. In the past, the time and effort involved in interviewing has hampered this goal. (A related problem arises from the nature of the persons who are influential. Generally they are very busy, active persons. Often they are called out-of-town on business matters or for conferences and the like. Thus, the longer the time period given over which to conduct the interviews the more likely will the researcher conclude more interviews. Though there were very few refusals for interviews in El Paso, for example, many persons were simply unavailable in the short, 16-day period set aside for inter- views. Much better response was achieved in Tucson where the inter- viewer remained in the community.) A larger sample would have enabled an increase in the extent of involvement of various classes of influentials. The model used by Freeman, e_t a}. (23. iii.) is an ideal one from the double standpoint of wide involvement in the influential group and efficient use of research time. The model could likely be improved with an evaluation of relative influence such as that described in this study . REFER ENC ES Agger, Robert E. (1956) Power Attributions in the Local Community: Theoretical and Research Considerations, Social Forces, 34: 322-331. and Daniel Goldrich (1958) Community Power Structures and Partisanship, American Sociological Review, 23:383-392.. and Vincent Ostrom (1956) The Political Structure of a Small Community, Public Opinion Quarterly, 20:81-89. Barber, Bernard (1957) Social Stratification: A Comparative Analysis of Structure and Process, New York: Harcourt, Brace and Company. Barth, Ernest A. T. and Baha Abu-Laban (1959) Power Structure and the Negro Sub-Community, American Sociological Review, 24:69-76. Bell, Wendell, Richard J. Hill, and Charles R. Wright (1960) Public Leadership in the United States, White Plains, New York, The Fund for Adult Education (mimeographed) cited in Freeman, e_t :11. , loc. cit. Cartwright, Dorwin (ed.) 1959 Studies in Social Power, Ann Arbor, The University of Michigan Press. Dahl, Robert A. (1960) The Analysis of Influence in Local Communi- ties, Social Science and Community Action (Charles R. Adrian, Editor), East Lansing, Michigan, Michigan State University, pp. 25-42. (forthcoming) Equality and Power in American Society, Power and Democracy in America (William V. D'Antonio and Howard J. Ehrlich, Editors), Notre Dame, Indiana, University of Notre Dame Press. D'Antonio, William V. (1958) National Images of Business and Politi- cal Elites in Two Border Cities, Ph. D. dissertation, Michigan State University. 115 D‘Antonio, William V., William H. Form, Charles P. Loomis, and Eugene C. Erickson (1961) Institutional and Occupational Representations in Eleven Community Influence Systems, American Sociological Review, 26:440-446. Davis, Harold E. (1958) Government and Politics in Latin America, New York, The Ronald Press Co. Dixon, Wilfrid J. and Frank J. Massey, Jr. (1957) Introduction to Statistical Analysis, New York, McGraw-Hill Book Company, Inc. Fanelli, A. Alexander (1956) A Typology of Community Leadership Based on Influence and Interaction Within the Leader Subsystem, Social Forces, 34:332-338. Form, William H. and William v. D'Antonio (1959) Integration and Cleavage Among Community Influentials in Two Border Cities, American Sociological Review, 24:804-814. and Delbert C. Miller (1960) Industry, Labor, and Community, New York, Harper's and Brothers. and Warren L. Sauer (1960a) Community Influentials in a Middle-sized City, General Bulletin No. 5 of the Institute for Community Development, Michigan State University. (1960b) Organized Labor’s Image of Community Power Structure, Social Forces, 38:332-341. Foskett, John M. and Raymond Hohle (1957) The Measurement of Influ- ence in Community Affairs, Research Studies of the State College of Washington, XXV: 148-154. Freeman, Linton C. and Warner Bloomberg, Jr. , Stephen P. Koff, Morris H. Sunshine, and Thomas J. Fararo, (1960) Local Com- munity Leadership, Paper No. 15 of the University College, Syracuse University. Hunter, Floyd (1953) Community Power Structure: A Study of Decision Makers, Chapel Hill, The University of North Carolina Press. (1959) Top Leadership, U.S.A. , Chapel Hill, The University of North Carolina Press. 116 Hunter, Floyd, Ruth Connor Schaffer, and Cecil G. Sheps (1956) Community Organization: Action and Inaction, Chapel Hill, The University of North Carolina Press. Klapp, Orrin E. and L. Vincent Padgett (1960) Power Structure and Decision—Making in a Mexican Border City, The American Journal of Sociology, LXV:400-406. Lynd, Robert S. (1945) Knowledge for What? Princeton, Princeton University Press, cited in William H. Form and Delbert C. Miller, loc. cit. and Helen Lynd (1929) Middletown, New York, Harcourt Brace and Co. , cited in William H. Form and Delbert C. lvliller, loc. cit. (1937) Middletown in Transition, New York, Harcourt, Brace and Co. Loomis, Charles P. (1960) Social Systems: Essays on Their Persistence and Change, New York, Van Nostrand Co. Merton, Robert K. (1957) Social Theory and Social Structure, Glencoe, The Free Press, Chapter X. Miller, Delbert C. (1958a) Decision-Making Cliques in Community Power Structures: A Comparative Study of an American and an English City, American Journal of Sociology, 64:299-310. (1958b) Industry and Community Power Structure: A Com- parative Study of an American and an English City, American Sociological Review, 23:9-15. (1957) The Prediction of Issue Outcome in Community Decision-Making, Research Studies of the State College of Washington, 25:137-147. Mills, C. Wright (1956) The Power Elite, New York, Oxford University Press. Nelson, Howard J. (1955) A Service Classification of American Cities, Economic Geography, 31:189-210. Olmsted, Donald (1954) Organizational Leadership and Social Structure in a Small City, American Sociological Review, 19:273-281. 117 Parsons, Talcott (1951) The Social System, Glencoe, The Free Press. (195 3) A Revised Analytical Approach to the Theory of Social Stratification, Class, Status and Power: A Reader in Social Stratification (Reinhard Bendix and Seymour M. Lipset, Editors), Glencoe, The Free Press. Pellegrin, Roland J. and Charles H. Coates (1956) Absentee—Owned Corporations and Community Power Structure, American Journal of Sociology, 61:413-419. Polsby, Nelson W. (1958a) The Sociology of Community Power: A Reassessment, Social Forces, 37:232-242. (1958b) Three Problems in the Analysis of Community Power, American Sociological Review, 24:796-803. Rossi, Peter H. (1957) Community Decision Making, Administrative Science Quarterly, 1:415-443. (1960a) Theory and Method in the Study of Power in the Local Community, Paper presented at the 1960 Annual Meeting of the American Sociological Association, (mimeographed). (1960b) Theory, Research, and Practice in Community Organization, Social Science and Community Action (Charles R. Adrian, Editor), East Lansing, Michigan, Michigan State University. and Robert A. Dentler (1960) The Politics of Urban Renewal, Glencoe, The Free Press. Survey of Buying Power (1959), Sales Management, 82-201-780. Sauer, Warren L. (1960) Labor-Business Images of Community Power: Convergences and Divergences, Ph.D. Dissertation, Michigan State University. Schulze, Robert O. (1958) Economic Dominants in Community Power Structure, American Sociological Review, 23-3-9. and Leonard U. Blumberg (1957) The Determination of Power Elites, American Journal of Sociology, 63:290-296. 118 Septimo Censo General de Poblacion, 6 de Junio de 1950, Baja California Territorio Norte. , Estado de Chihuahua. Siegel, Sidney (1956) Nonparametric Statistics for the Behavioral Sciences, New York, McGraw-Hill Book Company, Inc. Smith, Ted C. (1960) The Structuring of Power in a Suburban Com- munity, The Pacific Sociological Review, 3:83-88. Smuckler, Ralph H. and Georgé M. Belknap (1956) Leadership and Participation in Urban Political Affairs, Government Research Bureau, Michigan State University. Stewart, Frank A. (1947a) A Sociometric Study of Influence in South- town, Sociometry, 10:11-31. (1947b) A Study of Influence in Southtown, II, Sociometry, 10:273-286. Stouffer, Samuel A. (1955) Communism, Conformity, and Civil Liberties, Garden City, Doubleday cited in Robert O. Schulze and Leonard U. Blumberg, loc. git. U. S. Bureau of the Census. U. S. Census of Population: 1950. Vol. II, Characteristics of the Population, Part 3, Arizona, Chapter B. United States Government Printing Office, Washington. D. C. , 1952. Vol. II, Characteristics of the Population, Part 5, California, Chapter B. United States Government Printing Office, Washington, D. C., 1952. Vol. II, Characteristics of the Population, Part 31, New Mexico, Chapter B. United States Government Printing Office, Washington, D. C. , 1952. Vol. II, Characteristics of the Population, Part 43, Texas, Chapter B. United States Government Printing Office,- Washington, D. C., 1952. Vol. IV, Special Reports, Part 3, Chapter C, Persons of Spanish Surname. United States Government Printing Office, Washington, D. C., 1953. 119 Walker, Helen M. and Joseph Lev (1953) Statistical Inference, New York, Henry Holt and Company. Winnie, William W. , Jr. (1960) The Spanish Surname Criterion for Identifying Hispanos in the Southerwestern United States: A Pre- liminary Evaluation, Social Forces, 38: 363-366. Wolff, Carole A. (1961) The Role of Women in the Structure of "Society" in a Midwestern Industrial Community, M.A. Thesis, Michigan State University. Wolfinger, Raymond E. (1960) Reputation and Reality in the Study of "Community Power, " American Sociological Review, 25:636-644. New York, The Tressler Press. APPENDICES 120 APPENDIX I 121 122 Respondent number Date of interview Inte rviewe r In order to get the most complete and clearest picture possible of community leaders and their activities, we would first like to know something about your background. 1. Do you mind telling us when you were born? Year 2. Where did you grow up? City State Town? Farm? Nation City State Town? Farm ?— Nation 3. What is (or was) your father's regular occupation? (Be specific) Occupation Industry Example: Occ. mechanic Ind. airplane facto ry farm labo re r cotton farm 4. How long have you lived in this area? years. 5. Under what conditions, if any, would you consider leaving this com- munity? 6. How much schooling have you had? Numbe r of years How many, if any, were in Mexico? (If college or university) Degrees: Major subject in college 7 . 10. 11. 12. 13. 123 Do you speak any other language besides English? If so indicate what languages ? Well Fair Little . Do you have any church affiliation? . What is the name of your firm or organization? How long have you worked for this organization? years. What kind of work do you do? (Official title) (Job desc ription) How long have you been in this position? years. Could you please give us a brief outline of the high points in your occupational experience: What are the major steps you have taken to get to this present position. (Some of this data may be obtainable beforehand.) Fi‘rm or Dates Position Description Ogranization City State (years) 14. As a businessman and/or community leader, (use relevant phrase) What are the most important voluntary associations of which you are or have been a member? 15. 124 Attend Regularly Name of Organization Office held Committee Y/N 1. Why do you think of them as important? (Probe for differences between them) In any of the associations you have mentioned, do you sometimes think of yourself as representing a group such as business, professions, or government? No (Go to Question 16 directly) Other __ Yes (IF YES) 15a. What organizations ? 1. 4. Why is the representation of your (group respondent has mentioned above) important? 125 15b. Do you feel that your participation contributed to the associations' general policies or not? ' No Why not? Yes __ Which associations ? 1. 2. 3. Other _ (Explain) 16. What kinds of groups do you think ought to be represented in local government? 17. Is your group (mention name of relevant occupational group) now represented in the local government? (In this question group may refer to a specific business type such as insurance, or a church, or a profession). 18. (If not answered in 17, ask), Is this representation now at the admin- istrative or legislative level? (Probe to find out just who represents him) 19. 20. 21. 22. 23. 126 How do you usually proceed with respect to important issues: do you contact these men in government, or do they contact you? How would you rank the importance of the following kinds of political participation for persons of your occupational group? running for office being an active political party member contributing money to campaigns giving advice if called on prefer not to be involved in political activity Is your wife active in community affairs, through such organizations as League of Women Voters, AAUW, etc. ? (This information may be obtained elsewhere.) What sort of a role has your (name organization or firm) had in health matters in the community in recent years, that is, in such things as the Polio Drive, Hospital Fund Drive, etc. ? What particular types of community activities would you personally be most likely to become interested in? 127 24. Is this interest due to your occupational position or to some other factor ? (Probe) 25. What have been the major community activities in which you have 26. 27. 28. participated in the last five years or so? Have you personally ever taken an active role in general community health problems ? (If no, skip next question) Do you feel that you represent your occupational organization's interests in these activities, or is this something you do strictly on your own? Suppose that a major hospital project were before the community, one that required decision by a group of leaders whom nearly every- one would accept. If you were completely free to choose, which people would you choose to make up this group--regardless of whether or not you know them personally? 29. 30. 31. 128 l 5. 2 6. 3 7. 4. 8. Why have you chosen these people? (Write in number to left of name in appropriate blanks here.) Their personal qualities Technical knowledge Organizational affiliation Personal friends Other Whom would you choose if the project were a school bond issue? 1. 5. 2. 6. 3 7. 4. 8. Why the differences ? If a decision were to be made in the state capital that would affect a hospital or other health project in your community, who would be the best contact man to get in touch with state officials (besides local members of the legislature?) Why did you choose him? 129 31a. Who would be the best contact man for the school problem? Why ? 32. In your response to the hypothetical problem about the hOSpital project and getting an acceptable group together to put it across, you (didn't mention or mentioned only x number of physicians as the case may be). What would be the general role of the physicians in this community in such a project? Which doctors would be important here? Why the s e ? 33. Have the medical men here ever resolved any major health issues on their own? Yes No Why? 130 34. Now we would like to discuss how an issue actually gets resolved in We are not primarily interested in how you stand for or against a community issue or project. Let us consider the New Hospital Expansion Drive (or other such hospital projects as relevant): (Use same issue with every respondent) How did the issue arise in the first place? Who initiated it? Were you contacted on this issue? Yes No (If yes) How were you contacted? Personal call Informal chance meeting Private luncheon Other (specify) Committee meeting Who contacted you ? Did you contact others ? Yes Whom No What did you decide to do? Did you do it? (explain) Who else did you talk to about it? Why? (Follow action, involvement, etc.) 35. 131 What persons and organizations worked for this issue? Against the issue? (F - for; A - against) ORGANIZATIONS , PERSONS (If the issue is not yet resolved, ask) How, in your opinion, will the issue be resolved? (Probe as to why he feels this way?) Was it necessary to organize in some way to achieve community approval? How long after the issue first arose did this take place? (If not already clear from the previous discussion, ask) In the final analysis, whose influence counted most? Names of persons Organizations How did they put it across, or get it their way? Now let us compare this issue with another issue (inject title of issue as relevant but not a health issue. Use same issue for all respondents). How did this issue arise? Who initiated it? 132 Were you contacted on this issue? Yes No (If yes) How were you contacted? Personal call Informal chance meeting Private luncheon Other (specify) C ommitte e meeting Who contacted you ? Did you contact others ? Yes Whom No What did you decide to do? Did you do it? (explain) Who else did you talk to about it? Why? (Follow action, involvement, etc.) What persons and organizations worked for this issue? Against the issue? (F - for; A - against) ORGANIZATIONS PERSONS (If the issue is not yet resolved, ask) How, in your opinion, will the issue be resolved? (Probe as to why he feels this way?) 133 Was it necessary to organize in some way to achieve community approval? How long after the issue first arose did this take place? (If not already clear from the previous discussion, ask) In the final analysis, whose influence counted most? NAMES OF PERSONS ORGANIZATIONS How did they put it across, or get it their way? 134 36. NAMES OF TOP INFLUENTIALS l. 26. 2. 27. 3. 28. 4. 29. 5. 30. 6. 31. 7. 32. 8. 33. 9. 34. 10. 35. ll 36. 12. 37. 13. 38. 14. 39. 15. 40. 16. 41. 17. 42. 18. 43. 19. 44. 20. 45. 21. 46. 22. 47. 23. 48. 24. 49. 25. 50. Listed above are the names of people who are considered to be influential in this community. We have talked about some of them already. Now we would like you to consider the whole list and rank the ten you consider to be the most influential. If a name has been omitted which you believe should be here, please feel free to add it. They are listed in alphabetical order. 135 .om .om .wm .NN .oN .mm .vN .mm .NN .LN .ON .mL .wa .: .8: .m3 41 .md .NA .: O _4 0‘ HNM‘d‘LfiOl‘CD mM2d£ 30>. 80:3 £333 8808303 .053 Go mega: 83 xoono >358 E303 30> 3 fl 3.300.858 2303 03 $.38?» o>fl amen 03$ :3 mEoBOHQ 5.235800 30.55 amen vim: £033 mo3mmfi .3830 38338 peumfl 88: 03 .owmm age :0 .nm 136 .3 memos» 30m m03m3 ucmtomcfi “mos 03.3 and 38:5 >288 Emmeo mo oz a: 8 m:30 oobwmmflfl ovpmaq. wmv3mmfl acmuhoaefi $05 03? 0.3m ewes» «was. eouwmmflp .30 093mm >2mgecow 3.335500 e5. mo 9.pr wcgmemmufimflcmmm .05. m0 mmzfimfiaomoumoh E303 .mvm N. 0800330 m. 0x33 eat Howie 330.3m .Q3ooov 330.3m .9503 :23 :23 8368 8:3 8:82 m>mpou mcfiowm 23m mo3mm> 2.333.3an“ shoe 0:“ mo e50.» one was? .wm 40. 41. 42. 43. 137 Since (Although) this community has a large (small) percentage of people of Spanish-speaking background living here, we are also interested in knowing something about their participation in commun- ity affairs and the problems they present for community leadership, particularly in health matters. Does the community have a group of Spanish-speaking leaders apart from the top community leaders whom we have already talked about? WHO ARE THEY ? l 6. 2. 7 3 8. 4. 9. 5. 10. Are any of these Spanish-speaking leaders particularly interested in health matters in the community? (IF YES) Do they take an active part in community health projects? AT WHAT LEVEL do they participate? Decision making? Initiation of action Carrying out of policy How would you evaluate the extent of participation of the Spanish- speaking leaders in community affairs ? As much as other people More Less Other On what kinds of problems is it necessary to take Spanish-speaking leaders into account in important community projects? 138 43a. In important issues, at what stage do you solicit support from the 44. 45. Spanish- speaking group ? What kind of support do you need? It has been noted that persons of Spanish-speaking background here in the Southwest are not very strongly represented in the medical profession, at least not in proportion to their total numbers. How would you account for this? It has frequently been found in the Southwest that people of Spanish- speaking background have to work a lot harder, perhaps twice as hard, as Anglo-americans to get ahead. Would you say that this is true in this community? a. Would it be true for a Spanish-speaking person trying to get ahead in business ? b. Would it be true for a Spanish-speaking person trying to get ahead in government? c. Would it be true for a Spanish-speaking person trying to go into medicine, particularly if he wanted to become a physician? (IF DIFFERENCES) Why these differences? 139 46. In your opinion, do Anglo-americans here in the Southwest generally 47. like 0t dislike the Spanish-speaking Americans here? Like Dislike Other Are there any particular jobs or occupations in the area of health including hOSpital work which persons of Spanish-speaking back- ground seem to be better suited for than others. WHICH? In closing this interview we would like to know what experiences or lessons you have gained from your many years of community leader- ship with respect to working with other people? Has it been a worth- while experience? How so? What have been the major problems in trying to work with other people? APPENDIX 11 140 141 Table A. Service classification, estimated population, and percentage of persons with Spanish surname of six cities. Service Estimated Population Percentage of Classifi- of Central Cit at Spanish Surname City cationa Time of Study Population in- 1950 San Diego, California PbZPsF 522, 600 4. 6% El Paso, Texas T 263, 000 49. 0 Tucson, Arizona PsZPfF 110,000 21.1 Las Cruces, New Mexico Pb2R 22, 500 48. 7 Ciudad Juarez csC 250, 000d 96. 0e Tijuana csC 160,000f 97.0e Source: Selected from Table 1, D'Antonio e_t a_._l. , (1961:441). aTaken from Howard J. Nelson, "A Service Classification of American Cities," Economic Geography, 31 (1955), pp. 189-210. Key: Mf, Manufactur- ing; R, Retail Trade; Pf, Professional Service; T, Transportation and Com- munication; Ps, Personal Service; Pb, Public Administration; W, Wholesale Trade; F, Finance, Insurance and Real Estate. A "2" after the symbol signi— fies the city fell two standard deviations from the mean. The Nelson classification considers the proportion of the labor force engaged in performing a service for the 897 urban concentrations of 10, 000 or more persons in the United States. The mean proportion of persons engaged in a particular service in all 897 cities is calculated for each service group and standard deviation in excess of the mean is calculated. The classification of a city in one group arbitrarily reflects an excess of at least one standard devi- ation of the labor force employed in that service group. The service data is taken from the 1950 Census of Population Classification of Industry Groups (Vol. 2, Table 35, "Economic Characteristics of the Population. . . . " See Nelson, 22. c_i_t. Except where noted differently, these estimates are taken from Sales Management, "Survey of Buying Power," 82 (May 10, 1959), and are the estimates as of January 1, 1959, pp. 201—780. CThe classification of Mexican ‘cities are based on a service classifi- cation of Municipios of over 10, 000 population which include a city of over 10, 000. There were 37 municipios which fell within our criteria from the six border states. The Mexican and United States Census have non-compar- able categories for "trade" and "service. " For this reason two different categories were made for the Mexican cities. The Commercial (C) includes Nelson's Retail and Wholesale Trade, Finance, Insurance, and Real Estate. The Service (S) includes Nelson's Professional and Personal Service, and Public Administration. ' Estimate based on census taken by municipio government, January 1959. 6The percentage of Mexican persons in the population in 1950; from Septimo Censo General dSPoblacion, 6 clSJuniode 1950. Estimate obtained by researchers in the area. 142 Table B. Comparisons of All Influentials Named and Not Interviewed with Influentials Named and Interviewed for Certain Occupational Categories for San Diego, El Paso, Tucson, Las Cruces, Cd. Juarez, and Tijuana. Community and All Influentials Named Chi- Occupational Categories Not Interviewed Interviewed Total Square p (r1) (II) (n) (X?) San Diego Merchants 1 7 8 Banking , Financ e ' [\a p...- p—r Manufacturing and Construction 3 6 9 Professionals 2 3 5 Government Service 1 1 2 All other 0 1 1 Totals 9 27 36 . 858 <. 50 El Paso Merchants 6 6 12 Banking, Finance 2 5 7 Mfg. and Constr. 6 6 12 Professionals 9 10 19 Government Service 2 5 7 All Other 4 1 5 Totals 29 33 62 1.111 <.90 Tucson Merchants, Banking and Finance 1 19 20 Mfg. and Constr. 0 4 4 Professionals, Gov't Service, Others 1 19 20 Totals 2 42 44 . 000 >. 99 Las Cruces Merchants 6 4 10 Banking, Finance 1 5 6 Mfg. and Constr. 2 2 4 Professionals and Government Service 7 10 17 All Other 7 5 12 Totals 23 26 49 . 924 <. 70 Juarez Merchants 15 6 21 Banking, Finance 2 2 4 Mfg. and Constr. 11 6 17 Professionals 18 7 25 Government Service 7 2 9 Others 9 2 11 Totals 62 25 87 1.771 <.8O Not Ascertained Z O 2 143 Table B Continued Community and All Influentials Named Chi- Occupational Categories Not Interviewed Interviewed Total Square p (n) (n) (n) (X2) Tijuana Merchants 14 9 23 Banking, Finance 8 1 9 Mfg. and Constr. 0 2 2 Professionals 8 5 13 Government Service 6 3 9 Others 4 2 6 Totals 40 22 62 8 O 8 .521 <. 95 Not Asc e rtained Chi-Squares were computed with certain categories combined. When categories were combined the most nearly like occupations were grouped. APPENDIX III 144 APPENDIX III This thesis has been concerned with the association of certain social characteristics with status-roles evaluated as influential. With few exceptions it has not been our interest to consider the relative influence which specified influential status-roles are assumed to possess. The general argument which has been presented is that there are social characteristics which are conditions or indices of conditions the presence of which enhance a potential for influence. It may also be argued that degrees of influence may be associated with the degree to which the influential status-roles possess the total set of social characteristics. Two indexes are defined: Relative Influence Index. --The number of votes an individual received nominating him as one of the top ten influentials in the community. Relative Social Characteristics Index. --The number of highly evaluated social characteristics a given influential possesses. To determine the degree of a social characteristic which is to be classi- fied as "highly evaluated" each of the social characteristics (13) was to be dichotomized and scored "1" for high evaluation and "O" for Not high evaluation. Each item (social characteristic) in the index was scored consistent with the findings presented in Chapters VI and VII. For each of the six communities the following items were rates as follows: Item Score ”1" HO" Ethnicity Anglo Mexican descent Father's occupation White Collar Not White Collar Occupation Business Not Business Position Executive or Not Executive or Directive Directive Marital Status Married Not I/Iarried Orientation to Community Locally oriented Not Locally Oriented Sex Male Female 145 146 For each of the communities in the United States the item religious affiliation was scored, "1" for Protestant and "0" for Non-Protestant. For each of the communities in Mexico the item religious affiliation was scored, "1" for Roman Catholic and "O" for Non-Roman Catholic. For each of the remaining characteristics a score for each individual was available. It was reasoned that if x years in the com- munity is minimal to a potential for influence then x + 1 will be "better." The point at which the dichotomy of "+" and "-" years would be determined is to be the mean number of years given for the item for each community. Thus, for each of the following communities the score "1" was given at the level indicated. (A "0" score would be any number less than the one listed.) "1" scores given at anything equal to or greater than the level indicated for each social characteristic (item) listed for each of six communities. Communities San El Las Ciudad Item Diego . Paso Tucson Cruces Juarez Tijuana -------------------- years---------------------- Age 55 54 59 49 46 52 Residence in Community 40 40 34 25 18 28 Tenure in Firm 29 23 24 14 11 16 Tenure in Position l6 13 18 12 9 15 -Years in Education 15 15 15 16 14 13 Each individual received a score on each item of "1" or "0. " With 13 items in all the score for an individual might range from "0" to "13. " This cumulative score is the Relative Social Characteristic Index. A product moment correlation was then computed on the two indexes for each set of influentials by community. The results were: 147 Community r t p N San Diego -.0618 -0.3133 .60