1 . €3.31!» . .9. .7 a. v .1! . $1....“ 1.. 4x 3 E. 4 : 7... 3’71} 1 PI»; I} I v: truly... a . LN. . . : :7;qu _$.Zu.r..4u,.\353,..n . , 1:33;? 5:. . . . 17. . ‘ 7... 0’1???» . . f .~J.>.\_VF..A \ f. , f 3:. iii! .. I. . . . H 4i». til-.192! 1.: _ v v, .434 in 74.. 3.6 . I" 5? '7... p5..." ... it .. , wand}. .egfiqfiufiwké, .. . V ..u.:..:.n.i.b£ LIBRA R 1/ Michigan State . #1an C‘VI'}, g‘mw [#7. This is to certify that the , thesis entitled THE ORGANIZATION OF COMMUNITY STATUS STRUCTURES: AN ANALYSIS OF STRATA CONSISTENCY-INCONSISTENCY AMONG COMMUNITIES IN THE CONTINENTAL UNITED STATES presented by RICHARD BERNELL STURGIS has been accepted towards fulfillment of the requirements for jflLdegree in M9! 0 Mn. 5/ / Major professofj Date M 0-7639 I 2 I a we??? 3W , efl and (co the ana exister in com degree PositiI Si0ns . lkhed comm IIere s that V a the( divis: °f laI Persp‘ ABSTRACT THE ORGANIZATION OF COMMUNITY STATUS STRUCTURES: AN ANALYSIS OF STRATA CONSISTENCY-INCONSISTENCY AMONG COMMUNITIES IN THE CONTINENTAL UNITED STATES By Richard B. Sturgis The problem of the research was couched within a human ecology and (community) stratification orientation. The units of observation in the analysis were communities. One objective was to demonstrate the existence and varying forms of strata consistency and inconsistency with— in communities. The second and major objective was to explain the varying degree and types of consistency-inconsistency (the dependent variable). The dependent variable was determined by comparing the relative positions of communities on income, educational, and occupational dimen— sions. The degree as well as type of consistency—inconsistency was estab— lishedfor each community. An example of a "type” would be an income—high community, i.e., the occupation and education scores for the community were similar while the income score was high. This is one of 10 types that were developed. In order to account for the variations in the dependent variable a theoretical scheme was developed around the argument that there is a division of labor between communities. Under the canopy of the division of labor were brought together a number of traditional ecological Parspectives and variables, i.e., size of community, distance from neares (funcI indept white: femal natio from them hypoI wasl nati here for vari clue dam new The mun fun inc per The shI Im Richard B. Sturgis nearest SMSA (metropolitan dominance), and industrial specialization (functional specialization). These served as three of the five major independent variables. The remaining two were the percentage of non— whites in communities and the percentage of labor forces that were female. A random sample of 539 communities (urban places by Census desig- nation) were drawn from the 1960 Census. The communities ranged in size from 10,000 to 100,000. Twenty—four hypotheses were tested and seven of them were statistically significant; however, the results from 12 hypotheses did suggest support for the general rationale. Whenever it was possible the results were tabulated by region as well as for the nation. The conclusions in relation to the data analysis are presented here in very short summary form. In brief, there is some justification for stating the following conclusions: (1) Communities do exhibit various patterns of strata consistency—inconsistency. While this con— clusion was a necessary basis for the research, it had not been demonstrated previously. (2) There is good evidence that distance from nearest SMSA is related to some patterns of consistency-inconsistency. The evidence of a relationship is weaker and less clear between com— munity size and strata arrangements. (3) There is good evidence that functional specialization is related to some forms of consistency- inconsistency. (4) Finally, there is also good evidence that the percentage nonwhite in communities is related to strata arrangements. The evidence also speaks rather strongly for the existence of relation— ships between the percentage of the labor force that is female and forms of consistency—inconsistency. THE ORGANIZATION OF COMMUNITY STATUS STRUCTURES: AN ANALYSIS OF STRATA CONSISTENCY—INCONSISTENCY AMONG COMMUNITIES IN THE CONTINENTAL UNITED STATES By Richard Bernell Sturgis A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Sociology 1971 T somethin ciple p0 begin ”w becomes thanks 1; l tolerate college not Iner in me. HIE ina A“‘08 He Martin williar of lab ACKNOWLEDGMENTS To properly acknowledge those persons who have contributed to something like the culmination of a Ph.D. thesis is, perhaps, in prin- ciple possible, but in fact extremely difficult to impossible. To begin ”way back when” and start with one's fourth grade teacher, at best, becomes ludicrous. Nevertheless, let me take this occasion to express thanks publicly to a few of the more notables. To the parents who initiated my life, gave me sustenance, tolerated me, and, very importantly, financed most of my undergraduate college education. To Professor Evan T. Peterson who interested me in sociology, not merely by his scholarly teaching, but by taking a personal interest in me. To Professor William Form who, rather unknowingly, interested me in a ”structural orientation” to sociology. To Professors Emile Durkheim, Robert Park, Roderick McKenzie, Amos Hawley, Otis Dudley Duncan, Leo Schnore, Jack Gibbs, and Walter Martin for their provocative and insightful writings. To my committee members, Professors William Form, James McKee, William Faunce, and especially Professor J. Allan Beegle, who has demonstrated great patience and lent assistance during the few years of labor pains that have transpired in giving birth to the thesis. ii To Mrs. I wife, Margaret, And, fin; They have been 1 period of this joy, and made ” iii To Mrs. Linda Stoeckel, Mrs. Phyllis Betz, good friends, and my wife, Margaret, who volunteered many, many hours of tedious coding. And, finally, to my family, the most ardent loves in my life. They have been perhaps the major reason for the rather long gestation period of this thesis. However, living with them has provided my major joy, and made "the long wait" not only tolerable but rather pleasant. CHAPTER I. INTRODUCTI II. THEORY--P Introdu Divisio Competi Dominan Relatio Pr0posi HI- THEORY--P Introdu 0Vera11 Size a; FUnctic Females IV- METHODS I Introm Sample Operat Depe Inde CHAPTER I. INTRODUCTION . . . . II. THEORY-~PART I . . . Introduction . . . Division of Labor Competition . . . Dominance . . . . TABLE o u o Relation to Ecocomplex . OF CONTENTS 0 Propositional Statements of Theory III. THEORY—~PART II: Introduction . . . Overall Distribution . . Size and Distance Functional Specialization Females and Nonwhite . . IV. METHODS AND METHODOLOGICAL CONSIDERATIONS Introduction . . . Sample . . . . . . Operational Procedures Dependent variable: Independent variables HYPOTHESES . iv strata consistency—inconsistency PAGE l7 17 22 29 34 36 40 46 46 48 50 58 64 7O 70 CHAPTER Operath Control Diagram Statist V. DATA ANAL Introdu Distrib Size an Summary Functic Summar3 Females Summary Combint Sununarj VI- ISSUES A: Introd The Sa The Us TWO Ma Ineo Summer Sorne 5 BIBLIOGRAPHY APPENDIX A. APPENDIX c. CHAPTER Operational Statement of Hypotheses . . . . . . . . . . . . Control Variables . . . . . . . . . . . . . . . . . . . .. Diagrammatic Presentation . . . . . . . . . . . . . . . Statistical Analyses . . . . . . . . . . . . . . . . . . V. DATA ANALYSIS AND TESTING OF HYPOTHESES . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . Distribution of Strata Consistency—Inconsistency Types Size and Distance . . . . . . . . . . . . . . . . . . . . . Summary of Size-Distance Hypotheses . . . . . . . . . . . . Functional Specialization . . . . . . . . . . . . . . . . . Summary of Specialization Hypotheses . . . . . . . . . . . Females and Nonwhites . . . . . . . . . . . . . . . . . . . Summary of Female and Nonwhite Hypotheses . ~ . a n o o . Combined Variables . . . . . . . . . . . . . . . . . . . . Summary of Findings . . . . . . . . . . . . . . . . . . . VI. ISSUES AND SUMMARIES . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . .‘. . The Sample . . . . . . . . . . . . . . . . . . . . . . . . The Use of Census Data . . . . . . . . . . . . . . . . . . Two Major Variables: Distance and Degree of Consistency- Inconsistency . . . . . . . . . . . . . . . . . . . . . . Summary Statements . . . . . . . . . . . . . . . . . . . . Some Suggestions for Further Analysis . . . . . . . . . . . BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . APPENDIX A. Percentage of Wives Employed by Level of Family Income APPENDIX B. List of Sample Communities by State and Size . . . . - APPENDIX C. Distributions of Dependent Variables . . . . . . . . . PAGE 92 96 97 99 99 99 102 124 125 135 135 146 147 150 152 152 152 154 155 157 160 164 171 173 187 TABLE 1. Distributi lnconsis [\D « Ranked Ass Percenta Ranked L w - Correlatic SMSAs, f .a_\ - Percentage Extremal U1 - C0rrelatic 00nsiste SMSA . , ' PeTCEntage and Dist . PereentagE Extreme} . Percentage EXtreme; SMSAS at LIST OF TABLES TABLE PAGE 1. Distribution of Strata Inconsistency by Type of Inconsistency for the Nation and by Region . . . . . . . . . 100 Ranked Association Between Distance from SMSAs and the N . Percentage of Strata Consistent Communities that are Ranked Low . . . . . . . . . . . . . . . . . . . . . . . 103 3. Correlations Between Consistency Rank and Distance from SMSAs, for the Nation and by Region . . . . . . . . . . . . 104 4. Percentage of Consistent Communities that are Ranked Extremely Low by Size and by Distance from SMSAs . . . . . 106 5. Correlations Between Size of Community and Rank of Strata Consistency for the Nation and by Distance from Nearest SMSA . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 6. Percentage of Communities that are Strata Consistent by Size and Distance from SMSAs . . . . . . . . . . . . . . . . . . 110 7. Percentage of Strata Consistent Communities that are Ranked Extremely High by Size and by Distance from SMSAs . . . . . 113 8. Percentage of Strata Consistant Communities that are Ranked Extremely High for Communities Within a 50—mile Radius of SMSAs and All Other Communities . . . . . . . . . . . . . . 114 9. Strata Characteristics of Small Suburbs . . . . . . . . . . 116 vi TABLE 10. Percentage ._. ._. .._. ca) ,_. U‘l ,._. ox Occupat . Comparism to AllI Strata . Percentag Income Manufac - Percentag High to - Per Cent and by - Per Cent and Non Per Cent Special MUltiple and the f°r the Per CEnt and Ear High fl othert . PerCEnta. High A Low Le I1. TABLE PAGE 10. Percentage of Strata Inconsistent Communities that are Occupation High for Each Distance from SMSAs . . . . . . . 118 11. Comparison of Communities Within a 50—mile Radius of SMSAs to All Other Communities by Frequency of Occupation High Strata Inconsistency . . . . . . . . . . . . . . . . . . . . 119 12. Percentage of Strata Inconsistent Communities that are Income High, With and Without Specialization in Manufacturing, for Each Distance from SMSAs . . . . . . . . 120 13. Percentage of Strata Inconsistent Communities that are Income High for Employing, Diversified, and Residential Suburbs . 122 14. Per Cent of Income High Inconsistent Suburbs for the Nation and by Region . . . . . . . . .'. . . . . . . . . . . . . . 124 15. Per Cent of Strata Consistent Communities by Specialized and Nonspecialized Communities . . . . . . . . . . . . . . . 126 16. Per Cent of Strata Consistent Communities by Type of Specialization . I O 0 I I O I l I U 0 I I O l . . . . . . . 126 17. Multiple and Partial Correlations Between Size and Distance and the Dependent Variable, Degree of Strata Consistency, for the Nation and by Region——Nonspecialized Communities . . 127 18. Per Cent of Strata Consistent Communities for Nonspecialized and Each Type of Specialized Community . . . . . . . . . . . 130 19. Percentage of Strata Inconsistent Communities that are Income High for Communities Specialized in Manufacturing and All Other Communities . . . . . . . . . . . . . . . . . . . . . 130 20. Percentage of Strata Inconsistent Communities that are Income High for Communities with High (Specialized), Moderate and Low Levels of Manufacturing . . . . . . . . . . . . . . . . 131 TABLE 21. Community Inconsis 22. Percentage Educatic All 0the m w . Correlatir Consists the Nat 24. Percentag QUintiI 25- Percentag High an Labor F 26~ Percentag Ranking ' COrrelati Female by Regi 28' COrrelati by Regj 29. PerCQntaE QUintTJ 30. PerQentas IncOme Rankin TABLE 21. Community Characteristics Linked with Income High Inconsistency . . . . . . . . . . . . . . . . . . . . . . . 22. Percentage of Strata Inconsistent Communities that are Low in Education High for Communities Specialized in Education and All Other Communities . . . . . . . . . . . . . . . . . . 23. Correlation Between Size of Community and Degree of Strata Consistency for Communities Specialized in Education, for the Nation and by Region . . . . . . . . . . . . . . . . . 24. Percentage of Communities that are Strata Consistent by Quintile Rankings of Percentage of Labor Force Female . . . 25. Percentage of Strata Inconsistent Communities that are Income High and Income Low by Quintile Rankings of Percentage of Labor Force Female . . . . . . . . . . . . . . . . . . . 26. Percentage of High and Low Ranked Consistency by Quintile Rankings of Percentage of Labor Force Female 27. Correlation Between Percentage of the Labor Force that is Female and Degree of Strata Consistency for the Nation and by Region, Excluding High Income Communities . . . . . . . 28. Correlation Between Percentage of the Labor Force that is Female and Degree of Strata Consistency for the Nation and by Region, for all Communities . . . . . . . . . . . . . 29. Percentage of Communities that are Strata Consistent by Quintile Rankings of Percentage Nonwhite . . . . . . . . 30. Percentage of Strata Inconsistent Communities that have Income Inconsistency, Both High and Low, by Quintile Ranking of Percentage Nonwhite . . . . . . PAGE 132 133 134 136 137 139 142 142 I I I I TABLE 3L w N w w w 4_\ pow \10‘ LA) co Correlati for Nat . Correlati Consist . Correlati white 2 Nation Summary I - Centile T - Centile I « Centile I DiStribut ' Distribut and by ix TABLE PAGE 31. Correlation Between Per Cent Nonwhite and Consistency Rank for Nation and by Region . . . . . . . . . . . . . . . . . 144 32. Correlation Between Percentage Nonwhite and Degree of Strata Consistency for the Nation and by Region . . . . . . . . . . 145 33. Correlation Between Distance from Nearest SMSA, Per Cent Non— white and the Dependent Variable Consistency Rank for the Nation and by Each Region . . . . . . . . . . . . . . . .'. 148 34. Summary of Findings from Testing the Hypotheses . . . . . . . 151 35. Gentile Disttibfition and Codes for Occupational Strata Scores 187 36. Gentile Distribution and Codes for Educational Strata Scores . 187 37. Gentile Distribution and Codes for Income Strata Scores 38. . . . 188 Distribution of Strata Consistency-Inconsistency Degree Scores 188 39. Distribution of Consistency—Inconsistency Types for the Nation and by Region 0 o n o n ‘o o l a I v I o o o a o n a l a . . 189 FIGURE l. Diagramma 2. Histogram Various LIST OF FIGURES FIGURE PAGE 1. Diagrammatic Presentation of the Conceptual Scheme . . . . . . 98 2. Histogram Depicting Per Cent of Wives Employed Within Various Family Income Categories . . . . . . . . . . . . . . 171 The prob ecology and soc draw upon tradi within sociolog study which are The firs cept of status analysis of an m“TholOgy frar this chapter, The Com P°Sitions With 9'8', "stratum and P6rhaps th lizatiomu Th throllghoug thi for using this 1- Alt conSIStency or trated 0n is t to indiCate 81 Q0 %- temimogy wI inconSIStEnCy CHAPTER I INTRODUCTION The problem to be explored in this thesis is a problem in human ecology and social stratification. Many aspects of what follows will draw upon traditional notions of social stratification, as it is viewed within sociology. On the other hand, there are two major aSpects of this study which are less traditional and require some comment. The first divergence is a concentration of attention on the con- cept of status consistency—inconsistency, and the second is that the analysis of this thesis will be conducted within a human ecology—social morphology framework. The second point will receive attention later in this chapter. The concept or notion of units sharing relatively different rank positions within different ranked orders has undergone various namings, H H H H e.g., ”stratum consistency, status consistency, status congruency," and perhaps the most popular—~"status crystallization" or ”class crystal— lization." The term "strata consistency-inconsistency" will be used throughout this paper. We feel there are a number of very good reasons for using this terminology. 1. Although writers have consistently referred to the notion of consistency or congruency, what their research and essays have concen— trated on is the lack of consistency or congruency. The usual attempt to indicate status inconsistency has been to refer to low degrees of consistency. While the procedure is legitimate it tends to hide in terminology what much of the real interest has been—-a concern with inconsistency. Our use of the phrase consistency—inconsistency is a 1 straightforward examining. 2. The largely initiat tion: A non—ve tion of this a have often use rather than co when the term ‘ tried to conve English usage definite form,l him cr stal i.e., the unit positions in d: We have trying to grass: the term "crys concepts in so jargon well in present case t 2 straightforward attempt to indicate what it is we are interested in examining. . 2. The recent interest in status consistency—inconsistency was largely initiated by Lenski in a 1954 article titled ”Status Crystalliza— . . . . . 1 tion: A non—vertical dimenSIon of soc1al status.” Since the publica— tion of this article the rash of essays and research which have followed have often used the term "crystallization." The use of crystallization rather than consistency was unfortunate. The intuitive meaning conveyed when the term ”consistency” is used comes much nearer what writers have tried to convey than does ”crystallization." If we accept the standard English usage of crystallize, "to assume or cause to assume a fixed and definite form,"2 a major problem is evident. A unit's status could be highly crystallized, fixed and definite, in a highly uncrystallized form, i.e., the unit may hold definite and fixed but greatly differing rank positions in different status orders. We have noticed in teaching, the difficulty students have in ' when trying to grasp the concept of ”status consistency-inconsistency,' the term ”crystallization” is used. There is no rule which requires concepts in sociology to conform to standard usage, as sociological jargon well indicates; nevertheless, there are no reasons in the present case to cause confusion by not subscribing to common usage. —_————-—-———.__._.____ 1G. E. Lenski, ”Status Crystallization: A non-vertical dimension of social status," American Sociological Review, 19 (August, 1954), 405— 13. 2Webster's New Collegiate Dictionary, 2nd ed. (Springfield, Mass.: G. and C. Merriam Company, 1953). P- 201- It appea are in agreemen works. He now tion."3 3. Our practical reaso distinction be usage of the to in the stratifi be more to our to talk of "st become convent not merely in and out of soc. farting to a u] position in a : themselves, 2. status. Becau because it is consistency-i: its usage in I alter the phrz G. E. TEST," Americ 4M8}! Sociology, II University P 3 It appears that some of the most recent writers on the subject are in agreement with this point. Of note is one of Lenski‘s recent works. He now uses the term ”inconsistency" instead of crystalliza- tion."3 3. Our use of the term "status” (stratum and strata) is based on practical reasons. It is not, however, in keeping with the classical distinction between class, status, and party made by Weber.4 Nor is the usage of the term "class,” when talking about consistency—inconsistency in the stratification order, in keeping with his distinction. It would be more to our own liking, and in keeping with Weber's useful distinction, to talk of ”stratification consistency-inconsistency.” However, it has become conventional in American stratification literature to use status not merely in the prestigeful sense but in a more generic way. Both in and out of sociology, the term ”status” has become a catch—all term re— ferring to a unit's position in a particular ranked order, its overall position in a number of orders, and even to refer to the ranked orders themselves, e.g., economic status, occupational status, and educational status. Because the term ”status" has been used in the above ways, and because it is more often associated with the notion of stratification consistency—inconsistency than any other term, we have chosen to retain its usage in this thesis. Our only deviation in using "status” is to alter the phraseology to "strata" or "stratum." Since status inconsis— tency has been used most frequently by social psychologists, our 3G. E. Lenski, "Status Inconsistency and the Vote: A Four Nation Test," American Sociological Review, 32 (April, 1967), 298-301. 4Max Weber, "Class, Status, and Party,” Max Weber: Essays in Sociology, Hans Gerth and C. Wright Mills, editors (Oxford: Oxford University Press, 1946), pp. 180—195. L alteration of We are not int structural the orientation f inconsistency one considers strata consis traditional. proven the be writers in st to strata con recent attent. of the positit amount of attl the notion of gists). We fee orders within in America or sional strati 4 alteration of the term will help clarify our own usage of the concept. We are not interested in the psychological states of our units, but in structural characteristics. The use of "strata" distinguishes our orientation from the social psychologists and also more explicitly repre— sents our structural orientation. Whether it is correct to consider a concern with consistency— inconsistency as traditional or nontraditional is subject to debate. If one considers Weber's classic thesis5 as an essay basic to the notion of strata consistency—inconsistency, then our concern can be argued as traditional. It could well be the case that our hindsight has again proven the better sight. That is, Weber and other more traditional writers in stratification are viewed as having discussed issues central to strata consistency—inconsistency, now that the concept has gained recent attention, i.e., we have reinterpreted their writing. Regardless of the position taken, what is fairly recent (nontraditional) is the amount of attention that sociologists and social psychologists are giving the notion of status consistency—inconsistency (mostly by social psycholo— gists). We feel that the recognition of the existence of various ranked orders within American Society is basic to understanding stratification in America or any urban—industrial society. The notion of a unidimen— sional stratification order simply does not fit, empirically or theoreti— cally. Given that there are various ranked orders, questions follow: Does a position in one order indicate that the unit in question will occupy a relatively similar position in all other orders or in any of the other orders. On the other hand, if a unit becomes mobile in one ranked 51bid. order does it relationship b questions poin portant areas inconsistency stratificatior Our wor major and impr consistency-i much previous cerned with t cation, and o ranks will to In thi research on s that consistei cerned with t' tency or inco literature so dependent var Some o On the usage \——-— 6John itatus Struct ations" (un I 1966). While COHSiStEncy~j Badman, AIM Sonolo N (NE 5 order does it become mobile in the Other ranked orders? What is the relationship between the various ranked orders? These and related questions point the direction to what we feel is one of the more im— portant areas in the study of social stratification-—the consistency or inconsistency of a unit's position in the various ranks within the stratification order. Our work in this thesis has some similarities but differs in major and important ways from other literature concerned with status consistency—inconsistency. Nevertheless, one similarity shared with much previous research is in the ranked orders that we use. We are con— cerned with the three ranked orders resulting from ranking income, edu— cation, and occupation. Explication and operationalization of these ranks will follow at a later point in the thesis. In this thesis there are three major variations from most previous research on status consistency—inconsistency. The first variation is that consistency-inconsistency is the dependent variable. We are con— cerned with the factors which appear to influence the state of consis— tency or inconsistency of strata. Practically none of the available literature sources have consistency~inconsistency represented as a dependent variable.6 Some of the criticism of consistency—inconsistency has centered on the usage of the concept as an independent variable. There are 6John Stoeckel, ”The Impact of Metropolitan Dominance Upon the Status Structure and Status Consistency of Rural—Farm and Urban Popu— lations" (unpublished Ph.D. dissertation, Michigan State University, 1966). While not Specifically stated, some suggestion of status consistency-inconsistency as a dependent variable is made in Ronald Freedman, Amos Hawley, W. S. Landecker, and H. M. Miner, Principles of Sociology (New York: Henry Holt and Company, 1952), Chapters VII and— XIII. fi questions cone This issue has using degree 0 suggests that inconsistency the analysis 0 one of our maj other impingi as a dependen provide insig inconsistency The se mentioned. I The most freq a unit's place step has been relative posi' score represe‘ be with a sec ranked orders the units obt levels of inc 7Hubet Building: '1" BE, 31 (Fe Imconsistenc Journal 93 § ans 6 ”causal" properties as an independent variable. questions concerning its This issue has led Blalock to question the theoretical justification of using degree of consistency as an independent variable.7 Blalock further suggests that the specification and analysis of patterns of consistency— 8 In inconsistency is a necessary step in clearing the conceptual haze. the analysis of strata consistency—inconsistency as a dependent variable, one of our major concerns is with the varying patterns which result from other impinging factors. While this thesis is concerned with the concept as a dependent variable only, it is possible that such an approach will provide insight for delineating properties of strata consistency— inconsistency as an independent variable. The second major variation we take in this thesis has just been mentioned. It is a concern with the patterns of consistency—inconsistency. The most frequent treatment of the concept in question has been to observe a unit's placement in three or four different ranked orders. The next step has been to determine the similarity (or lack of same) of a unit's relative position in each ranked order and finally to calculate a single score representing the degree of similarity. Our concern will not only be with a score demonstrating a unit's comparable position in different ranked orders but to determine and demonstrate the patterns of similarity the units obtain. For instance, let us use the three ranked orders of levels of income, levels of occupations, and levels of education. It M 7Hubert M. Blalock, Jr., ”The Identification Problem and Theory Building: The Case of Status Inconsistency," American Sociological 337 View, 31 (February, 1966), 52—61; and Hubert M. Blalock, Jr., ”Status Inconsistency and Interaction: Some Alternative Models,” American Journal of Sociology,” 73 (September, 1967), 305—315. —-_. 81bid. i would be possi different and by the number levels--high, highly consist sistently med which could 1: levels within tally examine orders and 1e demonstration contention th or theoretica the patterns It is a single scor the concept 0 issues partia tion system, dimensional r tem is compo: social strat we take the ranked order lightenment tion system positions 11 would be possible for a unit to have consistent strata in a number of different and patterned ways. The number of ways would be determined by the number of levels used in each ranking order. If we used three levels-—high, medium, and low-~in each ranked order, a unit could have highly consistent strata in three ways, i.e., consistently high, con~ sistently medium, or consistently low. The logical number of patterns which could result from our example of three ranked orders and three levels within orders is 27. The number is rather cumbersome to empiri- cally examine and becomes close to impossible if the number of ranked orders and levels within orders is increased. While we contend that the demonstration and examination of patterns is necessary, it is also our contention that a long proliferation of patterns is neither heuristically or theoretically justifiable. A later portion of this thesis outlines the patterns which will receive attention. It is important to observe that if strata consistency is viewed as a single score, we nourish an old problem. One of the reasons for using the concept of Status consistency” has been that it helped to illuminate issues partially hidden under the notion of a unidimensional stratifica— tion system, We contend it is an important step to partial out of a uni— dimensional notion the various stratification orders of which such a sys— tem is composed. In short, what results is a more refined picture of social stratification. However, we shorten our step towards refining if we take the information about a unit‘s relative positions in various ranked orders and squeeze this information into a single score. The en— lightenment produced by forsaking notions of a unidimensional stratifica— tion system is shaded by using only a single indicator of a unit‘s positions in the various orders. Our solution is to consider the patterns é of consistency ' inconsistency. To thi (variations fr inconsistency analysis of 0 focus on the cussion which this study. Common of observatio within the co of observatio strata consis data we use w recent letter inconsistency us as it can couched in 5: some further “a...” 9The status consi "Status Crys PP. 405-13, high in one However, Ler Various pat1 "A Study of . tation, Uni“ loTh Robert Hodg 8 of consistency-inconsistency as well as the degree of consistency— inconsistency. To this point we have mentioned two attempts we are making at Ivariations from usual investigations of status consistency— inconsistency. An understanding of these two variations: (1) the analysis of consistency—inconsistency as a dependent variable, and (2) a focus on the resulting patterns,9 allows us to move to an area of dis— cussion which produces the third and perhaps most unique variation in this study. Communities have often served as units gf analysis but the units of observation have most generally been persons, groups, or institutions within the communities.10 The present study uses communities as the units of observation not as the units of analysis. Our concern is with the strata consistency—inconsistency of communities rather than persons. The data we use will represent properties of aggregates or populations. A recent letter from a critic stated flatly that ”status consistency— inconsistency” was an individual measure. This appears rather naive to us as it can be used for anything for which an appropriate rationale couched in scientific rigor can make of it, as long as it appears to offer some further understanding. That is, as we understand, how a discipline M 9The reader should note that specifying and examining patterns of status consistency—inconsistency is not new but is uncommon. Lenski, ”Status Crystallization: A non—vertical dimension of social status,” pp. 405—13, paid some attention to the differing effects of a unit being high in one rank and low in another as opposed to some other arrangement. However, Lenski neither claimed nor attempted to specify and examine various patterns. The work of most note in this regard is Ralph Spielman, "A Study of Stratification in the United States” (unpublished Ph.D. disser- tation, University of Michigan, 1953). 10This useful distinction was first called to our attention by Robert Hodges, ”Occupational Composition and Status Crystallization: An grows. The pa are those whic Our dependent sistency—incor will be asking and education; what patterns structure or : The va‘ viewed as fit Specifically, division of l intercommrunit dominance and ourselves wit United States dimensions ar munities. M Aggregate App University of 11We l “-3. Bureau < "place" and < follow in a I lee 1 using commun: 0f problem it! article, "Cu of Social Or 1959), 132-4 cOIImunities 9 grows. The particular aggregates and populations we are interested in are those which compose the social units we refer to as "communities.”ll Our dependent variable will be the degree and patterns of strata con— sistency—inconsistency which exists in communities. More explicitly, we will be asking if the distributions of income levels, occupational levels, and educational levels of communities are consistent or inconsistent and what patterns of consistency-inconsistency are demonstrated? It is the structure or form of the community with which we are concerned.12 The variables we will be using as independent variables are all viewed as fitting within the broader concept of the division of labor. Specifically, we will concern ourselves with the (l) complexity of the division of labor as represented by the size of the communities, the intercommunity division of labor as represented in (2) metropolitan dominance and (3) functional specialization. Lastly, (4) we will concern ourselves with two dimensions of the ascribed division of labor in the United States and their influence upon our dependent variable. The two dimensions are the nonwhite population and female labor force of com— munities. Aggregate Approach" (unpublished M.A. thesis, Department of Sociology, University of Chicago, 1961), P- 16- llWe have accepted as an operational definition of communities the U.S. Bureau of Census' unit, ”places.” An explication of the meaning of "place” and our justification for using these units as communities will follow in a later chapter. 12We are indebted for our first insight into the possibility of using communities as units of observation as well as analysis in this type 0f problem to Otis Dudley Duncan and Leo Schnore. They suggested in an article, ”Cultural, Behavioral and Ecological Perspectives in the Study of Social Organization,” American Journal 9f_Sociology, 65 (September, 1959), 132—46, that it would be possible to characterize "whole aggregates- communities and even societies-as more or less crystallized.” A numb inconsistency searched the inconsistency the general p only deviatio: The pro formulating t' community, st represents a for scrutiny. has centered central conce Althou variations an M 13,229 1950), p. 231 14Hodg 15Sp1e l6 Stoe l7Sinc using the dix % the analyt Labor in inner hA.thesis, Frank A. Cler Ofémerican ( mm Richard 1 unpublished 10 A number of writers have discussed the notion of consistency— 13 inconsistency as an aggregate phenomenon and a few people have re— searched the idea.14 Most of those who have researched the consistency— inconsistency of aggregates have been concerned with selected samples of 15 Stoeckel's work is the 16 the general population of the United States. only deviation and is concerned with aggregate data from counties. The previous aggregate studies have been very instructive in formulating the present study. Nevertheless, it is our opinion that the community, still unexamined in terms of strata consistency-inconsistency, represents a statistically analyzable and sociologically relevant unit for scrutiny. While thus far concern with status consistency—inconsistency has centered on individuals and on the general population, an area of central concern to sociology, the community, has gone without analysis.l7 Although the present study will not attempt to examine concomitant variations and relationships that may result from communities with varying l3Ibid.; Amos H. Hawley, Human Ecology (New York: Ronald Press, 1950), p. 231. 14Hodges, loc. cit.; Spielman, loc. cit.; Stoeckel, loc. cit. 15Spielman, loc. cit.; Hodges, 10c. cit. 16 Stoeckel, loc. cit. 17Since the initiation of this thesis, some work has been done using the division of labor as an orienting concept and human ecology as the analytical framework. See: Frank A. Clemente, "The Division of Labor in American Communities: An Ecological Analysis” (unpublished M.A. thesis, Department of Sociology, University of Tennessee, 1969); Frank A. Clemente and Richard B. Sturgis, "Industrial Diversification of American Communities," Urban Studies (forthcoming); Frank A. Clemente and Richard B. Sturgis, ”The Division of Labor in American Communities" (unpublished paper, July, 1970). strata arrangt A first step : strate degree: dicate some 0: terns. Such . Some av studied commu‘ analysis. Du of structural of itself. I analyses are analyses but fully agree w to examine va as a justifia The or most aociolog familiar or c human ecology ecology. Th: orthodox vie! -—-—_....._.... 180:1 FE Study _o_f ditors (Chi 19Dun 206st Peterson and Ecology and 11 strata arrangements, such concerns appear highly relevant for sociology. A first step is to proceed with the task as has been outlined, to demon— strate degrees and patterns of strata consistency-inconsistency and in— dicate some of the variables which participate in determining the pat— terns. Such a task requires a human ecology—social morphology framework. Some authors have suggested that the findings of ecologists who studied communities could provide "the base” for further sociological analysis. Duncanl8 has since argued that the demonstration and analysis of structural arrangements of communities provides a useful end in and of itself. In fact, Duncan, along with Schnore, has argued that such analyses are not simply stepping stones for other more insightful 19 We analyses but are central to traditional sociological analysis. fully agree with the thesis of Duncan and Schnore. It is our intention to examine variations and patterns of status structures in communities as a justifiable examination in its own right. The conceptual use of an ecological framework will be familiar to most sociologists; However, it is possible social morphology is less familiar or clear. We need to interject here that what we refer to as human ecology has been pointed out by others as but one view of human ecology. Theodorsen refers to the approach we are using as the neo— orthodox view.20 One of the best articulators of this approach has been W l8Otis Dudley Duncan, ”Human Ecology and Population Studies, The Study_ of Population, Phillip M. Hauser and Otis Dudley Duncan, editors (Chicago: University of Chicago Press, 1959), pp. 678— 716. 19Duncan and Schnore, loc. cit. 20George A. Theodorsen, Studies in Human Ecology (New York. Row, Peterson and Company, 1961), pp. 129— 34; George A. Theodorsen,’ 'Human Ecology and Human Geography," Readings ip Contemporary American Otis Dudley D as a concern ' lation aggreg always, but 0 the other thr ferred to the complex.” Concer morphology an with morpholo Population fl larities of D Schnore . 23 D only a few br presented . ——-.._.__.._._. Sociology, Jc Adams and Con 2J‘Dunc his hesitancg issues sugge: properties ," Duncan mentit designation : In one of D111 acceptable 21 Organization Farris, edit‘ The term "so made to the 22Man Duncan and H 23Lee American _J_o_u also found 1 1965), pp, 1 12 Otis Dudley Duncan,21 who delineates the field of study of human ecology as a concern with the interdependence and reciprocal influence of popu— lation aggregates, technology,environment, and social organization. Not always, but often, social organization is the dependent variable while the other three variables serve as independent variables. Duncan has re— ferred to the interdependence of these variables as forming the "eco— complex.” Concerning social morphology, a brief but useful explanation of morphology and related developments, their similarities and differences with morphology, can be found in the translator's preface to Halbwachs' Population and Society.22 A second important work which links the simi— larities of Durkheim's social morphology to human ecology is provided by Schnore.23 Due to availability and excellence of the above sources, only a few brief, relevant comments concerning social morphology will be presented. Sociology, Joseph S. Roucek, editor (Patterson, New Jersey: Littlefield, Adams and Company, 1961), pp. 339—57. 21Duncan, loc. git. In the aforementioned work Duncan indicates his hesitancy to BEE the term ”ecosystem" so as to ”avoid prejudgment of issues suggested by the term 'system' . . . with equilibrium maintaining properties," ibid., p. 684. It is interesting to note that one year later Duncan mentions the usefulness of the term "ecosystem” as a "heuristic designation for the ecological complex," Duncan and Schnore, 1223 cit. In one of Duncan's more recent works, he finds the term "ecosystemfi__ acceptable enough to include in the title of the article, "Social Organization and the Ecosystem,” Handbook 9£_Modern Sociology, Robert Farris, editor (Chicago: Rand McNally, 1964), pp. 36—82. The term ”ecocomplex” will be retained in this thesis when reference is made to the aforementioned interdependent variables. 22Maurice Halbwachs, Population and Society, trans. Otis Dudley Duncan and Harold W. Pfautz (Glencoe: The Free Press, 1960), pp. 7—30. 23Leo F. Schnore, "Social Morphology and Human Ecology,” American Journal 2£_Sociology, 63 (May, 1958), 620—34. The article is also found in Leo F. Schnore, Ehg_Urban Scene (New York: Free Press, 1965) , pp. 1—28. Social separate but j initial develr of interest. championed to nurtured in t The to strates the s Social form is d mass of 1 they are tion of a lationshi the DOM it is con how citie by the so kind of b nature of Phenomena immediate Here, the interest and the 5 Single Sc 13 Social morphology and human ecology can be seen as having separate but parallel developments. They are parallel in time of initial development and identification with sociology24 and in areas of interest. Social morphology had its early development in France, championed to a great extent by Durkheim,25 while human ecology was nurtured in the United States. The following statement is from Durkheim's writings. It demon— strates the similarity in orientation of morphology and ecology. Social life rests on a substratum whose size as well as its form is determinate. This substratum is constituted by the mass of individuals who make up a society, the way in which they are distributed on the soil, and the nature and configura— tion of all sorts of material things that affect collective re— lationships. The social substratum differs according to whether the population is large or small or more or less dense, whether it is concentrated in cities or dispersed over the countryside, how cities and houses are constructed, whether the area occupied by the society is more or less extensive, and according to the kind of boundaries that delimit it. On the other hand, the nature of the substratum directly or indirectly affects all social phenomena in the same way that all psychological phenomena are immediately or ultimately related to the state of the brain. Here, then, is a whole group of problems which obviously are of interest to sociology and which——inasmuch as they refer to one and the same object—~must be placed under the jurisdiction of a single science. It is this science that we propose to call social morphology. The literature discussing these questions at the present time appears in various disciplines. Geography studies the territorial forms of states; history traces out the evolution of urban and rural groups; demography deals with everything involving the distribution of population; etc. We believe there is interest in drawing these fragmentary sciences out of their isolation, bringing them into contact with one another, and uniting them under a common rubric; in this way, they will gain a feeling of their unity. I I n o o a o o c a n o o n I I u l n a a o o a a 24Social morphology had a slightly earlier beginning, being elaborated by Durkheim in the latter part of the nineteenth century, while human ecology is usually identified as evolving during the second decade of the twentieth century. 25Durkheim's discussions of social morphology appear in various editions of L' Année sociologigue (old series). Social tional sc for them; investige political their frc inquire i are, how quently, the form analysis; see how i it quite conditior branches history z Schno agebetween: ecology.27 “010w with the field of weaknesses o cauSa1 relat m0rphology a an attempt a CouPled with will be Exam \ 26m 11(1897- 1 9 . 98) ~l 27Sci 28J. 1962): 1:. 2e 14 Social morphology, moreover, is not merely a simple observa- tional science which desaribes . . . forms without accounting for them; it can and must be an explanatory science. It must investigate the conditions responsible for variations in the political areas of peoples, the nature and significance of their frontiers, and the uneven density of population; it must inquire how urban groups arise, what the laws of their evolution are, how they are recruited, what their role is, etc. Conse- quently, it does not merely consider the social substratum in the form it takes at a given time in order to make a descriptive analysis; it observes the substratum in process of becoming to see how it is formed. It is not purely static science; rather, it quite naturally covers the movements from which result the conditions that it investigates. Moreover, like all other branches of sociology, it finds indiSpensable auxiliaries in history and ethnography. Schnore has provided an excellent argument demonstrating the link— age between social morphology, as discussed by Durkheim, and human ecology.27 Beshers has since referred to the identification of human ecology with social morphology as the "most significant development in the field of human ecology.”28 He accurately points out that one of the weaknesses of human ecology has been an "inadequate conceptualization of causal relationships.”29 Some of Durkheim and Halbwachs' work in social morphology attempt causal explanations. While the present thesis is not an attempt at "causal analysis" in the most rigid sense, social morphology coupled with human ecology does suggest direction in the relationships we will be examining. 26Emile Durkheim, "Morphologie sociale,” L'Annde sociologigue, 11 (1897—98), 520-21. Quote is also found in Halbwachs, pp. cit., pp. 9—10. 27Schnore, The Urban Scene, pp. 1-28. 28.1. M. Beshers, Urban Social Structure (New York: Free Press, 1962), p. 26. zglbid. If one the variables Schnore31 sug ferences in c Schnore has p neglected the 'hnce enviror garded as wet ables-~most l Schnor attempt will theframeworl In rex Social strat: consistency_; consistency_; degree of co] Plained) not thedeSTEe o “O“Sistency- est; the uni consistently. “fincome, 0 \ 3 0Haw 31Dun 323th 15 If one accepts human ecology (at least in part) as a concern with the variables of the ecocomplex at the community level30 (Duncan and Schnore31 suggest the inclusion of the societal level as well), the dif— ferences in orientation between ecology and morphology disappear. Schnore has pointed out that Durkheim's discussion of social morphology neglected the ecocomplex variable of environment. Schnore concluded that "once environment is brought into the picture, modern ecology can be re— garded as working with essentially the same array of independent vari- ables——most broadly, papulation, technology, and the environment.32 Schnore's conclusion is accepted in this paper. No further attempt will be made, except for clarification, to distinguish between the frameworks of social morphology and human ecology. In review, this study concentrates on a problem concerned with social stratification. The specific issue of the analysis is with strata consistency-inconsistency. There are three aspects of our concern with consistency~inconsistency which vary from most previous studies: (1) the degree of consistency—inconsistency is the dependent variable (to be ex— plained) not an independent variable; (2) we are concerned not only with the degree of consistency-inconsistency but also the patterns of consistency-inconsistency; (3) the third variation is perhaps the great— est; the units within which we are to observe and analyze strata consistency—inconsistency are communities. That is, the ranked orders 0f income, occupation, and education will be analyzed to determine their 3OHawley, pp, cit., p. 180. 31Duncan and Schnore, loc. cit. 32Schnore, The Urban Scene, p. 16. 16 degree and patterns of consistency—inconsistency among communities. The explanatory orientation for the analysis draws largely from human ecology and social morphology. These particular frames of reference are suited for dealing with aggregate data of the nature involved in the present study. CHAPTER II THEORY—~PART I Introduction William Dobriner made the following statement in regard to analyzing communities: Generally speaking, there are four analytical perspectives from which community organization can be sociologically de— fined: (1) the demographic, (2) the ecological, (3) the patterns of formal and informal social organization, and (4) the normative and ideological ethos which largely defines the specific character of the demographic, ecological, and organizational facets of the community. Using Dobriner's classification as a reference point, we are con- cerned, directly or indirectly, with parts of the first three perspec— tives. We will attempt to develop some conceptual webbing between the three perSpectives by providing a general, encompassing conceptual scheme.2 In fact, one of the three perspectives offers a frame of reference which can be inclusive of at least part of the other two. The lWilliam Dobriner (ed.), The Suburban Community (New York: G. P. Putnam's Sons, 1958), pp. xiv—xv. 2In regard to the fourth perspective Dobriner lists, we do not have space to elaborate, but must point out that from our perspective he con- fuses the verb usage in the sentence. We would have it read ”the norma— tive and ideological ethos which largely ii defined py_the specific charac- ter of the demographic, ecological, and organizational facets of the com— munity.” For general support of our perspective see: Alvin W. Gouldner and Richard Anderson, Notes on Technology_ and the Moral Order (Indiana— polis: The Bobbs-Merrill Company, Inc., 1962); Julian Steward, Theory of Culture Change (Urbana: University of Illinois Press, 1955); T. B. '__ Bottomore (ed.), Karl Marx: Selected Writings 13 Sociology 33d Social Philosophy (New York: McGraw—Hill Book Company, 1956); Emile Durkheim, The Division of Labor in Society (New York: The Free Press, 1964), B00k Two; Bronislaw MalinOWSki, A Scientific Theory_ of Culture (Chapel Hill: The University of North Carolina Press, 1944). l7 l8 orientation which offers the most in allowing us to develop an inte— grated conceptual scheme is human ecology. We are not suggesting that there is only one major conceptual framework in human ecology, but that the general sociological orientation which provides the best fit for our problem is found within the writings generally labeled ”human ecology.” Theory in ecology, like most other substantive areas in sociology, is referred to as ”just developing,” ”beginning to mature,” and ”in need of sharpening." As with most areas in sociology, the statements are semi- accurate appraisals of the situation. The position is depicted in the introductory remarks Duncan felt it necessary to make at the beginning of a presentation about human ecology to a group of scholars "outside” the area of human ecology. . . . These prOpositions are more or less plausible in terms of generalized notions about the nature of the human community. However, they could not be said to represent careful deduction from a body of rigorously organized theory. Nor can it be claimed that the verification of these propositions is satisfac— tory as yet, in the absence of extensive comparative studies. But at least they will illustrate some of the approaches and methods of contemporary human ecology.3 One may be caused to wonder why ecology which flourished so strongly in the first part of this century, under such dynamic personalities, and at outstanding institutions of higher learning has not yet deveIOped past the "just developing stage.” One very important reason for the state of theory in ecology is the strong descriptive orientation that imbued much of the early work in human ecology, at least in the United States. This descriptive orientation resulted, as Hawley has noted, from "a subordina— tion of interest in functional relations to a concern with the spatial Otis Dudley Duncan, "Population Distribution and Community Structure," Cold Harbor Springs Symposia pp_Quantitative Biology, 22 (1957), 357-71. l9 patterns in which such relations are expressed.”4 An example of Hawley's indictment is offered from a definition of human ecology by McKenzie. "Human ecology deals with the spatial aspects of the symbiotic relations of human beings and human institutions.”5 While ”spatial" aspects are one of the indicators that exhibit the action of an ecological complex in human society, they are not, in and of themselves, the main or only con- cern of ecologists.6 We find ourselves in agreement with the following position expressed by Gibbs and Martin, and the orientation of this paper will attempt to conform to this position. This explicit rejection of spatial analysis as the major concern of human ecology is essential to the revitalization and further development of a once promising discipline. The conception of human ecology as the study of sustenance organiza— tion appears to be more consistent with the nature of ecology in other fields and to be potentially more fruitful than spatial ecology for the develOpment of useful theory and meaningful empirical prOpositions. 4Amos Hawley, Human Ecology_(New York: The Ronald Press Company, 1950), p. 69. 5Roderick D. McKenzie, "Human Ecology," Encyclopedia p£_the Social §2$§2§g§3 Edwin R. A. Seligman, editor, 5 (1931), 314. It needs noting that Hawley has recently pointed out that McKenzie in his later notes, in reference to spatial patterns, indicated that they "should be subordinate and incidental to the analysis of sustenance relations." See Amos Hawley (ed.), Roderick D, McKenzie gp_Human Ecology (Chicago: University of Chicago Press, 1968), pp. xiii—xiv. 6To study only the Spatial patterns is to ignore Durkheim's third rule for the observation of social facts. "When, then, the sociologist undertakes the investigation of some order of social facts, he must en— deavor to consider them from an aspect that is independent of their in— dividual manifestations." Emile Durkheim, Th3_§gl§§_g£_300101081031 Method (New York: The Free Press, 1964), p. 45. 7Jack Gibbs and Walter Martin, "Urbanization and Natural Re- sources: A Study in Organizational Ecology," American Sociological Review, 23 (June, 1958), 267. 20 The idea that collective austenance activity and the related forms of social organization is of central interest to human ecology is not entirely new. Park's8 notions of an economic base to society founded upon symbiotic relationships is consistent with a stress on sus— . 9 . . tenance as were many of McKen21e's interests. What is new is the em— phasis being given, in various forms, to sustenance activities as a cen— tral concern of ecology. One major orientation in present day ecology which we have previously mentioned may appear to be divergent from an emphasis on sus— tenance. It is the ecocomplex-—population, organization, environment, and technology (POET).lO The POET scheme provides an orientation which suggests what variables ecology takes into account. If the writings 8Robert Ezra Park, "Human Ecology,” The American Journal pf Sociology, 42 (July, 1936), 1—15. Roderick D. McKenzie, "The Scope of Human Ecology," Publications g£_£hg American Sociological Society, 20 (July, 1926), 141—54; and Roderick D. McKenzie, ”Demography, Human Geography, and Human Ecology," Thg Fields gpd_Methods pf Sociology, L. L. Bernard, editor (New York: Ray Lang and Richard Smith, 1934), Chap. 4. The preceding references are both reprinted as chapters two and three in Hawley, Roderick D. McKenzie 9p_Human Ecology. _ lOLeo F. Schnore, "Social Morphology and Human Ecology," American Journal pf Sociology, 63 (May, 1958), 620-34; Otis Dudley Duncan, "Human Ecology and Population Studies,” Th9 Study pf_Popu1ation, Phillip M. Hauser and Otis Dudley Duncan, editors (Chicago: The University of Chicago Press, 1959), pp. 678—716; Otis Dudley Duncan and Leo F. Schnore, "Cultural, Behavioral, and Ecological Perspectives in the Study of Social Organization,” American Journal pf_Sociology, 65 (September, 1959), l32~46; Otis Dudley Duncan, "From Social System to Ecosystem," Sociological Inguiry, 31 (Spring, 1961), 140—49; Otis Dudley Duncan, "Social Organization and the Ecosystem," Handbook 2: Modern Sociology, Robert E. L. Paris, editor (Chicago: Rand McNally and Company, 1964), pp. 36—82. 21 which include a discussion of the POET scheme are examined in any detail, sustenance activities and up being the cloth upon which specific inter— dependencies among the variables are embroidered.ll In one way or an— other, the social organization of man which is directly related to collec— tive Sustenance processes remains central. Based on the preceding discussion we are going to introduce and explicate12 three very broad concepts-—division of labor, competition, and dominance (as one form of collective power). These three concepts will provide, at a very general level, an overall scheme by which we can order the more specific concepts we will introduce. These specific concepts that are central to our theoretical framework and that will be ordered within the framework suggested by the broader concepts are: size of community, metropolitan dominance, functional specialization, and two ascribed bases of the division of labor, which are represented in llSee especially Duncan and Schnore, ”Cultural, Behavioral, and Ecological Perspectives in the Study of Social Organization," pp. 132- 46; and Duncan, Handbook pf_Modern Sociology, pp. 36—82. Our use of the term "explication" is in keeping with the general explanation of the notion given by Hempel. See Carl C. Hempel, Funda- mentals pf_Concept Formation 13 Empirical Science (Tenth Impression; Chicago: The University of Chicago Press, 1969). For a specific state— ment which characterizes his treatment, the following quote is from page 12: "An explication of a given set of terms, then combines essential aSPects of meaning analysis and of empirical analysis. Taking its de— parture from the customary meanings of the terms, explication aims at re— ducing the limitations, ambiguities, and inconsistencies of their ordinary usage by propounding a reinterpretation intended to enhance the clarity and precision of their meanings as well as their ability to function in hypotheses and theories with explanatory and predictive force. Thus understood, an explication cannot be qualified simply as true or false; but it may be adjudged more or less adequate according to the extent to which it attains its objectives." For another discussion of explication, see Richard G. Dumont and William J. Wilson, ”Aspects of Concept Forma— tion, Explication, and Theory Construction in Sociology," American Sociological Review, 32 (December, 1967), 985—95. 22 communities as females in the labor force and percentage nonwhite. It is our intent to develop the general concepts first and demonstrate that the more specific concepts can be better understood within the notion of a competitive, economically based division pg labor. Division of Labor For those who are interested in what is somewhat hazily called "macro" sociology there are few other cencepts of greater theoretical utility than "division of labor." There is no other form of social organization which provides as encompassing and as essential of an element to societal existence than what is loosely termed "the division of labor." For sociologists who are in one way or another concerned with social organization there appears to be no concept which offers a greater range of generality and abstraction. We are aware, as with most concepts of any longevity in sociology, there is a lack of agreement and precision in the various usages of the concept. Most important, there has been little empirical research done with the division of labor as a major analytical variable, one of our main complaints in this thesis. However, it is in response to this particular complaint that we will be giving much of our attention. We are not suggesting that there is great explanatory power emanating from the usage of division of labor as a concept. It does provide, however, a major orienting and perspective setting frame of reference. Much of the lament in sociology over the lack of integrated and developed theory is due to the fact that there has been no framework Within which to integrate on a very broad scape. Much debate (useful in and of itself) has centered around strictly analytical frames of 23 reference, e.g., functionalism, conflict, and now systems. Division of labor denotes a concept which is translatable into comparative, empirical analysis. It can be logically construed into a broad substantive orientation, which will permit the integration of various less general theoretical schemes. We present these notions with very short discussion and little de- tail. Given the limited space available, within the present context, we want to make clear that we do not view the division of labor as the only concept to possess such integrative potential, but we do want the reader to be aware that we assume the concept has the capacity (presently rather dormant) to perform such a function. It is on this assumption that we will use the division of labor as our most general concept, providing overall integration of our more specific concepts and their theoretical relationships. It is to some analytical dimensions of the concept, im— portant for this thesis, that we now give our attention. As we have emphasized, there is probably no single concept in sociology that is used more frequently in an array of substantive sub— fields than the division of labor. It is the type of concept which has possessed the amount of generality to allow its use in talking of primi— tive societies to explaining aspects of modern, technologically based, bureaucratized nations. Nevertheless, the concept's empirical anchorages have been minimal. The concept of division of labor has had a somewhat strange career in the history of sociology. On the one hand, the concept has achieved wide acceptance, particularly since Durkheim's classic treatment. On the other hand, it is rarely employed in the genera— tion of testable hypotheses. This is even true for the field of human ecology where, like competition, the concept is often in— voked in pure theory but remains in the background as far as research is concerned.13 Labovitz has suggested that: "The paucity of empirical treatment of the division of labor stems, at least in part, from (1) inadequate conceptual analysis and (2) a lack of rigorous specification of its dimensions."14 There has been a general orientation to regard the division of labor as synonymous with occupational differentiation. Gibbs and Martin have recently attempted to explicate the concept of division of labor and develop testable hypotheses. In the process they have stressed the importance of including more than the idea of occupational differentia— tion. There are two general ideas associated with the concept. First, there is the suggestion of occupational differentia— tion. However, more is involved than individuals "doing dif— ferent things." In addition to differentiation there is functional interdependence. A second idea associated with the concept is often confused with the first. In the process of differentiation a person's occupational status may be determined, more or less, by bio- logical characteristics, ethnic—caste status, or territorial location. These distinctions may be called the bases pf_£hg division of labor, but they are not to be confused with the degree of_fhe division of labor. Occupations in a society may be closely correlated with non occupational distinctions but, at the same time, the number of different occupations may be small. This means a low degree of division of labor. [Emphases ours.] There are two central notions in the preceding comments, to which we have given emphasis, that are important for the further development of m l3Jack P. Gibbs and Walter T. Martin, ”Urbanization, Technology, and the Division of Labor: International Patterns,” American Sociologi— 331 Review, 27 (October, 1962), 669' l4Sanford Labovitz, ”Technology and the Division of Labor" (un— published Ph.D. dissertation, Department of Sociology, University of Texas, 1963), p. 2. 15Gibbs and Martin, "Urbanization, Technology, and the Division Of Labor: International Patterns," p. 669. 25 our theoretical framework. The notion that division of labor involves functional interdependence and that analysis of the bases, as well as the degree, contributes toward understanding the total picture of a division of labor. First, the implication of stressing functional interdependence. No community in the continental United States, today, is completely isolated from other communities. While the interdependence between communities may be more visible in a suburban and metropolitan re— lationship, the interdependence between communities is no less real 16 for more rural communities. Some authors have discussed and ex- plored the influence of communities on one another, especially those interested in metropolitan dominance, and more recently and popularly 17 those who emphasize intraregional interdependencies. The more recent emphasis includes the interests of geographers, economists, and various 16For a book which takes as its central theme the dependence of a small more rural community on the more urban communities see Arthur J. Vidich and Joseph Bensman, Small Town in Mass Society (Princeton: Princeton University Press, 1958). See also, Donald J. Bogue, The Structure g£_the Metropolitan Community—jA Study of Dominance EE§.§EE" dominance (Ann Arbor: Horace H. Rackham School of Graduate Studies, University of Michigan, 1950); Otis Dudley Duncan and Albert J. Reiss, York: John Wiley and Sons, Inc., 1956); C; J..Galpin, The Social Anatomy 3: EB Agricultural Community, Research Bulletin 34 (Madison: University of Wisconsin Agricultural Experiment Station, May, 1915). l7Otis Dudley Duncan and Others, Metropolis and Region (Baltimore: The Johns Hopkins University Press, 1960); Donald J. Bogue, "An Outline of the Complete System of Economic Areas,” American Journal of Sociology, 60 (September, 1954), 136—39. The Census Bureau's adoption and use of the concept "economic area" as designating a useful statistical area is a good example of acceptance of intraregional interdependencies. U.S. Bureau of the Census, U.S. Census of Population: 1960. Selected Area Reports. State EconomiE_Areas. Final Report PC(3)—lA (Washington, D.C.: U.S. Government Printing Office, 1963). 26 urban analysts. Their general recognition of functional interdependen— cies can be found within the discussions centering around a number of the more frequently used concepts, e.g., ”basic—nonbasic economic 1118 11 11.19 and ” base, central place theory, system of cities."20 The con— cepts are not necessarily used exclusive of one another, but each does represent a central organizing concept for a number of authors. The im— portance for our analysis is that each of the concepts stresses functional interdependencies among communities. If, as we stated earlier, functional interdependency among units is one aspect of the division of labor, then quite clearly we have a division of labor among communities. Given their political, economic, and organization import, communities represent im— portant units in a societal division of labor. Recently Gibbs and Martin21 analyzed and discussed sociologically 18John W. Alexander, ”The Basic—Nonbasic Concept of Urban Economic Functions,” Economic Geography, 30 (July, 1954), 246—61; Charles M. Teibout, ”The Urban Economic Base Reconsidered," Land Economics, 32 (February, 1956), 95-99; Walter Isard, Location apd Space—Economy (New York: John Wiley and Sons, 1956). H 19Edward Ullman, "A Theory of Location for Cities, American Journal pf Sociology, 46 (May, 1941), 835—64; John E. Brush and Howard E. Bracey, "Rural Service Centers in Southwestern Wisconsin and Southern England," Geographical Review, 45 (October, 1955), 559—69; Brian J. L. Berry and William L. Garrison, ”The Functional Bases of the Central—Place Hierarchy," Economic Geography, 34 (April, 1958), 145—54; August Losch, The Economics pf_Location, trans. William H. Woglom with the assistance of Wolfgang F. Stopler (New Haven: Yale University Press, 1954); Walter Christaller, Central Places ip_Southern Germany, trans. Carlisle W. Baskin (Englewood Cliffs: Prentice—Hall, 1966). 20Eric E. Lampard, ”The History of Cities in the Economically égvanced Areas,” Economic Development and Cultural Change, 3 (1954-55), 3-29. 21Gibbs and Martin, "Urbanization, Technology, and the Division of Labor: International Patterns," p. 669. 27 relevant dimensions of an international division of labor, but, to our knowledge, the advancement of a conceptual framework and an analysis of the division of labor among communities on a societal basis has gone un— explored.22 Those who have come closest within sociology to the approach we are suggesting are the ecologists in studying metropolitan dominance and in analyzing geographical regions.23 The point of relevance is that the division of labor has been used as an ancillary concept, to help in developing such concepts as metropolitan dominance and to some extent functional specialization of communities. Our contention is that the concept of division of labor is the more general concept, the one which can provide a conceptual webbing to encompass such notions as metropolitan dominance and functional specialization. It is not a question of seman— tics, only, but a question of theoretical development. The issue now be— comes what are the major factors shaping the functional interdependencies (division of labor) among communities? The possibilities we will explore (as independent variables) in this analysis are size of community, metro— politan dominance, and functional specialization. It is our intent in 22Since the initiation of the present study, there has been an analysis of industrial diversification among communities within the con— tinental United States. The analysis was conducted within the ordering conceptual framework of the division of labor. See Frank A. Clemente, ”The Division of Labor in American Communities: An Ecological Analysis" (unpublished M.A. thesis, Department of Sociology, University of Tennessee, 1969); Frank A. Clemente and Richard B. Sturgis, "Population Size and In— dustrial Diversification,” Urban Studies (forthcoming); and Frank A. Clemente and Richard B. Sturgis, ”The Division of Labor in American Com— munities: An Ecological Analysis” (unpublished paper, July, 1970). 23Some authors have explicitly used the term ”division of labor” in their analyses. Examples from three of the major writers in this area can be found in Bogue, The Structure of the Metropolitan Community, P- 3; Rupert Vance and Sara Sutker, The Urban South (Chapel Hill: The University of North Carolina Press, 1954), p. 114; Roderick D. McKenzie, 28 what follows to develop a conceptual framework which depicts community size, metropolitan dominance, and functional specialization as major organizational influencers in the functional interdependencies (division of labor) among communities. The second important "notion" outlined by Gibbs and Martin in our earlier quote was that the bases as well as the degree were important in the total picture of the division of labor. The concepts of size, metro— politan dominance, and functional specialization represent bases of the division of labor. In the present analysis they serve as the bases for influencing certain morphological characteristics of communities, our dependent variable (the status consistency-inconsistency of communities). One additional dimension needs to be added to our bases of a societal division of labor among communities. The previously mentioned three concepts are all identifiable as traditional ecological concepts and variables. There is another dimension which is a characteristic of the occupational force involved in the division of labor; it can be viewed in terms of a community characteristic; and is also an important basis of the division of labor, which in turn influences our dependent variable. We refer to the nonwhite population and female labor force. Each of the concepts we have introduced as independent variables will be developed in turn. They have been introduced here to support our contention of the utility of the division of labor as a major orienting and integrating concept, within which can be logically tied together a number of different subconcepts. In terms of communities, the concepts we have introduced point to the functional The Metropolitan Community (New York: McGraw—Hill, 1933), p. 113. 29 interdependencies among communities and to the bases upon which some of these functional interdependencies are founded. Before we discuss each of the subconcepts, it will be necessary to introduce two other concepts, which along with the division of labor serve as a major theoretical orientation. The concepts are competition and dominance (as a form of collective power). To introduce the dis— cussions it is only necessary to point out the obvious; the division of labor in any industrial, urban society, and certainly in the United States, is fundamentally economic. Further, in the United States there has been a basic ideology supporting a competitive economic system. Competition The concept of competition has often been used by ecologists, but seldom clearly explicated. For Park, Burgess, and McKenzie the notions of competition and dominance, centering around the economic dimensions of communities and society, were of central concern. We are taking an important divergence from these ”classical" ecologists, at least from part of their writings. Park argued, conceptually, that underlying all social order was an unplanned biotic order or level. The next level, based upon the biotic, but forming the basis of the political and moral 24 social orders, was the economic. What we want to clearly set forth is the notion that the biotic dimension is inextricably bound up with the 25 economic dimensions of man's collective existence. Within the framework .1..._______________ 24Robert Park, ”Human Ecology,” American Journal pf_Sociology 42 (July, 1936), 1—15, reprinted in George A. Theodorsen (ed.), Studies ip Human Ecology (Evanston: Row, Peterson and Company, 1961), pp. 22—29. 25One of the basic criticisms from the early critics of ecology 30 of the division of labor, economic ties become one of the more important fibers woven into the web of life. In the social organization (division of labor) which results from and continues as a collective adaptation of man in providing, maintaining, and distributing sustenance, economic ties are basic. It has been unfortunate for the further growth and development of human ecology that economic analyses almost always get translated into ideas of rational and planned patterns of activities.26 This is especially true when the idea of competition enters into the analysis. Nevertheless, there is much that enters into a societal division of labor, involving economic ties among communities, the re— lationships involved and their outcomes being influenced by competition 27 It is some in the economic order, which are nonrational and unplanned. of these unplanned factors, at the community level, which operate as our independent variables, i.e., community size, metropolitan dominance, and functional specialization. It is important to make two points at this particular juncture. First, the early human ecologists in the United States argued the was that they saw an "unreal" separation of the biotic and other levels of man's existence. Our stress on the inseparability of these levels will allay some of the criticism. See, for instance, Milla A. Allihan, Social Ecology (New York: Columbia Press, 1938). 26Rutledge Vining, in discussing Christaller's notions of urban spatial location, points out the impossibilities, in any practical sense, of invoking notions of rationality in understanding the interdependencies among communities. For Vining's insightful analysis see ”A Description of Certain Spatial Aspects of an Economic System,” Economic Development and Cultural Change, 3 (1954—55), 160—69; see also, Duncan, "From Social System to Ecosystem,” p. 142. 27In the present analysis we are referring to the societal division of labor on a community basis, but there is no reason to limit our state— ment to that level. The statement holds for the diviSion of labor in general. 31 importance of competition for space.28 Closely allied with the spatial dimension was economic competition. In fact, the competition for space (land) was in terms of its potential economic benefits. In the attack on the descriptive, spatial, and analogy reification aspects of human ecology, the critics indiscriminately attempted to discard all of human ecology. Gibbs and Martin have suggested that there was ”a tendency on the part of sociologists, in their reaction to economic and geographic determinism, to throw the baby out with the bath.”29 30 is the grease that lubricates Since some form of economic system the most simple to the most complex division of labor, to ignore its im- portance is to conceptualize an immobile division of labor. What we want to stress here is that some form of exchange (economic system) arises to handle the sustenance producing, maintaining, and distributing activities both within populations and between populations. These activities we 28Robert E. Park, Ernest W. Burgess, and Roderick D. McKenzie (eds.), The City (Chicago: University of Chicago Press, 1925). 29Jack P. Gibbs and Walter T. Martin, ”Urbanization and Natural Resources: A Study in Organizational Ecology," American Sociological Review, 23 (June, 1958), 266—67. 3OBronislaw Malinowski, Crime gpd_Custom i2 Savage Society (London: Routledge and Kegan Paul, Ltd., 1926), pp. 39—45; Claude Levi— Strauss, ”The Principle of Reciprocity,” Sociological Theory: A_Book of Readings, Lewis A. Coser and Bernard Rosenberg, editors (Third Edition?“ Toronto: The Macmillan Company, 1969), pp. 77—86; Raymond Firth, Primitive Pglynesian Economy (Hamden, Conn.: Archon Books, 1965). In the preceding references we have referred to "economic system" in very general terms. What we want to stress is that some form of ex— change (economic system) arises to handle the sustenance gaining, main— taining, and exchanging activities. 32 refer to as processes of the division of labor. In the society under analysis in the present research, the monetary dimensions of the eco— nomic system are obvious. A second point is that we do not want to present the appearance of accepting a completely unplanned, competitive economic order, impos— ing its effects helter skelter on community structure. For our analysis, competition is one concept which fits logically into our theoretical framework, but not to the extent that we can forget historical develop- ments, cultural variations, planned industrial locations, and other fac— tors that play influential parts in structuring communities.31 That any knowledgeable sociologist should ppp forget such factors is most heartily accepted by us. On the other hand, that we can neglect them as unimportant for our particular analysis is also accepted by us, and is the position from which we are developing the theoretical frame- work of this thesis. If the view of a competitively oriented division of labor can be accepted, we can present some of the implications of such a system at the community level. Our attempt to discuss and logically move from concern with division of labor to competition and to systems of community stratification, as we have just indicated, does not give attention to the rational aspects of man's collective existence. Let us emphasize the ‘———-—————_~____—_ 31See, for instance, William Form, ”The Place of Social Structure in the Determination of Land Use,” Social Forces, 32 (May, 1954), 317—24; Leonard Cottrell, "Death by Dieselization,” American Sociological Review, 16 (June, 1951), 358-65; Walter Firey, ”Sentiment and Symbolism as Eco— lOgical Variables," American Sociological Review, 10 (April, 1945), 140— 48; Evon Z. Vogt and Thomas F. O'Dea, ”A Comparative Study of the Role of Values in Social Action in Two Southwestern Communities,” American Sociological Review, 18 (December, 1953), 645-54. 33 idea that out of the processes by which men attempt to collectively sus— tain themselves, unplanned as well as planned competition ensues. Within a competitive network, positions of dominance and power may be obtained inadvertently as well as rationally. While Weber had in mind relation— ships among individuals and organizations, his following comments con— cerning competition are instructive. IE may 13 various ways, EE.EE unanticipated consequence g£.g course pf_socia1 action and its relevant conditions that certain types of social relationships . . . will be adversely affected. in their opportunities to maintain themselves or to arise. [Emphasis ours.] And at another point he made a comment which is more directly con— ceivable at the community level: Even on the utopian assumption that all competition were com— pletely eliminated, conditions would still lead £2.E latent process of selection, biological or social, which would favor thet types _best adapted_ to the conditions, whether their relevant qualities were mainly determined by heredity or by environment. [Emphasis ours.] Given the division of labor among communities, and given the economic dimensions of the division of labor, the question can now be asked: are there factors of an unplanned or "latent" nature which place communities in dominant or advantageous positions within the competitive, economic division of labor? We intend to explore some of these possibili— ties and their outcomes, specifically as they relate to the general status consistency—inconsistency of communities. 32Max Weber, The Theory_ of Social and Economic Organization, trans. A M. Henderson and Talcott Parsons, Talcott Parsons, editor (Glencoe: The Free Press, l947),p . 135. 33. Ibid., p. 134. 34 Dominance The concept of dominance along with competition was pivotal to the classical ecologists and has remained important in the study of metropolitan dominance. Duncan and Schnore34 have recently suggested the similarity between the concepts of dominance and power. Neverthe— less, almost all discussions of power in sociology have been at the inter— personal, intracommunity, or organization level. Duncan and Schnore point to the need for developing the concept in general and suggest its possible usage within an ecological framework. One treatment of the concept which can be adapted for analysis of power among collectivities (populations), of the nature involved in this thesis, is the classic discussion of the concept by Robert Bierstedt.35 He suggests that "power would seem to stem from three sources (1) numbers of people, (2) social organization, and (3) resources.”36 Given our par— ticular problem and Bierstedt's analysis the linkage is apparent. (1) "Numbers of people" is directly translated into size of popula— tion. Within the present framework, economic dominance (power) within the division of labor is maintained, generally speaking, by larger communities. In fact, Duncan and Reiss in their 1950 census monograph arrived at the following conclusion: ”Of all the differences among communities of dif— ferent size revealed in this study, perhaps the most striking is the pro— nounced direct relationship between size of place and income."37 -————.__..________ 34Duncan and Schnore, ”Cultural, Behavioral, and Ecological Pers- pectives in the Study of Social Organization,” p. 139. 35Robert Bierstedt, ”An Analysis of Power,” American Sociological Review, 15 (December, 1950), 730-38. 351bid., p. 737. 37Duncan and Reiss, 92. cit., p. 103. 35 (2) An advantageous social organization which gives the competitive edge (power) to a community is closely bound to size. It is generally argued that the change in organization which develops with size is one of the main reasons size makes a difference. It is this very argument that Bogue used in stating the thesis of metrOpolitan dominance. The metropolis is usually the largest and most complex (the farthest removed from the ”average” city) of all of the cities in the territory. Because it is able to assembly cheaply a varied array of raw materials and products from all parts of the world; because a large number of specialized components and skills are required to sustain human beings at their present level of living; because up to a certain point machine production increases in efficiency with an increased scale of Operations; and because certain mutual benefits appear to accrue to business enterprises from their location in proximity to each other, the large city is able to produce and distribute more varied goods and services than is a smaller city. The more Specialized the goods, and the more the goods are amenable to mass production, the greater these industrial and commercial advantages of large cities seem to become.38 Gras also argued for the dominance (power) of metropolitan centers in terms of organizational superiority. We may think of metrOpolitan economy as an organization of peOple having a large city as nucleus. Or we may put it this way, metropolitan economy is the organization of producers and consumers mutually dependent for goods and services, wherein their wants are supplied by a system of exchange concentrated in a large city which is the focus of local trade and the center through which normal economic relations with the outside are established and maintained. While the comments of Bogue and Gras include both the dimensions of size and location, the contributions of size toward superior organiza— tion, which in turn gives competitive economic advantage, is readily apparent. 38Bogue, The Structure gf_the Metr0politan Community, pp. 5—6. 39N. S. B. Gras, AE_Introduction £2_Economic History_(New York: Harper and Brothers, 1922), p. 184. 36 (3) The third factor in Bierstedt's analysis is resources. Al— though, in our present analysis we do not expand upon this point to a great extent, its possibilities are easily seen. What resources are most important for economic advantage will vary with time and are re— lated to many other faCtors; nevertheless, the question is one that can be explored empirically. We can suggest the obvious advantageous re— sources such as a deep sea port (a natural resource“+0 or a large manu— facturing base (man—made or technological resource).41 Factors one and two, size and organization, are taken into account when we discuss the influence of size and metropolitan dominance. Factor three, resources, is at least partially taken into account when we dis— cuss functional specialization. All three factors, of course, are inter— related in rather complex ways, but they are useful concepts for under— standing the dimension of power or dominance among communities. In fact, they become central concepts in trying to understand the dynamics of the division of labor in a free market society. Relation to Ecocomplex Thus far we have introduced three broad concepts——division of labor, competition, and dominance——as central concepts, which will provide meaning for our more specific concepts and for the derivation of our hypotheses. Since the ecocomplex is a frequently used conceptual 4OLeo F. Schnore and David W. Varley, ”Some Concomitants of Metro~ politan Size,” American Sociological Review, 20 (August, 1955), 408—14. 41Lampard, op} cit., pp. 92—102; Edgar M. Hoover, The Location of Eggnomic Activity (New York: McGraw~Hill Book Company, Inc., 1948). 37 scheme and mnemonic device (POET) for suggesting what human ecology is about, a few statements indicating its relevance for the present study are in order. 42 who have been the leading proponents of the Duncan and Schnore, ecocomplex, have suggested in general and provocative terms, ways of con— ceptually utilizing the POET scheme. The point we want to make is that up to the present time POET functions as a frame of reference, a point of orientation for a number of macro oriented sociologists. The thing that POET is not, and that to our knowledge its proponents have not claimed 44 for it, is a theory.43 In Merton's categorization it fits what he calls "general sociological orientation" (at the macro level). Stoeckel45 used the POET scheme in setting the stage for developing his theoretical framework, but there is no theoretical development of the POET scheme 42See, especially, Duncan and Schnore, "Cultural, Behavioral, and Ecological Perspectives in the Study of Social Organization,” Duncan, ”From Social System to Ecosystem;" and Duncan, "Social Organization and the Ecosystem.” 43The statement of the POET scheme probably comes closest to a statement of theory in the Handbook, but it still remains a broad inte— grative conceptual scheme. 44Robert K. Merton, Social Theory and Social Structure (Revised and Enlarged Edition; Glencoe: The Free Press, 1961), pp. 87—89. A short statement from Merton's discussion will indicate the relevance of his categorization: "Much of what is described in textbooks as socio— logical theory consists of general orientations toward substantive materials. Such orientations involve broad postulates which indicate types of variables which are somehow to be taken into account rather than Specifying determinate relationships between particular variables. In— dispensable though these orientations are, they provide only the broadest framework for empirical inquiry.” 45John Stoeckel, ”The Impact of Metropolitan Dominance Upon the Status Structure and Status Consistency of Rural—Farm and Urban Popula— tions” (unpublished Ph.D. dissertation, Michigan State University, Department of Sociology, 1966), pp. 7—lO. 38 itself, i.e., there are no interlocking propositions which link the concepts of the ecocomplex. What we do not want to do is depricate the development and utility of the ecocomplex. It has, after all, been ex— tremely important in guiding our own thinking. What we do want to demon- strate is that our preceding discussions provide a good conceptual fit, and, consequently, theory developing potential for the POET scheme. We will demonstrate the fit by looking, briefly, at each component of the ecocomplex. (I) As Duncan and Schnore46 have pointed out, organization, al— though not necessarily, most frequently is viewed as a dependent vari— able. The morphology (organizational structure) of communities in terms of status consistency-inconsistency is our dependent variable. (2) Population is treated by us in terms of size, quite in keeping with the ecocomplex. Important additional factors have been introduced in terms of viewing communities as interdependent within a societal division of labor and size possessing certain competitive economic, I advantages. What we have done is provide some of the "why" (at least in one case) for population influencing organization. It is one thing to suggest that population influences organization; it is quite another to suggest and explore the dynamics involved. Nevertheless, it is these dynamics that relate concepts in the form of propositions and generate theory. 47 (3) The third concept of the ecocomplex is environment. The 6 Duncan and Schnore, ”Cultural, Behavioral, and Ecological Perspectives in the Study of Social Organization,” p. 136. 47Duncan has a useful discussion of regional analysis as one attempt to deal with the concept of environment. See Hauser and Duncan (eds.), The Study 9§_Population, pp. 701—06. 39 concept of environment is a weak concept in ecology and in sociology generally. It is one of those terms which has frequently served as a residual category into which are dumped various and sundry leftovers. For example, such statements as ”X is influenced by variables A and B and 'general environmental forces,'” have not been uncommon. Without engaging in a lengthy critique and explication of environment let us note that region and distance have frequently been used as indicators of en— vironment when community comparisons have been made. While we will have occasion to use region as a control variable in our analysis, we are in 48 observation that environment is most simply inter— keeping with Duncan's preted as space (distance). Distance between communities always in terms of metropolitan dominance, fits, then, into the ecocomplex as "E" (environment). (4) There is one remaining concept in the ecocomplex, technology. For our analysis technology is most clearly conceived in terms of the in— dustrial bases of communities, more specifically, as functional speciali— zation. Unfortunately technology as a concept suffers from the same prob— lem as environment; it is frequently used but only, if ever, vaguely de— fined.49 Even when it has been used as a variable within the ecocomplex scheme it has generally remained ill defined. What we want to make clear is that our analysis includes the four concepts of the ecocomplex and translates them into variables for analysis. This aside is important given the frequent usage of and -——-———_—.—.._.__.— 481bid., p. 685. 49For recent treatment of and bibliography on technology, see Harvard University Program on Technology and Society, Fourth Annual Report, 1967—1968 (Cambridge: Harvard University Program on Technology and Society, 1968). 40 reference to POET. What we also hope to have made clear is that the general concepts of dominance and competition within a competitive, economically based division of labor extends the notion of POET. We move from a perspective which says "look here are the important concepts” to one which says ”look here is one way of understanding how the concepts become interrelated." We have attempted to develop a broad conceptual framework which provides what Gibbs and Martin50 have called the ”connecting mechanisms.” The general discussion and development of our theoretical frame— work up to now, provides the basis for the section which follows. While the statements in the following section may appear able to stand by them— selves, their meaning is entirely dependent on the context of what has preceded them in this thesis. Propositional Statements of Theory The remainder of this chapter includes a restatement, in succinct propositional form, of our basic theory. We are not suggesting that such a presentation makes the statements of relationship we are interested in any more ”real” or "true” than a rambling discussion form of presentation. The advantages of stating a theory in a propositional format are: (l) clarity, (2) simplicity, and (3) Vulnerability to justified criticsm.51 50Jack P. Gibbs and Walter T. Martin, "Toward a Theoretical System of Human Ecology,” Pacific Sociological Review, 2 (Spring, 1959), 3. 51A discussion which gives general support for what we are suggest— ing, the mechanisms involved in presenting propositions, and some of the ensuing problems and benefits can be found in Hans Zetterberg's 93_Theory and Verification in Sociology (Third Enlarged Edition; New York: The Bed— minster Press, 1965), Chapters 4 and 5. 41 Similar goals can be pointed to for all three of the previous points, but basically it comes down to stating the conceptual framework without a lot of verbal trimmings. The garnishment of many theories makes them more tasty, and may be the only thing that makes some of them conceptually palatable. It is our contention that scientific theory can develop only to the extent that we clearly know what the theory is saying. While over— simplification of complex relationships is always a threat, there must be an attempt to state the relationships in their clearest and most simple form.52 An outcome of stating theories in the form we have indicated is that other analysts can more readily evaluate, bggh logically and gm— pirically, the proposed system of explanation. Our propositions, which have resulted from earlier discussions in this chapter, are very general in scope. They are what Zetterberg53 calls "theoretical propositions.“ Due to their ”high informative value," these propositions differ from ”ordinary propositions.” It will become apparent in what follows, that lower level propositions (ordinary propo— sitions and hypotheses), can be fitted within the more general, theoretical propositions. The weaving of empirically testable propositions within the more general propositional network can be done both deductively and inductively. 52Simplicity and parsimony have been suggested as criteria for evaluating theory. Those theories which can (I) simply stated, (2) en— compass a greater range of phenomena, are valued above those theories which do not meet these criteria as well. See Melvin H. Marx, ”The General Nature of Theory Construction," in Theories in Contemporary Psychology, Melvin H. Marx, editor (New York: The Macmillan Company, 1963), pp. 19-21. 53Zetterberg, 22. cit., p. 80. 42 The first proposition is very basic and one that will be readily accepted. We believe its simplicity is more apparent than real, but it serves us, only as an initial step in developing our theory. Proposition I: The obvious empirical support which abounds for this proposition would allow its acceptance as fact by most observers. What is important and has ramification for sociology is a parallel proposition and a third proposition closely related to it. Proposition II: If man is £3 live collectively (a§_a population), hg must engage in collective sustenance activities. Proposition III: If man engages in collective sustenance activities, a division of www- It will be useful to clarify a number of issues at this point: (I) It is with propositions II and III and their logical union—— if man is to live collectively (as a population), a division of labor will occur——that Durkheim was concerned in The Division of Labor.54 What he did was to explore the effects on the division of labor resulting from changes in some dimensions of the population, e.g., size, physical and social density.55 Again, let us restate that the propositions we are setting forth are broad in scope. As we develop (deduce and induce) 54Emile Durkheim, The Division 9f_Labor in Society (New York: The Free Press, 1964). 55Ibid., Book Two; Schnore, loc. cit. 43 more specified propositions, capable of refined empirical analysis, within the framework of these more macro propositions, scientific theory can evolve. Durkheim's work was a move in that direction. If we stop with very general and encompassing propositions we border between sociology and social philosophy or history, but not scientific (empiri— cally anchored) theory. 0n the other hand, if we step with the specific, ordinary propositions, we end up with, at best, empirical generaliza— tions.56 (2) The emphasis on the division of labor as a phenomenon of a population makes it important for sociology in general. If the emphasis is on the population as a whole it becomes of specific interest to human ecology. That the interests of human ecology are basically population interests has been clearly articulated by a number of ecologists.57 Further, Schnore58 has ably demonstrated the centrality of ”the division of labor” within human ecology, and the connection is also apparent in the previous development of our theory. The bond may become clearer if we give re-emphasis to two points: (a) ”the primary focus of ecological attention is logically the organization of functional relationships,"59 and (b) that the particular functional relationships are those centering around sustenance activities. We have, then, a population (in our case 56 Merton, 22, cit., pp. 95-96. 57Park, op. cit., pp. 14-15; Duncan, ”Human Ecology and Population Studies,” pp. 678—716; Hawley, Human Ecology, Chapters 10, 11, and 12; Hawley, Roderick D. McKenzie 22 Human Ecology, Chapter 3; Schnore, loc. cit. _- 58Schnore, loc. cit. 59Hawley, Human Ecology, p. 179. 44 a large population with smaller sub populations rather than persons) engaged in sustenance activities, which involve functional inter— dependencies. The functionally interdependent patterns evolving in these activities are an important form of social organization——the division of labor. The preceding account has been aptly caught in the phrase "social organization is an adaptation of a population to its en- vironment." This leaves us a final point to make before we proceed to the next prepositional statement: (3) It is necessary to provide a definitional statement for the division of labor. I Definition: Egg, 33d distributing collective sustenapgg. Proposition IV: If_§p organization fp£_producing, maintaining, gpd distributing collective sustenance occurs, £p§£§_yill develop unplanned positions within the system which effect controls pp the producing, maintaining and distributing pf_collective sustenance (i.e., dominating positions). Proposition V: I: unplanned positiopg within the system which effect controls 33 the producing, maintainipg, and distributing pf_collective sustenance (i.e., dominating positions) develop, those positions will accrue 3 dis— other positions ip the system. It is with propositions IV and V that this thesis is concerned. We will develop our hypotheses within the broader scope of the theoreti— cal propositions. Along with a reminder that our units of observation ,. “1*.“ --_, . .. ..-.n.. . _ 45 are communities in a division of labor, it is important that we draw attention to one notion before turning to the following chapter and the more specific prOposition——our hypotheses. In a large pOpulation and especially a large, mobile, techno- logically developed nation, some form of monetary exchange is essential. Producing, maintaining, and distributing sustenance becomes inextricably linked with the gaining, maintaining and distributing of money. While there is not a literal translation from sustenance to money, legally speaking, money is translated into forms of sustenance. We mention this aspect of sustenance activities to emphasize, again, the importance of the economic dimension in the division of labor. "_ __ ,_.__,..' _" S.._._'.._?.—.T. .~.’ a-..w_,—._ .w._—,--.-.-_- -~'_' - _ .. -- _.__.. _ ._._.._.—._..._ CHAPTER III THEORY——PART II: HYPOTHESES Introduction This chapter attempts to provide some logical conclusions to statements and issues in the preceding chapter. Hypotheses evolving from the relationships between the independent variables and the dependent variables will be presented. More specifically and in order of their presentation, we will discuss (1) some general expectations about the overall distribution of inconsistency, (2) community size and metro- politan dominance, (3) functional specialization, and (4) the nonwhite population and female labor force. Before we proceed, however, it will be useful to make some comments on the basic premise of this whole study. We are working under the assumption that among communities can be found both the characteristics of strata consistency and of strata incon— sistency. While such characteristics have been demonstrated for 3 , l . . . . . . indiViduals, families,2 counties, and the general distribution of income 1G. E. Lenski, "Status Crystallization: A non—vertical dimension of social status,” American Sociological Review, 19 (August, 1954), 405— 13. Almost all of the work reported on status consistency—inconsistency has used individuals as the units of observation. We use Lenski‘s article as a reference because it serves as a groundbreaking classic in the par— ticular area, and it is typical in its use of individuals as units of observation. 2Ralph Spielman, ”A Study of Stratification in the United States" (unpublished Ph.D. dissertation, University of Michigan, 1953). 3John Stoeckel, "The Impact of Metropolitan Dominance Upon the Status Structure and Status Consistency of Rural—Farm and Urban Popula— tions" (unpublished Ph.D. dissertation, Michigan State University, 1966). 46 47 and education in relation to occupation for the United States,4 there is no direct evidence to demonstrate the consistency or inconsistency of communities. Nevertheless, it is quite logical, in keeping with other findings which used different units of observation, to suppose that some communities are consistent across stratum dimensions while others are not. Studies using differing units of observation have produced vary— ing results. Spielman5 found that 33 per cent of his national sample demonstrated status consistency. The census of 1960 showed 29 per cent of its sample to demonstrate status consistency while the remainder ex— hibited some degree of inconsistency. Since stratum scores of communities are summaries of many diverse individual scores, we can look for some of the extreme diversity of statuses among individuals to be concealed within the overall community stratum scores. We can expect to find less diversity of scores, and, consequently, more strata consistency among communities than has been found among individuals or families. Some observers may be led to conclude that no inconsistency will be found at the community level. The argument being, as suggested above, that at such a macro level all inconsistencies will disappear. However, it is the basic premise of the 4Robert Hodges, ”Occupational Composition and Status Crystalliza— tion: An Aggregate Approach” (unpublished M.A. thesis, Department of Sociology, University of Chicago, 1961). 5Spielman, pp. cit., p. 36. 6 The information was tabulated from data presented in the U.S. Census of Population and Housing: 1960, l/l,OOO and l/l0,000 Descrip— tion and Technical Documentation (Washington, D.C.: U.S. Government __ Printing Office, 1960)- 48 present investigation that there are factors at a macro level, which are associated with the stratum arrangements of communities, and which pro— duce varying degrees of consistency—inconsistency among those arrange— ments. This premise still remains to be demonstrated through analysis of the data. Overall Distribution Although they are not part of our more specific hypotheses, cer— tain characteristics of the overall distribution can be hypothesized. Given the general propositions we have already introduced, and in regard to inconsistency, there is justification for expecting each of the three hierarchies income, occupation, and education to occur with differing degrees of frequency.7 Let us re—emphasize two aspects of our general orientation. First, within the division of labor certain competitive advantages and dis— advantages are associated with different aspects and dimensions of com— munities. Second, the kind of advantages we are concerned with in this analysis are directly or indirectly economic. Hypothesis l——Distribution Income will occur most frequently pg the inconsistent dimension in strata inconsistency. 7We will explain our usage of ”Strata consistency—inconsistency” in operational terms in the next chapter. However, our present refer— ence to greater or lesser ”occurrence” with strata inconsistency requires clarification. We want to make clear that when strata inconsistency characterizes a community there is always one straUmIdimension——income, Education, or occupation——which is most deviant (inconsistent). Concep— tually, this is easy to understand. Since we are concerned with three strata dimensions whenever there is inconsistency one dimension, due to numerical determination, will be ”most” deviant. 49 Since income is obviously bound up in the economic aspects of the division of labor, we expect the income stratum to be most sensitive to those factors which contribute to economic advantages or disadvantages in communities. Consequently, since it is most sensitive, it will, in relation to occupational or educational strata, occur more frequently as the inconsistent stratum in strata inconsistency. Hypothesis 2——Distribution Education will occur least frequently pg the inconsistent dimension ip strata inconsistency. The major factor leading to this hypothesis is that education is the most controlled, through organization and legislation, of any of the three strata dimensions. State laws as well as some federal guide- lines contribute toward greater homogeneity among communities on this particular strata dimension, basically through attempting to raise the lower end of the distribution. This does not mean that communities and even regions do not present a hierarchy of educational status. On the other hand, because of the attempts to regulate the educational obtainment of populations, we expect nonplanned, economic factors to have less effect on the educational stratum than on occupation and income strata. One remaining aspect about the functionally interdependent division of labor among communities needs to be mentioned. While the point is not directly included in the explanation of the previous two hypotheses, it will be influential in the outcomes predicted by those hypotheses. Every community develops its own division of labor. While com- munities do vary in this regard, a minimal division of labor is neces— sary. Within each community the division of labor presents a range of occupational positions. This range of distribution of the occupational 50 stratum positions within each community hinders communities from obtain— ing overall extreme rankings on the occupational stratum. While it was pointed out that there has been a conscious attempt to influence educa— tional obtainment of community populations, there are, on the other hand, some ”natural" influences restricting the range of the occupational stratum in communities. "Natural” is used here to refer to the organiza— tional adaptation of a population to meet its sustenance needs in terms of a division of labor. We expect occupation to occur more frequently than education but less than income as the inconsistent stratum. As we indicated, this would support the predicted outcomes of the previous two hypotheses. Size and Distance We find it difficult to talk of size of community and distance from metropolitan centers apart from one another. A simple example will demonstrate the point. Let us suppose we are interested in certain characteristics of communities X and Y. Both of them have a population of 20,000. Both of them share another point in common, they each contain a college with 12,000 students. However, community X is on the outskirts of a metropolitan center of one and a half million inhabitants. Community Y, on the other hand, is the largest community within a 75 mile radius. Obviously, one would expect quite different things, say for example, in regard to the impact of the local college in the two communities. On the other hand, if X and Y were two communities both 70 miles from a relatively large metropolitan center, but X was 10,000 in size and Y 35,000, one would also expect some major differences. In the statement and explanation of the hypotheses to be 51 presented here, we will frequently look at the simultaneous effects of a community's size and its distance from a metropolitan center. On some occasions the two dimensions will be treated singly for a particular hypothesis. Before we present any hypotheses, it is necessary to make clear certain characteristics of the notion of consistency—inconsistency and to refer to some general attributes of communities. A unit, in this case a community, may have consistent strata in a number of ways. For instance, the community may have consistently high, medium, or low strata. That is, in terms of the number of levels used, say three, the community might be high in all three, towards the middle in all three, or low in all three. A community, then, can receive as high of a rating for being consistent but low in strata as it can for being consistent but high. We have indicated in the previous chapter, both theoretically and with reference to empirical findings, that average income is higher in larger communities. Hathaway, Beegle, and Bryant,8 in their census monograph, present findings which give some support to a decreasing average income with distance from metropolitan centers. Stoeckel9 has also presented similar evidence. Nevertheless, the findings associating decreasing size and increasing distance with decreasing income do not say anything about the degree or type of consistency—inconsistency. The phrase low, medium, or high ranked consistency refers to one of three general types of consistency. Using the term ”rank” gets away 8Dale E. Hathaway, J. Allan Beegle, and W. Keith Bryant, People pf Rural America, a 1960 CenSus Monograph (Washington, D.C.: U.S. Government Printing Office, 1968), pp. 192—98. 9Stoeckel, pp. cit., pp. 48—51. 52 from the awkward phrase, low (medium or high) strata strata consistency. Hypothesis 3-—Size and Distance Among strata consistent communitigg, with increased distance from consistent communities. The hypothesis, at first glance, may appear somewhat cumbersome and involved, but a little reflection will indicate that it follows from our general propositions. Given the dominance of larger population centers, communities most closely associated, spatially, to the large communities will reap a greater amount of the status benefits than will more distant communities. Ex— clusive suburbs functionally linked to metropolitan centers afford the most obvious example. Our argument, however, is not purporting any one— to-one relationship of increased distance and decrease in status. The theoretical framework from which we are operating suggests that there are other variables we will have to take into account presently. On the other hand, that same framework does indicate the importance of distance.10 While varying types of consistency may occur for a number of reasons, we expect low ranked consistency to become more prevalent as distance increases. We are concentrating only on low ranked consistency in the hypothesis. There is some value to be gained in a brief discussion concerning our focusing on only one type of consistency in the hypo- thesis. _————___.____._._ 10Some empirical evidence is also available which discourages any distance—only explanation. See ibid.; and Hathaway, Beegle, and Bryant, loc. cit. 53 Stoeckelll predicted linear relationships, in his analyses of counties, with regard to distance and status characteristics. It is our contention that the ideas involved in metropolitan dominance do not suggest such association. Bogue,12 in his classic treatment of the subject, talks of subdominant communities. Many analyses using the notion of metropolitan dominance suggest some type of hierarchy of dominance.13 The 1960 census monograph, People p£_Rp£pl America, shows some supporting evidence in this regard. When commenting on some of the data showing the earnings of operatives in relation to distance of counties where operatives are located, from metropolitan centers, the authors state: the medium earnings of operatives in SMSA's were usually higher than earnings of operatives outside SMSA's and that the differentials tended to rise with distance from the SMSA——pp least pp_£p_p_ppipp. [Emphasis ours.] While "subdominance" has been used to refer to varying sized and distant communities and spatial units, we suggest a general principle can be invoked from our theoretical framework. Wherever a community has, because of resources, location, or size (which is central to the present argument), come to dominate in the func— tional division of labor, relative status advantages accrue. The more llStoeckel, pp, cit., pp. 16—17. 12Donald J. Bogue, The Structure pf_the Metropolitan Community—— Q Study p: Dominance and Subdominance (Ann Arbor: Horace H. Rackham School of Graduate Studies, University of Michigan, 1950), p. 23. 13For example, see Otis Dudley Duncan and Others, MetrOpolis and Region (Baltimore: The Johns Hopkins University Press, 1960); and Bogue, loc. cit. 14Hathaway, Beegle, and Bryant, pp. cit., p. 193. 54 distant a community is from a metropolitan center the less direct the influence from the center. On the other hand, the distant community, while not on the same scope as the metropolitan center, has its own hinterland. Its own hinterland which, in terms of organization and economic resources, it dominates.15 Given the complexities created by the influence of subdominance, strict linear hypotheses are difficult to justify theoretically or demonstrate empirically. What we have done is to look only at one type of arrangement. By looking at more refined dimensions of strata arrange— ments, which can be supported by our theoretical framework, we hope to be able to find empirical results which more closely fit our theory. Given the differing types of consistency that can occur it is impossible, from our vantage point, to derive hypotheses fitting all types. However, with the awareness of varying kinds of consistency, it is possible to suggest hypotheses which fit a particular type. Hypothesis 4--Size and Distance ppppg strata consistent communities, pp varied distances fppp frequency pf_lpp ranked communities. The preceding hypothesis follows from the rationale that larger communities will accrue more status advantages. The result would be either higher ranked consistency or some form of inconsistency with one —-—————____.____._ 15We are concerned here with distance, but communities may, for varying reasons such as control of important resources, location, etc., become functionally dominant over a hinterland (including neighboring communities). Such communities will be expected to deviate from a linear decreasing of status characteristics, with distance from the metropolitan center. 55 or two strata dimensions exhibiting the advantage of size. Implicit in the previous logic is the notion that even though medium ranked consistency may be common in many distant communities high ranked consistency is very unlikely. Given the theory from which we are operating the smallest, most distant communities will virtually never have high ranked consistency. One point of clarification needs insertion before we present other hypotheses. As we will explain in detail in the next chapter, the smallest communities in our analysis do not go below 10,000 in population. The possibility of our smallest communities becoming subdominants is a real possibility. If we were to include smaller communities, our discussion would have to take into account the implausibility of, say, a community 1,000 in population accruing relative status benefits because of its own dominance. Our next hypotheses, related to the size—distance factors, are concerned with communities closest to the metropolitan centers. Hypothesis 5——Size and Distance Strata consistency pill pppp£_pp£p frequently EE.EEEl$ suburbs Eppp pp other communities. Suburbs, because of their very close and involved interdependence with metrOpolitan centers, will show more dramatic effects than more distant communities. One of these effects is the homogeneity in terms of strata dimensions which is permitted in small suburbs. Because the metropolitan center fulfills many of the consuming and employment needs of the suburban population, there is no necessary development in the Suburb of an extensively diversified, community supporting, division of labor. 56 Some major issues with the orientation we are operating from can be raised here, and a brief discussion is necessary. We are contending that the societal division of labor, with its requisites and outcomes, is a very useful framework for our analysis. We have also said that the ecological orientation is concerned with the unplanned consequences of the competitive struggle within the division of labor. Some of the critics of human ecology get most concerned about its lack of attention to culture and man's own ability to rationally plan communities. These concerns may be expressed by some critics in regard to suburbs and their development, e.g., the planning by city or county planning commissions, the development and plans by realtors and land development groups, and the residential tastes of middle class Americans (or whomever). These suggestions and many more can be offered as factors contributing to the morphology of suburbs. Our framework does not deny the importance or influence of such factors. What we would stress is that the very possibility of the existence, of a small community functionally interdependent with the large metropolitan center, was planned by no one. A burgeoning popula— tion, national immigration, movement from farm to city, technological development including travel and communication sources are some of the factors which provided the bases for suburban development. While realtors and developers, for example, may develop and sell a certain type of suburban development, they are dependent upon the size and type Of metropolitan center with its particular industrial and occupational composition. Many questions and issues can be raised about what in— fluences what the most and which comes first. The only point, ppp_ppp point we want to make, is that there are very legitimate, unplanned 57 factors which come about through community functional interdependence, and which, in turn, influence community morphology. Our previous hypothesis would not be expected to hold for larger When suburbs get larger they, of necessity, develop a more suburbs. We complex and diversified division of labor to meet their own needs. would expect such diversification to be reflected by less frequently consistent strata hierarchies. Given the dominating influence of the metropolitan centers, the possibility for smaller suburbs to enjoy the status benefits accruing to dominant centers is present. Hence, the following hypothesis: Hypothesis 6-—Size and Distance High ranked strata consistency will pp most prevalent among small suburbs. The next set of hypotheses, 7 and 8, are still concerned with size and distance, but now our interest turns to patterns of inconsistency. Hypothesis 7—-Size and Distance pp inconsistent strata arrangement with the occupptional stratum high will occur more frequently among suburbs than among other communities. Hypothesis 8-—Size and Distance pp inconsistent strata arrangement with income high will occur more frequently among suburbs pppp among other communities. The preceding two hypotheses are both generated from the same precipitating factors. Given the income advantage that is associated with Size, and given the general finding of an increasing proportion of white collar occupations with community size, 6 metropolitan centers would be M l6Otis Dudley Duncan and Albert J. Reiss, Social Characteristics 58 expected to rank high on these two dimensions. With the interdependen— cies among communities, eSpecially between metropolitan centers and their suburbs, we have suggested that status advantages of the metro— politan centers contribute to advantages for some of the suburbs. Many suburbs become residential centers which feed off of the industrial and occupational characteristics of metropolitan centers. The benefits, in— come and occupation—wise, will be experienced by many suburbs in such a way that we expect suburbs to be disproportionately over—represented among communities which have income and occupation high inconsistency. As we indicated earlier in this chapter, education is less free than are the other two dimensions to fluctuate. More positively stated, it is the one stratum that is under the most regulated control. Thus, while we can posit economic and occupational advantages for some suburbs, it does not follow that the same thing holds on the educational dimen— . l7 Sion. Functional Specialization The next set of hypotheses are related to whether or not a com- munity is functionally specialized, and in.some cases we will be con- cerned with the particular type of specialization. Given the multipli— City of specialization and the varying degree to which different m ‘“_______.________—__— and Occupational Composition with City Size" (unpublished M.A. thesis, University of Chicago, 1959), pp. 90—94. 17For some evidence on this point see: Hathaway, Beegle, and Bryant, pp. cit., p. 142. They found no relationship between distance and education level for counties. 59 specializations can or do influence community structure, the web of relationships is very complex, at best. We will concern ourselves with some of the more general relationships and major specializations. Hypothesis 9——Specialization Wipp ppp exception pf_£pp smallest communities, communities Kipp pp major specialization Bill.h§i§ consistent strata pppp frequently Eppp ppecialized communities. Specialization among communities is possible only because of func— tional interdependence with other communities and the accessibility or control of some natural resource. Depending on what type of functional relationships a community has or the importance of its particular re— sources, positions of dominance or subordinance arise within the division of labor. These positions will have associated with them relatively ad- vantageous or disadvantageous status characteristics. On the other hand, if communities are not specialized, one source contributing to differen— tial advantages or disadvantages is absent. This, along with the fact that a form of specialization may have differing effects on the separate dimensions, results in the preceding hypothesis. Most simply, the rationale is that in diversified (nonspecialized) communities the strata hierarchies will tend toward parallel development (in terms of vertical positions). The case of small communities will be taken up in a later hypothesis. Hypothesis lO——Specialization IE communities pipp pp major specialization, ppp_multiple affects g: decreasing community pppp_ppp increasing distance pill_produce pp increasing degree p§_strata consistency. 60 Here we have combined the effects of three independent variables and suggest their multiple influences result in a higher degree of con— sistency. Hypothesis ll——Specialization Specialized pppll communities yi£p_ppp_exception p§_£pppp Here again we have combined variables. In smaller communities the effects of specialization are more pervasive; consequently, if a community is both small and specialized, the specialization will shape the community toward consistency. In smaller communities the specialization will have enough influence to actually shape all three of the strata with which we are concerned. Whether it is high, medium, or low ranked consistency de— pends on the specialization, distance from metropolitan center, and other factors. Regardless, the influence is toward consistency. Hypothesis 12—-Specialization Communities specializing pp manufacturing pill PE inconsistent pipp Epp income stratum pigp. Manufacturing is by far the largest single industrial category. Its influence on community structure has been well demonstrated.l8 Depending on the type of relationships one is interested in, differing types of manufacturing may be more important than others, e.g., durable- nondurable. For our present problem we will consider manufacturing as a whole. Since manufacturing is a huge employer it includes differing occupations of varying position in an occupational stratum hierarchy. '—-—_—___ 18Duncan and Reiss, pp. cit., pp. 253—273. 61 To suggest that specialization in manufacturing produces any particular patterns may, at first glance, appear unlikely. There are, however, a number of factors associated with manufacturing that should be kept in mind. First, manufacturing's size (as an employing, industrial category) makes any influence it might have important directly or indirectly for all communities. Another facet of manufacturing is its crucial position in the overall functional division of labor. Especially in a large, highly technological society, those industrial activities that transform and combine goods, both raw and otherwise, directly or indirectly, into objects of consumption, are crucial. Manufacturing processes are, of course, the industrial activities performing these functions. A third characteristic of large scale manufacturing is very important for our analysis. Unlike most industrial activities which produce raw goods, and might be argued as being most crucial in the division of labor, manufacturing brings about large concentrations of workers. It will be recalled that one of the factors forming the basis for power is organi— zation. Given, then, manufacturing's propensity for organization, i.e., numbers in proximity, and its importance in the division of labor, we expect communities specialized in manufacturing to exhibit characteristic traits. One trait we expect is that income status will accrue dispropor— tionately to educational and occupational status. The tendency will ex— hibit itself in the form of income high, inconsistent status. There is some empirical evidence to support our expectation for income high inconsistent status among manufacturing communities. Duncan 62 19 and Reiss compared communities specialized in manufacturing to com— munities lowest in manufacturing. They found that high manufacturing communities tended to be lower in education and higher in income.20 In regard to higher status (white collar) occupations, they concluded: that manufacturing centers have low proportions of white—collar workers, not only because the manufacturing industries themselves employ relatively small numbers in white—collar jobs, but also because the concentration in manufacturing exerts an indirect or selective influence on the remainder of the community's occupational structure. What we are doing in this thesis is making general comparisons of communities Specialized in manufacturing. Duncan and Reiss give comparisons only for high and low manufacturing communities. Also, and equally important, we have presented a logical set of general propositions and accompanying hypotheses to account for their earlier findings. The Duncan and Reiss findings and our hypotheses are to be expected, given the part manufacturing plays in the overall division of labor, and given the characteristics of manufacturing which have facilitated labor organiza— tions in the manufacturing industries. These factors combine to give communities high in manufacturing a competitive, dominating (power) advantage within the division of labor, which is exhibited in high income status in relation to educational and occupational status. Hypothesis l3-—Specialization Communities specialized pp education HEAA.PE inconsistent Kipp education high pp income low strata arrangements. Hypothesis l4—-Specialization lglbid. ZOIbid., p. 270. 211bid., p. 268. 63 The smaller the community specialized pp education the greater the degree p: inconsistency. The rationale for the above hypotheses is centered around three points: (1) the number of high stratum occupational positions associated with education; (2) the very nature of the specialization, education, will tend to give the community high ranking on education; (3) we expect that a disproportionate amount of status, occupationally and educationally, is associated with educational specialization in comparison with income. Education due to its close linkage with occupation will tend to be similar stratum wise. In fact, an early hypothesis indicates our expec— tancy for inconsistency to be most prevalent in relation to income. That expectation is partially due to a closer connection between education and occupation. Since communities specializing in education will, by the nature of the specialization, be high in education, occupational ranking will also be high. While educational processes may be seen as essential in the over— all division of labor, they are not directly involved in the allocation and distribution of sustenance. In such a position, those communities specialized in education are not in as dominant a position as their educational position might seemingly indicate. We are suggesting, then, that the result is less income in relation to occupation and education. While we look for the general influence of educational specialization, it is expected to be more pronounced in small communities, where the simultaneous influences of other dimensions of the division of labor and Strata placements are not so pronounced. 64 Females and Nonwhite Despite the great range of individual diversity a few simple distinctions are widely used in the distribution of functions. Sex and age differences for example, serve uni— versally as bases of functional differentiation in human aggregates. Everywhere, from the most simply to the most complexly organized groups, the sex dichotomy is mirrored in the distribution of functions and privileges. . . . Racial heterogeneity, wherever it_ occurs, forms a third generally recognized basis for the division of labor. they perform and are usually set apart, too, by various prescriptions and other marks of distinction. [Emphasis ours.] The present study does not concern itself with age differences. However, the hypotheses to be presented are directly related to the other two bases for the division of labor——sex and race. Our procedure in presenting the remaining hypotheses differs from the previous format. The large body of literature which presents empirical demonstration of female and nonwhite subordinance within the division of labor and the resulting status characteristics, precludes the use of our general for— mat. It is not necessary and would be ludicrous to predict the situation 0f females and nonwhites from our propositions when the outcomes have a1~ ready been demonstrated many, many times. It does remain, however, as a future exercise to logically demonstrate how the major propositions we are Operating under can provide a partial explanation for the subordinant Position of females and nonwhites. What remains to be done in the present analysis is answer the question: Given the generally subordinant status of the female and nonwhite populations within the division of labor, what effects can we expect on the dependent variable where we have a relatively 22Amos Hawley, Human Ecology (New York: Ronald Press, 1950), pp. 183—84. 65 high number of females in the labor force or a relatively high number of nonwhites in a community? While there are some similarities between the positions of the female and nonwhite populations in the division of labor, there are also important differences. We will look at the effects of females in the labor force first and then nonwhites. Hypothesis 15——Females Except for the highest income communities, the greater the ro— .——————_——.—.—_—— strata inconsistency. Hypothesis l6——Females Except for the highest income communities, the greater the inconsistency will 23 income high inconsistency. Hypothesis l7—-Females Except fp£_£hg highest income communities, EEE greater Ehg preportion pf females lE.EEE.l2§2£.£2£EE.E§E greater Eh: degree pf strata inconsistency. All three of the preceding hypotheses are couched within the same rationale. While females are in a general position of subordinance with regard to occupation and income, the overall effect of their employment on community strata is to raise the income dimension. Analyses of occu: pational status continue to be based on the occupations of males. While there is good argument that this approach does not reflect the total‘ picture,23 it is a realistic approach in terms of how status gets 23Ernest A. T. Barth and Walter Watson, ”Social Stratification and 66 assigned and then transferred intergenerationally, i.e., by family units, most usually with male heads. Analyses of the occupational stratum of communities have used a summary indication based on the distribution of male occupations. Again, it does not give the total picture, but it does realistically reflect that aSpect of occupational status to which people most frequently pay attention and react. The result is that the female's occupation frequently gets glossed over or ignored, but her contribution income—wise is very real. Female employment contributes to family income, where both male and fe— male are present and employed, and to overall income in the community. While her income may be less than male income, i£_i§ income. With greater numbers of females employed there are, proportionately speaking, more persons bringing in money. It is obvious, then, why greater female employment is expected to result in greater inconsistency and why this would more frequently be income high inconsistenCy. It leaves for explanation, however, the qualifying phrase in the hypotheses, "excluding the highest income communities.” 24 and from national data on From working with census tract data family income,25 we have found evidence that female employment contributes very little toward income in the upper income brackets. We expect the same pattern to hold at the community level. The explanation for this occurrence is rather simple. For whatever other reasons they might have, the Family in Mass Society," Social Forces, 45 (March, 1967), 392-402; L. H. Day, "Statue Implications of the Employment of Married Women in the United States,” American Journal 2; Economics, 20 (July, 1961), 391-97. 24In class exercise assigned to students, we have consistently ob- served that the highest average income tracts in a community are always among the lowest in percentage of females employed. 25See Appendix A, Figure 2, which demonstrates the point. 67 a good proportion of females work to supplement family income. Supple— menting income is most crucial where there is little of it and least essential where there is a lot of it. Income supplementing, by female employment, is most often practiced, then, by lower and middle income families. There is no need for supplementing income in the highest in— come brackets. If we were to include the highest income communities, e.g., some of the high income suburbs, they would cloak the relationship which we expect to hold for the remaining communities. Hypothesis l8——Nonwhites The greater the proportion of nonwhites in communities the greater the freguency pf strata consistency. Hypothesis l9--Nonwhite The greater the proportion pf nonwhites 33 communities the more likely £223 strata consistency Hili.§§.i2fl' Hypothesis 20—-Nonwhites TEE greater phg proportion pf nonwhites ip_communities £32 greater Ehg degree gf_strata consisteppy. Perhaps the expected effects on the dependent variable are more obvious for these hypotheses than for any others. A little reflection on the female position will make the nonwhite position even more stark. Females, for the most part, make up only one part of the earning and consuming unit, the family. What they lack in equal position on the income and occupational dimensions they, speaking of the family unit, get compensated for through the male's prestige positions. Secondly, in regard to education females rank equal with or superior to males.26 m 26The median school years completed for all females in 1960 was 68 While their higher ranking educationally highlights the status inconsis— tency for females in general, it does make the point that they have relatively high status on that dimension. If one is nonwhite, however, there are no compensations.27 In the family unit the other members are also nonwhite and have no compen— sating prestige positions to offer. Not only are there a lack of advan— tages income and occupationally, but educational status is also low. Due to the discriminatory practices from the dominant whites, the nonwhites have been kept in subordinant positions within the division of labor. This subordinant position has accrued to the nonwhite population the associated low positions on the three strata dimensions. Consequently, wherever communities have a large proportion of nonwhites the predicted hypotheses are expected to hold. The following chapter will present the methodological procedures to be used in this thesis. Included in the presentation will be the previous hypotheses stated in Operational form.28 10.9 compared to 10.3 for males. In all age categories females are at least equal and usually higher than males in median school years com— pleted. For national data see: U.S. Bureau of the Census, yy§3 Census pf Population: 1960. Subject Reports. Educational Attainment (Washington, D.C.: U.S. Government Printing Office, 1963), Table 1, pp. 1—3. 27We are not concerned here with whether the status positions of nonwhites are changing or how much if they are. We are looking only at their subordinant positions as they now exist and exploring the conse— quences for strata Consistency—inconsistency. 28We find ourselves in agreement with recent authors who have pointed to the difficulty of connecting the conceptual and the empirical realms. Not only is it difficult, but often a neglected concern by many researchers. We will state the hypotheses in an operational form to demonstrate how we have attempted the connection. The strengths and weaknesses of the linkages are more apparent when such transformation statements of the hypotheses are made. Some hypotheses, depending on 69 their conceptual statement, are almost in operational form, while the operational form of others is not so obvious. The following references are to some discussions which argue for the importance (and the past neglect) of linking the conceptual and empirical realms. Hubert M. Blalock, Jr., and Ann B. Blalock, Methodology ip_Social Research (New York: McGraw—Hill Book Company, 1968), Chapter 1; Gideon Sjoberg and Roger Nett, A_Methodology £33 Social Research (New York: Harper and Row, Publishers, 1968), pp. 33—38. For a slightly older but classic discussion of the general problem see Chapter VII, "Epistemic Correlations and Operational Definitions,” in F. S. C. Northrup's Th5 Logic of the Sciences and the Humanities (New York: The Macmillan Com- pany, 1947). CHAPTER IV METHODS AND METHODOLOGICAL CONSIDERATIONS Introduction Moving from the conceptual to the empirical realm within an ecological framework can be problematic. Hawley has commented that the "breadth of conception is one of the great virtues of human ecology and 1 also the source of its major difficulties." Part of this "source" is the general methodological problems encountered in a macro approach. Translating what might be a titillating conceptual analysis, partially because of its breadth, into a meaningful empirical counterpart is not always a graceful exercise. Blau has made the following comments and broad procedural sugges- tions in regard to analyzing problems like the one under investigation. Both the study of the determinants and the study of the consee quences of social organization are objectives of sociology, but the former is less easily done in empirical research and requires appropriate modifications of the prevalent research methods. To investigate the characteristics of social structure as the dependent variables to be accounted for by various antecedents, including other aspects of the social structure, necessitates (1) that dif- ferent organized collectivities be examined rather than individual differences within one, however the boundaries of organized collec— tives are defined; (2) that organized collectivities be treated as units of comparative analysis; (3) that the empirical data, though usually referring to observed conduct of individuals, be converted into measures of social structure, such as division of labor, status hierarchy, or homogeneity of beliefs; and, ideally. (4) that a large sample of collectivities be studied, because the organization of collectivities differs in so many respects that lAmos H. Hawley (ed.), Roderick 2. McKenzie 23 Human Ecology (Chicago: The University of Chicago Press, 1968), p. xiv. 70 71 only multivariate analysis of many cases can hope to distinguish causal connections from correlated biases. In the sections that follow and in some detail we will present our attempt to comply with the procedures suggested by Blau, i.e., (l) the organized collectivities are communities; (2) communities are the units of comparison; (3) indicators of the division of labor and status dimensions in the form of Strata consistency-inconsistency are the basic data; and (4) while how large is large may be debatable, the sample contains 539 communities. We will describe the sample first, then operational procedures, comment on major control variables, present a diagrammatic representation of our presentation up to that point, and comment on statistical analyses. Sample The sample is a disproportionate, stratified random sample of communities from 10,000 to 100,000 in population in the continental United States. The sample is stratified by size categories of 10,000 through 24,999, 25,000 through 49,999, and 50,000 through 99,999. It was neces- sary to select differing proportions from each of the size categories. The total number of communities in each of the categories under discus- sion is size 1 (IO—24,999) = 225, size 2 (ZS-49,999) - 214, and size 3 (SO-99,999) s 100, which results in a total of 539 communities.3 The decreasing number of communities with increasing size is apparent in the previous figures. There is a basic reason for this. The actual number M 2Peter Blau, ”Objectives of Sociology," A Design for Sociology: ___2 j ierstedt editor (Philadelphia: Sco e Ob ectives and Methods, Robert B , The Adademy of Political and Social Sciences, 1969), pp. 51-52. 3See Appendix B for a listing of the communities in the sample. fi 72 of communities in the nation decreases as size increases. If there had been no limitations on expense and all the communities in the nation had been used, there would Still be an inverse association between size and numbers of communities. Given the necessity of a large number of cases for comparative analysis, especially with control variables employed, it was necessary to draw different proportions from each size category. The procedure resulted in a 10 per cent sample from size 1, a 50 per cent sample from size 2, and a 50 per cent sample from size 3. The differing proportions were taken into account and proper weights employed, when it was necessary to develop and compare scales or measures representing national characteristics of communities. Our purposive limitation of the size range of communities re- quires some comment. The community variable that has received the most attention in research through the years is size.4 While there is a lot remaining to be explored about the effects of community size on other characteristics of communities, it is also true that we probably know more about the effects of size than we do about any other community vari— able. Based upon the expectation that size would have some effect on status consistency—inconsistency, but not wanting the influence of size to overwhelm the effects of other variables, we limited the size range. M 4See, for example, William F. 0gburn, Social Characteristics of Cities (Chicago: International City Managers' Association, 1933); Fenton KEYES, ”The Correlation of Social Phenomena with Community Size (unpub- lished Ph.D. dissertation, Yale University, 1942); Otis Dudley Duncan and Albert J. Reiss, Jr., Social Characteristics of Urban and Rural Communi— ties, 1950 (New York: John Wiley and Sons, Inc., 1956); Howard B. Kaplan, "An Empirical Typology for Urban Description" (unpublished Ph.D. disser— tation, New York University, 1958); Jeffrey K. Hadden, A Systematic Study- of the City as a Unit of Analysis" (unpublished Ph.D. dissertation, The University of Wisconsin, 1963); and Jeffrey Hadden and Edgar F. Borgotta, American Cities (Chicago: Rand McNally and Company, 1965). 73 The lower size limit was determined on practical grounds, but with some theoretical justification. The census data do not provide the same amount of information for communities under 10,000 as they do for larger communities. For some of the information needed in this study we could not get comparable information for communities less than 10,000 in size. On the theoretical side, the question can be asked of how small a community can be and still maintain an internal division of labor com— parable to larger communities? A population of 10,000 is large enough to reflect an industrial and occupational diversity with enough variation for comparison with larger communities. It may be that 10,000 is getting close to the point where the internal division of labor is appreciably different. While we cannot answer this in the present research, we do feel confident that 10,000 represents a relatively small community which still possesses adequate comparative characteristics. The upper size limit of 100,000, while somewhat more arbitrary, is defensible. On methodological grounds, the number of communities of larger size decreases so rapidly that comparative analysis becomes more difficult. It becomes a very real problem when controls, e.g., on region, are used. A second reason for excluding communities of a larger size is the increasing organizational complexity associated with larger size. This factor in itself needs further exploration and is reason for--not against-—further exploration of larger communities. Nevertheless, in the present analysis the possible ramifications of increased complexity are Purposefully excluded by limiting the size of communities in the sample. Hopefully this exclusion will permit a clearer view of relationships between the other specified variables of the study. 74 In regard to the sample one further point needs clarification. It is, in actuality, a discussion of the operational procedures used for defining ”community.” The discussion is presented here to help describe the sample. We will present the section "operational procedures” which treats each of the other major variables, after the present dis- cussion. In the present analysis we use the U.S. Census classification "urban place” as equivalent to ”community." The concept of community has long been a central one to human ecologists. It has been possible to symbolically use the term in varied, theoretically suggestive ways. Nevertheless, when it has come to an empirical analysis a unit has been used which varies from the conceptual one. There have been attempts to develop an empirical unit which provided a high degree of isomorphism with the conceptual unit.5 Even where isomorphism can be demonstrated the results are not practically applicable to future analysis. Data are still basically collected and presented in terms of political units. We suggest that the problem be recognized, but that it not be considered as Stymieing. What most analysts have done is use one of the classifications provided by the Census. The Census designation of urban place in 1960 corresponded to the political definition with some emphasis also given to other considerations. The term ”place” as used in reports of the decennial censuses refers to a concentration of population regardless of the existence of legally prescribed limits, powers, or functions. Most of the places listed are incorporated as cities, towns, villages, or boroughs, however. In addition, the larger unincorporated places outside the urbanized areas were delineated and those with a popu- lation of 1,000 or more are presented in the same manner as incor- porated places of equal size. Each unincorporated place possesses W 5Allan Gunner Feldt, "The Local Ecological Cgmmunity: An Investi- gation of Relative Independence in an Urban Society (unpublished Ph.D. dissertation, University of Michigan, 1962). 75 a definite nucleus of residences and has its boundaries drawn so as to include, if feasible, all the surrounding closely settled area. Unincorporated places are shown within urbanized areas if they have 10,000 inhabitants or more and if there was an expression of local interest in their recognition. The towns in New England and townships in New Jersey and Pennsylvania recog— nized as urban are also counted as places, as is Arlington County, Va. Beverly Duncan in one reported research used economic areas as spatial units in her analysis. While her concern (as is much of ours) was with economic related factors, she does make the statement "we do not suggest that this system of areas has any particular merit, as com— pared with alternate systems.”7 What areas are used must depend upon the conceptual problem at hand and then the data limitations. Duncan suggests that the situation has been stated well by Vining, we agree. The spatial structure of human economy should be regarded conceptually as virtually a continuum. As in other studies of phenomena having volume or spatial extension, empirical observa— tions must be made upon the contents of finite and arbitrary spatial units, these empirical observations being viewed as providing an approximate conception of what would be viewed were the Spatial limits made smaller while the contents were being made more dense. While we concur with Duncan, that Vining's position has to be taken for empirical analyses, we do suggest there is merit in the particular units (urban places) we are using. It is true that communication and transportation have made community distinctions less apparent and in all hm 6The definition of ”urban place” used by the Census can be found in many of its publications. See, for instance, U.S. Bureau of the Cen— Sus, ELS. Census of Population: 1960. Number of Inhabitants, United States Summary. Final Report PC (1)-A (Washington, D.C.: U.S. Govern— ment Printing Office, 1961), p. xxii. 7Beverly Duncan, ”Population Distribution and Manufacturing Activity: The Nonmetropolitan United States 1950," Papers and Proceedings: The Regional Science Association, 5 (1959), 96. 8Rutledge Vining, "Delimitation of Economic Areas: Statistical 76 probability less real. On the Other hand, it is the very interdepen— dencies made possible by communication and transportation that we are investigating in our analysis. In general, the decision to use urban places results from two basic considerations. One, the relatively high quality and mammorh amount of data provided by the Census, and it is a unit designation which reflects some important characteristics of our conceptual notions of community. A second reason is that many analysts have followed a similar approach and the present study should provide some accumulative and comparative results. Operational Procedures Dependent Variable: Strata Consistency—inconsistency The operationalization of strata consistency—inconsistency has been accomplished through two separate but complementing procedures. 9 One method is tailored after Lenski's original procedures and the second borrows from the U.S. Census's10 method. However, before a dis— cussion of the mechanics involved can be reviewed, it will be necessary to present the operational processes followed in determining an educa— tional, income, and occupational ranking for each community. It is these separate indicators that are combined to form the more general variable, strata consistency—inconsistency. Conceptions in the Study of the Spatial Structure of an Economic System," Journal 2: the American Statistical Association, 48 (March, 1953), 44. 96. E. Lenski, "Status Crystallization: A non—vertical dimension of social status,” American Sociological Review, 19 (August, 1954), 405— 13. 10U.S. Bureau of the Census, Methodology and Spores 2f Socio- economic Status, Working Paper No. 15 (Washington, D.C.: U.S. Government Printing Office, 1963). 77 The first two dimensions, education and income, were determined rather straightforwardly and require only brief comment. For education .11 the U.S. Census figures under ”median years of school completed' for each community were used. For income the figures under "median family "12 were used. income Family income was used for a number of reasons. The family repre- sents the best single indicator of total income received in a community. Due to their proportion (family income units) in every community and their comparability from community to community, family income units present the most logical choice among median income measures. Also, family income reflects the contribution of both spouses, and that is important in the present research. Our measure of the occupational stratum is less direct. We used Duncan's13 occupational ranking scale on each community. The U.S. Cen— sus's collapsed l4 category occupational distribution for males was used in each community. The proportion employed in each occupational category was multiplied by the appropriate scale score and the totals summed. The summation became the occupational score of the community.14 The occupational categories used and the assigned scores were as follows: llU.S. Bureau of the Census, U.S. Census pf_Population: 1960. General Social and Economic Characteristics, Final Report PC (1) (Washington, D.C.: U.S. Government Printing Office, 1961), Table 73. 121bid., Table 76. l3Otis Dudley Duncan, Occupational and Social Status, Albert J. Reiss, 25 pl. (Glencoe: The Free Press of Glencoe, Inc., 1961), Chapters VI, VII, and Appendix B. 14See Appendix C, Table 35, for the centile distribution of 78 - Socioeconomic MElEE Occupation Cgppp Index Professional, technical, and kindred workers 75 Managers, officials, and proprietors, exc. farm 57 Sales workers 49 Clerical and kindred workers 45 Craftsmen, foreman, and kindred workers 31 Occupation not reported 19 Operatives and kindred workers 18 Service workers, exc. private household 17 Farmers and farm managers 14 Farm laborers and foreman 9 Private—household workers 8 / Laborers, except farm and mine 715 Duncan's scale is particularly applicable to the problem under analysis: (1) he derived his measure from census data and so the applica— tion of his occupational categories to ours is direct and easy. (2) Dun- can16 has provided an analysis of the variation of the scale by region and under various demographic centrols which concludes in supporting the utility of the measure on a comparative intranational basis. (3) The scale was devised from aggregate measures; consequently, it fits the problem under discussion, both in concept and in its empirical deriva— 17 tion. (4) The very precedure of ranking occupations used by Duncan is W occupational status scores and coding procedures. 150m; Dudley Duncan, 22. cit., p. 155. 16Ibid., pp. 162-238. 17This argument and others have also been made by Stoeckel. See John Stoeckel, "The Impact of Metropolitan Dominance Upon the Status 79 advantageous to one of the tasks of this thesis. Occupational ranking was determined by the amount of education and income associated with each occupational category. The very process used to rank occupations insures a high correlation with income and educational levels. Since by rank definition there is an association between the three variables, the variation which occurs between them, and which we are interested in ex— plaining, can be more readily attributed to other sources than scale construction. In short, it presents a conservative bias toward strata consistency and against inconsistency. (5) One remaining advantage is worthy of note. Duncan has shown that his scale correlates highly with the NORC scale. He has provided a table for transforming occupational rankings by his scale to NORC prestige scores.18 After the separate status scores had been determined a distribu— tion for each of them was formed and decile levels determined. Every community received a score from zero to nine on each status dimension, depending into which decile it fell. These three standardized (centile) scores provided the data for determining the type and degree of consistency—inconsistency. An illustration to clarify the previous discussion may be useful. If community X had a median score of school years completed from 9.9 through 10.3, it would have been assigned an educational rank score of 2. If it had received an occupational score from 32.27 through 33.97, it would have received an occupational ranking of 2. On the income W Structure and Status Consistency of Rural—Farm and Urban Populations" (unpublished Ph.D. dissertation, Michigan State University, 1966), pp. 27‘28 a 18Otis Dudley Duncan, 92, cit., Appendix B. 8O dimension, a score of 2 would have been assigned if the median family 19 In the income for the community was from $5,104 through $5,441. example community X would obviously be consisrent in overall strata, i.e., it was in the same centile_strata on each dimension. In the following paragraphs the actual operational procedure followed to determine degree and type of consistency-inconsistency is outlined. Degree—~as was indicated earlier, Lenski's computation of status crystallization is used as one of our measures.20 The major difference between the procedure used here and the original approach of Lenski is that our units of observation are communities and not individuals. A second difference of some importance is the procedure used to develop common (standard) scales for each of the separate status dimensions. As Lenski has indicated, "without common scales, a measure of status ”21 crystallization would be impossible. The procedure used in the present research was to use centile scores. For Lenski: Frequency distributions were established for each hierarchy. Using these distributions as a basis, scores were assigned for each of the various positions (or intervals) in each hierarchy on the basis of the midpoint of the percentile ranged for that position (or interval). 2 The basic difference in the derivation of the strata scores is that Lenski's scores result from the accumulated percentage of cases in each predetermined category, e.g., income-—$l,OOO-$l,900; $2,000—$2,900, N.— 19For the centile distribution and codings procedures for education and income see Appendix C, Tables 36 and 37, respectively. 20Lenski, loc. cit. 2 21mm. , p. 407. 2mm. 81 etc. The scores used in our research result from the accumulated percen— tag; of actual gaggs, e.g., income-—the lowest 10 per cent, the second 10 per cent, etc. Since the outlined procedures were consistently used and since statistical analyses are relative to the scoring procedures used, there appears to be no inherent advantage in the Lenski approach. What advantages there are in one approach appear to be counter balanced by possible advantages in the other. Nevertheless, we do mention one ad— vantage of the centile procedure. Intuitively, a centile score compari— son between dimensions, e.g., a 3 in income and a 5 in occupation, is more readily perceived than with the Lenski approach. The computation of the consistency~inconsistency score is the same as was developed by Lenski. The difference we have pointed out concerns only the scores used in the computation not the computation itself. This was accomplished by taking the square root of the sum of the squared deviations from the mean of the three hierarchy centile scores 9£_£hg community and subtracting the resulting figure from one hundred. The more highly consistent . . . a community's status, the more nearly 135 consistency—inconsistency score approached one hundred; the less consistent . . . 135 status, the more nearly $55 consistency-inconsistency score approached zero. The us of squared deviation from the mean rather than simple deviations was employed to emphasize the effect of larger deviations and to minimize the effect of smaller deviations. This was considered desirable since the techniques employed in quantifying positions (or intervals) in the several hierarchies were sufficiently crude so that no great importance could be attached to small deviations. The technique of subtracting the resulting figure from one hundred was employed so that communities whose status was highly consistent would have numerically higher consistent scores than those whose status was inconsistent. This was done solely to avoid semantic difficulties.23 [Italics refer to our change of s u I s I I s s o I o s 0 23Ibid., pp. 407—08. For the distribution of status 82 terms so that the quotation fits the present problem.] Reference to Appendix C, Table 38, will indicate that the scores obtained from using this procedure ranged from 100 to 35. The Lenski procedure tells us about the degree of consistency-inconsistency. We now turn our attention to a complementary aspect of the concept. Type——the importance of this dimension of consistency— inconsistency has been indicated. For further example, two communities may be 100 per cent consistent in strata but one consistently high and the other consistently low. To lump both communities into the same category cloaks as much as it reveals. The procedure used for typing consistency— inconsistency follows, in general, the procedure developed and used by the U.S. Bureau of the Census.24 After determining the centile scores for every community on each of the three separate dimensions, consistency— inconsistency types were determined in the following ways: 1. If the range between the highest and the lowest centile scores was two centiles or less, the community was considered to be consistent and a code of l, 2, 3, or 4 was assigned.25 a. A code of l was assigned if two or all three of the centile scores for the community were in the 1 through 9 or 10th percentiles. b. A code of 2 was assigned if two or all three of the centile scores for the community were in the 20th, 30th, or 40th percentiles. M..— consistency-inconsistency scores obtained in the present analysis see Appendix C, Table 38. 24Methodologyand Scores of Socio—economic Status, pp. 2—3. 25A comparison between the Census and Lenski procedures is discussed later in this chapter. 83 c. A code of 3 was assigned if two or.all three of the centile scores for the community were in the 50th, 60th, or 70th percentiles. d. A code of 4 was assigned if two or all three of the cen- tile scores for the community were in the 80th or 90th percentiles. 2. If the range between the highest and the lowest centile scores was three centiles or more, the community was considered to be inconsis— tent and a code of 5, 6, 7, 8, 9, or 10 was assigned. a. If the range between the highest and the medium centile was greater than the range between the medium and the lowest centile: — a code of 5 was assigned if the income centile was highest — a code of 7 was assigned if the educational centile was highest - a code of 9 was assigned if the occupational centile was highest. b. If the range between the medium and the lowest centile was greater than the range between the highest and the medium centile: - a code of 6 was assigned if the income centile was lowest ~ a code of 8 was assigned if the educational centile was lowest - a code of 10 was assigned if the occupational centile was lowest. 84 The resulting consistency-inconsistency types may be described as follows: Strata Consistency- Inconsistency Type 1 All three components consistent with overall very low rankings. 2 All three components consistent with overall low rankings 3 All three components consistent with overall high rankings 4 All three components consistent with overall very high rankings 5 Occupation and education most consistent; / income high 6 Occupation and education most consistent; income low 7 Occupation and income most consistent; education high 8 Occupation and income most consistent; education low 9 Education and income most consistent; occupation high 10 Education and income most consistent; occupation low The only situation not defined by the preceding rules is one particular case of inconsistency. It occurs when two of the centile scores, one high and one low, are equidistant from the medium centile. This occurs in the sample when the two extreme centiles deviate two — " i 85 centiles from the medium (15 cases). With the few cases in question, we arbitrarily considered the lower centile as the most extreme and scored the community accordingly. We chose a score of 80 from the Lenski procedure as the dividing point between consistency and inconsistency. Besides an intuitive appeal, i.e., on scales with 100 as maximum 80 or above is generally considered "fairly high” in everyday parlance, there are methodological justifica- tions as well. (1) It takes a range of 3 centiles or more between centile scores, using the Census method, for a community to score below 80. The Lenski procedure, when we use 80 as the dividing point, and the adaptation of the Census procedure are complementary. Since we are using two separ- ate operational procedures to get different dimensions of our problem, an / important criterion is that they do not conflict. If two separate opera— tional definitions produce two conflicting measurements, the procedure is always open to the criticism that two different concepts are actually being measured. This criticism is not justifiable in the present analysis. (2) A second justification for the dividing point comes from empirical sources. If the distribution of consistency—inconsistency scores are treated as continuous data, the mean is 82. The actual scores closest to the mean are 84 and 78. We come as close as possible to dividing the distribution in half at the actual mean by using 80 as the dividing point. The actual distribution that occurs is 58 per cent of the communities are consistent and 42 per cent are inconsistent in strata.26 An example.of five hypothetical communities with varying centile M 26See Appendix C, Table 39. 86 scores will demonstrate the general principles that have just been dis- cussed in detail. Range Between Highest Communities Centile Scores Lenski Score and Lowest Gentiles I_ §_ 9 Consistent V 8 8 7 92 l W 8 7 6 86 2 Inconsistent X 8 7 5 78 3 Y 8 7 4 7l 4 . Z 8 7 3 63 5 Independent Variables Size. The designation for community size is the total population of an urban place as given in the 1960 Census. Distance. The measure of distance and the rules for defining its usage are basically those developed in the Hathaway, Beegle, and Bryant census monograph.27 The indication of distance refers to distance from an SMSA.28 For the previously mentioned authors the distance value is determined by concentric circles of 50—mile wide bands ”having their common center in the geographical center of the largest central city 27 Dale E. Hathaway, J. Allan Beegle, and W. Keith Bryant, People 2: Rural America, A 1960 Census Monograph (Washington, D.C.: U.S. Government Printing Office, 1968), pp. l7—18. 28SMSA refers to Standard Metropolitan Statistical Area. For a detailed definition see "Area Classifications,” in Introduction of the 1960 Census 2£_Population, General Social §Ed_Economic Characteristics, Series PC (1), p. vii. The general meaning can be understood from a short statement by Hathaway, E£.El" pp. 232', p. 17: ”A standard metropolitan statistical area is a county or counties having at least one city with a population 87 of each SMSA."29 The center circle is scored as zero, the next SO-mile band is scored as l, the next 2, and so on, with increasing numbers for each successive band. A community receives a distance score depending on which band it falls within.30 There are two variations from the Monograph procedures that need clarification at this point. The procedures developed for the Monograph used counties as the units of observation. The authors used the pro— cedures just outlined, but they were interested in developing indicators for counties. Since the distance measure used in the present research was drawn from their data, an interpolation was necessary. The county of each of the communities in the sample was identified and the county's distance score assigned to the community. This procedure means that a community is always assigned the distance indicated by its county's closest boundary to the nearest SMSA. It is possible that some communi- ties would have received a higher distance score, i.e., they could be located in a distant end of the county. However, the bands are 50 miles wide, which suggests the probable inclusion of a number of counties in each band and reduces the occurrence of the problem. To the extent our procedure does produce some error, the error is consistent. The consis— tency is of a fashion that will not alter interpretation in the present analysis. Mm over 50,000, or two cities having contiguous boundaries with a combined population of over 50,000, or a county which is metropolitan in character adjacent to a county with such a city.” 29Hathaway, §£_al., loc. cit. ____._ 30For specific rules for a SMSA which overlaps two bands, see ibid. 88 The second variation involves a slight alteration in deter- mining distance. Our argument has been that metropolitan centers (SMSAs) dominate surrounding communities. We are not concerned with effects on SMSAs themselves, but with their affects on outlying com- munities. In the Monograph procedure central cities are scored as zero along with all other communities in the first 50—mile band. It was necessary for us to exclude central cities when trying to assess the effect of dominance. One further distinction is made between com— munities that are within the first SO—mile band. We had available the data from one type of ”suburb” designation (see following section). These are communities within the first SO—mile band but generally closer to central cities than other communities also within the first band.. Since such a large number of sample communities are within the first band, we refined it. Our designation for communities closest to central cities is 'suburbs.” The next distance desi nation will 8 be ”distance 2” communities. These are those communities within the first SO—mile band, but outside of SMSAs. From that point on our procedure for designating distance is identical to the Monograph's, i.e., ”distance 3” communities are within the next SO—mile band, and so forth. Since we have based our basic procedure so heavily on the Monograph, the following presentation will show the similaries and differences in procedures. 89 Monograph Designation Actual Milgs 935 Designation Distance 0 Within first SO-mile band - Central Cities " " " " — Suburbs " " " " Suburbs All other communities within 50—mile band " " " " Distance 2 Distance 1 51 to 100 miles Distance 3 Distance 2 101 to 150 miles Distance 4 Distance 3 151 to 200 miles Distance 5 Distance 4 201 to 250 miles Distance 6 Suburb. Our classification of suburbs is taken from the 1963 Municipal Year Book.31 We have used their definitions and have desig— nated each community as either a central city, a suburb, or an inde— pendent community. The definitional rules are: Central city: the largest cities of the Standard Metropolitan Statistical Areas (SMSA); usually have a population of 50,000 or more. Suburb: all other urban places over 10,000 located within an SMSA. Independent city: all urban places over 10,000 population that are located outside SMSAs.3 Functional Specialization. In defining the specialization of communities we have followed the basic procedures suggested by Duncan and Reiss.33 For determining manufacturing specialization, we used the percentage of the employed resident population34 in manufacturing as the 31Victor Jones, Richard L. Sorstall, and Andrew Colliver, ”Economic and Social Characteristics of Urban Places," ThgpMunicipal Year Book: 1963 (Chicago: The International City Managers Association, 1963), pp. 85—157. 321bid., p. 111. 33Duncan and Reiss, 599;. gi_t. 34The distinction of "resident population“ in determing 90 determinant. The percentage employed in manufacturing was determined for each community.35 A distribution of all the percentages was determined and communities located in the upper quintile (20 per cent) of the dis— tribution are categorized as Specialized in manufacturing. Specialization in education was determined on the basis of the percentage of the 20—28 year old population enrolled in school. The implication is that the social structure and the economic base of the community are conditioned to some extent by a college, university, or professional educational institution. These effects generally are thought to include such things as the size and composition of the student body and faculty, its general ”excellence," and its contributions to the reputation of the community as an ”educational center. ”37 Duncan and Reiss used the ages of 20—24 years as their determining age range. Given the increasing enrollment in graduate training the ex— tension of the age range reflects a more "realistic” specialization in higher education. As with manufacturing, those communities in the upper quintile on this dimension are categorized as specialized in education.38 manufacturing specialization, separates this approach from the alterna- tive of using per capita value added by manufacturing (dollars) . The first approach uses the characteristics of the population regardless of place of employment. The second procedure places emphasis on manufactur— ing located in the community. The problem now under analysis indicates the greater utility of the first approach. For a discussion of the two approaches see ibid., pp. 219—23. 35 U. S. Census of Population: 1960, General Social and Economic Characteristics, Table ”74. -—.—.——’ —.————-———-—_ 36Ibid., Table 73. 37Duncan and Reiss, pp. pi£., pp. 274-345. 38The procedure followed in this classification allows a com- munity to be specialized in more than one functional area. For a dis— cussion of this procedure see ibid., pp. 215—19 For the most recent classification using the alternative approach of allowing only one form of specialization per community, see Jones, Forstall, and Colliver, loc. cit. 91 There are numerous specializations, based on percentage employed in each industry, that could be analyzed. We continue to concern our- selves, in this research, with manufacturing and education. However, we will utilize other forms of specialization in the analysis and they need to be mentioned. Duncan and Reiss39 demonstrated that communities specialized in retailing or wholesaling present distinctive character— istics. While we do not give specific conceptual attention to these specialties, we do control their influence in the analysis and in the Statement of the operational hypotheses. The procedure used to deter- mine communities specialized in these functions is identical to the pro— cedure used to determine manufacturing specialization, i.e., the com- munities in the top quintile of the respective distributions. Female and Nonwhite. These two variables represent straightfor— ward usages of the Census data. The proportion of the labor force that is female and proportion of nonwhites in the population were used for each community.40 The remaining operational procedures to be discussed are the statements of the hypotheses of this thesis, stated in their operational and testable form. The list of Operational statements is lengthy and not always easy to follow without referring back to the arguments which generated them. Nevertheless, their inclusion at this point is in 39Duncan and Reiss, pp. cit., pp. 274—345. 40The data for nonwhites and percentage of labor force female were taken from Tables 77 and 75, respectively, U.S. Census 9: Population: 1299, General Social and Economic Characteristics. 92 keeping with the flow of this thesis, i.e., from the general symbolic realm to more precise statements, and finally to testable operational statements. Each hypothesis will be presented again.in the next chapter along with the data and discussion that relate to its statistical test- ing. We draw attention to this point in that the operational hypotheses may be easier to evaluate in the context of their presentation in the next chapter rather than in the immediately following presentation. Operational Statement of Hypotheses Given the preceding operational definitions, the following operational forms of the hypotheses can be stated. Hypothesis 1 - Distribution Income inconsistency will occur with greater freguency than occupation p3 education inconsistency. Hypothesis 2 - Distribution Education inconsistency will occur with less freguency than occupation p3 income inconsistency. Hypothesis 3a - Size and Distance (Consistent Communities Only) nearest §M§Aipppgppp freguency p£_lpy ranked consistent communities. The following hypothesis is a more powerful operational statement, i.e., it takes into account more information than the preceding hypo- thesis. HyPOthesis 3b — Size and Distance (Consistent Communities Only) nearest SMSA and the consistency rank 2E communities. 93 Hypothesis 4a - Size and Distance (Consistent Communities Only) Controlling for distance, there will 22.3 negative correlation _______—._—____.__————_—_. consistency. The following hypothesis is a more powerful operational statement of the preceding hypothesis. Hypothesis 4b — Size and Distance (Consistent Communities Only) Controlling pp distance, EEEEE.E$$1.§E.E positive correlation Hypothesis 5a — Size and Distance ppppg suburbs 10,000 pp 19,000 pp_pppp, ppppp yppp_pp p greater freguency pp strata consistent communities pppp.ppppg communities Hypothesis 5b - Size and Distance épppg suburbs 10,000 pp 19,000 13 pips, ppppp Kill.22 p greater freguency pp strata consistent communities pppp_ppppg communities pp_ppylppppp distance pppp nearest ggpp. Hypothesis 6a - Size and Distance ppppg suburbs 10,000 pp 19,000 pp pppp, ppppp_ypppippip greater freguency pp_ppgp ranked consistency pppp ppppg communities pp_ppy afterw- Hypothesis 6b — Size and Distance ppppg suburbs 10,000 pp 19,000 pp_pppp, ppppp Kill 23 p greater freguency p§_ppgp ranked consistency pppp.ppppg communities pp aggrega- Hypothesis 7 - Size and Distance (Inconsistent Communities Only) Suburbs will have 3 greater freguency pp occupation high 94 inconsistency pppp_communities pp_ppy ppppp distance pppp pppp nearest SMgp. Hypothesis 8 — Size and Distance (Inconsisrent Communities Only) Suburbs pppp_ppyp p greater freguency pp_income ppgp inconsistency pppp communities pp_ppy_ppppp distance pppp nearest pypp. Hypothesis 9 - Specialization Excluding communities pppp 10,000 pp 19,000 pp_§ppp, ppp— specialized communities Eill.§éi£.é greater freguency pp consis- pppp communities pppp pppp specialized communities. Hypothesis 10 - Specialization Looking pppy_pp nonspecialized communities, ppppp_p§_p_mu1tiple correlation between ppp independent variables, decreasing ppp; _____—__———————_ —— the dependent variable increasing degree pp consistency. Hypothesis 11 - Specialization Excluding communities specialized pp_manufacturing ppp education, specialized communities 10,000 pp 19,000 ifl.§l§£.fllll.h§l§.§ greater freguency pp consistency pppp nonspecialized communities. Hypothesis 12 — Specialization Communities specialized pp_manufacturing pip; ppyp_p_greater freguency pp income ppgp inconsistency pppp_plp.ppppp communities. Hypothesis l3 - Specialization Communities specialized pp education pppp_ppyp_p_greater freguency pp education high and income low inconsistency than all other communities. Hypothesis l4 — Specialization Looking only pp communities specialized pp education, there will 95 pp_p_positive correlation between ppp.pppp_pp_ppp community ppp ppp degree pp_consistency. Hypothesis 15 — Females Excluding communities pp_ppp_ppppp guintile pp_ppp income pppf proportion pp_ppp ppppp_ppppp_pppp_pp_female ppp_ppp_freguency pp consistency. Hypothesis 16 — Females (Inconsistent Communities Only) Excluding communities pp ppp_ppppp guintile pp ppp income pppf tribution, ppppp_pppp pp_p positive correlation between ppp proportion pp_ppp_lpppp_ppppp_pppp_pp_female ppp_ppp freguency pp income ppgp inconsistency. Hypothesis 17 — Females Excluding communities pp_the upper guintile pp_the income dis— proportion pp the labor force that pp female and the degree pp consistency. Hypothesis 18 — Nonwhite There will pp p positive correlation between the proportion nonwhite 23 communities and the freguency pp_consistency. Hypothesis 19 — Nonwhite (Consistent Communities Only) There will pp_p_negative correlation between the proportion —__—.___—_—____—__—_ Hypothesis 20 — Nonwhite There will pp'p positive correlation between the proportion _.___..______—___ 96 Control Variables For the most part the controls to be taken into account have been specified in the statement of the hypotheses. Nevertheless, region, an important and pervading factor, must be given attention. The variable of region is frequently used as a control in studies similar to the present one. Duncan, Cuzzort, and Duncan41 have indicated that region is most frequently a control on our ignorance. That is, region includes so many complex dimensions in different relationships with each other it is often unclear what, Specifically, is being controlled. To function as an explanatory variable it would be necessary to specify the phenomena that vary by region and that were also of importance.to the particular problem. The apparent point is that if a researcher were aware that regions varied, say in regard to industrial composition, it would be necessary to control for industrial composition and not region. Duncan and Reiss state the problem appropriately: Perhaps region may be thought of as a "contextual" rather than an "analytical" variable. The four nominal regions differ widely as a result of many causes-~geographic, historical, cul—, tural, and economic. Therefore, in "controlling” region it is by no means clear just what factors are being controlled, since regional differences may reflect differences in climate, ethnic background, crops produced, traditional customs, and a host of other factors.4 In the present study region will consistently be used as a "contex— tual” control variable. Although the hypotheses are stated and will be tested for the nation in general, the tables will also depict tests of W 41Otis Dudley Duncan, Ray P. Cuzzort, and Beverly Duncan, Statistical Geography (Glencoe: The Free Press, 1961), p. 146. ———_——__ 42Duncan and Reiss, pp. cit., p. 30. 97 the hypotheses by region. The procedure may contribute to the development of more specific information about regional differences and to limiting conditions of the hypotheses. When the number of cases in a particular analysis become too small to break down by region, only national data will be presented. Diagrammatic Presentation Figure 1 presents a summary statement of the present and preceding chapter. On the left of the diagram are the major constructs of our con- ceptual framework. The use of upper case letters with the first three is to indicate their level of generality in the scheme. The broken lines indicate relationships between constructs that are, at present, based mostly on unverified conceptual arguments. The single, solid lines indicate connections between constructs for which there is empirical support; although, the amount of support varies substantially. The double lines indicate the tie between the constructs and their operationalized statements. The use of the letter "v," along with the same number used to identify the construct, stresses the methodological translation from one language to another; the translation is from a strictly verbal system of symbols to a quantitative system of symbols. Statistical Analyses The statistics used in testing for significant differences and measures of association between appropriate variables are all standard measures. The'test or tests used will be indicated with the presenta— tion of the results of each hypothesis. The .05 level will be used for making decisions about statistical significance. 98 wemgom Hmsummocou ego wo :OHumucmmme oHumEEmprfiQ «99 x(\ mEoocH was mnowummsooo .COHumosvm How mmfiwucflno mo mcflxsmm m>flumsmano mHQMHum> uawwaoawm_mmv 'lntllilluilllrwll: mmflnumsvcH voflMfloomm .H muswflm \\Wo J kocmumfiwcoocthodmumflmfloo wumuum muHGSEEou GU m0 dfl wo>oamfim cOHuuoaoum 2 Hoozum as woaaoucm wNIoN wow< mo GOHuuoaoum udmwcomowcH \ coquNH who-.m mmmwm wwn_uwm< ZOHHHHMMEOU - 525. 08 . omawvu.pnm douflaoaou o2 \\\\\\\dHCNHVRQuwa:m \ k \ Hmscfl fiamwuo \\\\\\\\\/ \\ / .l I.II.I.uV 11.1 Ivmh.l.l Ho // / \ /, momHQ A QmNHQ< as as sales as ea sense amuse aeuee amiss sales sales assesses unoumwwflou HmuOH mefifim56£fi.fiw unww m< \ \ a.ee w.~e a.mm m.mm A.Hs memuomHV . \ \ \ elm woodcuwwo A e V m am m \ \ \ \ o.os m.He m.~s m.wm A.as aeH1ooaV A a V w.me \ e consumes m .Ne \ \ \ \ \ m.~e w.Hm A.co m.em A.Hs mmIomV Ame V a m museums: w.me \ o.o o.ooa o.ooa o.ooa. o.os o.oa N.ae H.am A.Ha om snauasV Ase V N assesses eeNV o.~e o.ma o.oe a.mw w.sw w wm m mm e.~m m.ma e.em mausaem A assesses solos owlom males solos em10m meroe AWWu %A N owmuo>< m Iausafioo Houoa mpdmwsoca oH ouwm m¢ ADV fie \ \ .\ \ x as ca 98 ad Teamemucnd ©lm WUUGNHmHH— A20 o.o \ \ \ \ e. oé 9.0 06 96 43.3783 q muflmuwwn Amm v m.m \ \ x x x 0.0 0.0 m.NH m.m A.fla mmlcmv m mafiwumwn Aom V H.N0 \ « 0.00M 0.00M 0.00M 0.00M w.N¢ 0.mm m.m0 A.wfi 0m uflguwhv N mummumwn Aemfiv «.00 0.0 m.mm 0.0m m.mm 0.0m «.HN q.m¢ m.~m 0.Hw mnusnam A20 moamumwm 00:00 00:00 anION 00100 0mlom malew 00:00 mmlom mfllcm communes mofiu kg N wwwum>< desafioo uamu wfidwmflonfi :H wNflm Imwmfioo HmuOH mmum cowmom Loom kn pom aoaumz ecu wow xamm mucmuwflmcou manmwum> ucwvconmm mg“ oufltssoz uaoo pom .<1 ><><3 NIX} ><><><1 >< X XX 175 Size Category Community 1 2 3 State Total California (cont.) Watsonville X West Covina X West Hollywood Whittier ><><1 Colorado 6 Boulder X Durango X Fort Collins X Longmont X Pueblo X Wheat Ridge X Connecticut 11 Briston Enfield Groton X Hamden X Meriden X Middletown New London Norwalk X Stratford Wallingford West Hartford X XX ><>< ><1>< Delaware Newark X Florida 16 Brownsville X Carol City X Coral Gables X Fort Lauderdale X Fort Myers X Fort Pierce Hollywood Lakeland Leesburg X Miami Beach X Palatka X Panama City X Pensacola St. Augustine X Tallahassee X Winter Park X ><><>< 176 Size Category Community 1 2 3 State Total Georgia 8 Albany X College Park X East Point X Forest Park X Macon X Midway—Hardwick X Rome X Smyrna X Idaho 4 Boise City X Caldwell X Pocatello X Twin Falls X Illinois 34 Arlington Heights X Berwyn X Bloomington X Brookfield Centralia X Champaign Danville Decatur X De Kalb East Peoria Elgin X Evanston X Freeport X Glencoe X Granite City Highland Park Jacksonville Lansing Macomb Maywood X Monmouth Naperville North Chicago Oak Lawn X Oak Park X Park Ridge Quincy Rantoul X Skokie X South Holland X Waukegan Westchester X Wilmette X Zion X N NIX ><>< XXX ><><>< NIX} NM Community Indiana Bloomington Crawfordsville Hobart Kokomo Logansport Marion Mishawaka Muncie Portage Richmond Wabash Iowa Burlington Cedar Falls Cedar Rapids Davenport Fort Dodge Mason City Muscatine Sioux City Kansas Atchison Emporia Hutchinson Junction City Olathe Prarie Village Winfield Kentucky Ashland Covington Hopkinsville Newport Paducah Richmond Louisiana Bastrop Bossier City Gretna Kenner Lake Charles New Iberia Ruston 177 Size Category 1 xx ><>< ><1>< >< ><>< ><>< ><>< 3 NIX: State Total 11 178 Size Category Community 1 2 3 State Total Maine 4 Augusta X Lewiston X Portland X Sanford X Maryland 9 Arbutus—Helethrope~Riley X Cumberland X Dundalk X Hagerstown X Hyattsville Overlea Rockville X Suitland—Silver Hill X Wheaton X ><>< Massachusetts 26 Adams X Belmont Braintree Brookline X Dedham X Everett X Fall River X Framingham X Haverhill X Lawrence X Lexington X Lynn Medford Methuen X Milford X Natick Northampton Pittsfield X Reading X Revere X Somerville X Swampscott X Taunton X Wellesley X ‘ Weymouth X Winthrop X ><>< xx XX 3 Michigan 2 Allen Park X Ann Arbor ‘ X Berkley X Birmingham X Community Michigan (cont.) East Lansing Eastlawn Garden City Grosse Point Park Hazel Park Inkster Jackson Lakeview Lincoln Park Midland Mount Clemens Oak Park Pontiac River Rouge Royal Oak St. Clair Shores Southfield Troy Wyandotte Minnesota Austin Bloomington Brainerd Crystal Hibbing Minnetonka New Ulm Rochester St. Louis Park South St. Paul Winona Mississippi Greenville Hattiesburg MaComb Meridian Missouri Bellefontaine Neighbors Clayton Columbia Hannibal Jefferson City Kirkwood Moberly Rolls St. Joseph University City 179 Size Category 1 ><><>< ><>< XX XX 2 XX >< ><>< NM 3 State Total X X X X X 11 X 4 10 X X Community Montana Anaconda Butte Great Falls Nebraska Beatrice Grand Island Norfolk Nevada Reno Sparks New Hampshire Laconia Nashua New Jersey Atlantic City Belleville Bloomfield Bound Brook Clark Cranford East Orange East Paterson Edison Ewing Garfield Gloucester City Hawthorne Hoboken Irvington Linden Livingston Maple Shade Middletown Millburn New Brunswick New Milford North Bergen Orange Paramis Pennsauken Plainfield Raritan (Monmouth County) Ridgewood Roselle Scotch Plains 180 Size Category 1 2 3 X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X State Total 3 38 Community New Jersey (cont.) South River Union City Verona Vineland Westfield West Orange Woodbridge New Mexico Alamogordo Grants Hobbs Roswell New York Amsterdam Baldwin Bellmore Cheektowaga—Northwest Copiague Dunkirk Eggertsville Elmont Freeport Fulton Glens Falls Hudson Ithaca Jericho Kingston Levittown Lindenhurst Lockport Massapequa Massena New Hyde Park New Rochelle North Tonawanda Ogdensburg Oswego Plainview Port Washington Rockville Center San Remo Schenectady South Westbury Troy Wantagh Westbury 181 Size Category xx ><>< ><><>< NM NV: 2 ><><>< XX XX ><><>< xx 3 State Total X X 4 34 X X X X X 182 Size Category Community 1 2 3 State Total North Carolina 10 Asheville X Burlington X Chapel Hill X Gastonia x Hickory X High Point X Kannapolis X Lumberton Salisbury Wilmington X ><>< North Dakota 3 Bismark Grand Forks JameStown X ><>< Ohio 31 Alliance X Athens Berea Brunswick Chilicothe Cleveland Heights X Cuyahoga Falls X Deleware X Elyria X Fairview Park X Garfield Heights X Greenville X Hamilton Lakewood Lorain Mansfield Marion X Martins Ferry X Middletown X Niles X Norwood X Oregon Reading Sandusky South Euclid Springfield Struthers X Upper Arlington Urbana Westlake Wooster ><><><1>< ><><>< ><>< x x x ><><>< r1“ 1W Community Oklahoma Bartlesville Bethany El Reno Lawton Midwest City Norman Sapulpa Oregon Albany Dalles (The Dallas in 1950) Roseburg Pennsylvania Abington Aliquippa Baldwin Bethlehem Bloomsburg Butler Castle Shannon Chester Columbia Donora Easton Ellwood City Harrison Haverford Hazleton Lancaster Lansdale Lebanon Lower Burrell Middletown Middletown (UT) Mount Lebanon Nanticoke Norristown 011 City Penn Hills Pottsville Ridley Sharon Shenandoah Stowe Uniontown Upper Darby Waynesboro 183 1 xx xx xx N><>< ><>< ><>< xxx Size Category 2 3 ><>< ><>< ><>< State Total 7 37 Community Pennsylvania (cont.) West Mifflin Williamsport York Rhode Island Bristol Newport Pawtucket South Carolina Anderson Charleston Florence Greenville Orangeburg Spartanburg South Dakota Huron Tennessee Bristol Dyersburg Inglewood Johnson City Lebanon Oak Ridge Red Bank — White Oak Woodmont-Green Hills~ Glendale Texas Abilene Baytown Bellaire Brownsville College Station Denton Edinburg Grand Prairie Greenville Hurst Irving Lamesa Laredo Longview Mercedes Mesquite New Braunfels Odessa 184 Size Category 1 2 ><>< ><>< ><>< xx 3 State Total 185 Size Category Community 1 2 3 State Total Texas (cont.) Pharr X Port Arthur X San Marcos X Temple X Terrell X Texas City X Typer x Weslaco X Utah 3 Murray X Provo X Ogden . X Vermont 0 Virginia 8 Charlottesville X Covington X Hampton X Hopewell X Petersburg X Roanoke X Springfield Winchester ><><§ Washington 5 Bremerton X Kennewick Pasco Richland Vancouver X ><><>< West Virginia 5 Clarksburg X Dunbar X Huntington X Parkersburg X South Charleston X Wisconsin 13 Appleton X Chippewa Falls X Eau Claire X Green Bay X Janesville X Kaukauna X Manitowoc X Neenah X 992312931 Wisconsin (COHC-) Racine Sheboygan Two Rivers Waukesha West Allis Wyoming Casper Rock Springs 186 Size Category 2 3 X X X X X X X w APPENDIX C DISTRIBUTIONS OF DEPENDENT VARIABLES Table 35. Centile Distribution and Codes for Occupational Strata Scores Distribution of Scores Codes 31.02 and under . . . . . . . . . . . . . . . . . . . . . . . . . . 31.03 — 32.26 . . . . . . . . . . . . . . . . . . . . . . . . . . . 32.27 — 33.97 . . . . . . . . . . . . . . . . . . . . . . . . . . . 33.98 - 34.94 . . . . . . . . . . . . . . . . . . . . . . . . . 34.95 ~ 35.75 . . . . . . . . . . . . . . . . . . . . . . . . . 35.76 — 37.00 . . . . . . . . . . . . . . . . .-. . . . . . . 37.01 — 38.62 . . . . . . . . . . . . . . . . . . . . . . . . 38.63 — 41.33 . . . . . . . . . . . . . . . . . . . . . . 41.34 — 45.39 . . . . . . . . . . . .'. . . . . 45.40 and over . . . . . . . . . . . . . . KOCDNQW-L‘WNI—‘O Table 36. Centile Distribution and Codes for Educational Strata Scores Distribution of Scores Codes 1 and under . . . . . . . . . . . . . . . . . . . . . . . 2 — 9.8 . . . . . . . . . . . . . . . . . . . . . . . . . 9 — 10.3 . . . . . . . . . . . . . . 4 — 10.8 . . .v. . . . . . . . . . . . .9 — 11.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 — 11.8 > . . . . . . . . . . . . . . 9 — 12.0 . . . . . . . . . . . . . 1 — 12.3 . . . . . . . . . . . . . . . . . . . . . 4 ~ 12.6 . . . . . . . . . . . . . . . . . . . . . 7 and over . . . . . . . . . . . . . . . . . . . . . . \OQNONU'IDUJNI—‘O 187 188 Table 37. Centile Distribution and Codes for Income Strata Scores Distribution of Scores Codes $4,312 and under . 0 4,313 — $5,103 . . . . . . . . . . . . . . . . 1 5,104 — 5,441 . . . . . . . . . . . . . . . . 2 5,442 — 5,781 3 5,782 — 6,124 . . . . . . . . . . . . . 4 6,125 - 6,557 . . . . . . . . . . . . . 5 6,558 — 7,020 . . . . . . . . .-. . . . . . 6 7,021 — 7,695 . . . . . . . . . . . . . . . 7 7,696 — 8,456 8 8,457 and over 9 Table 38. Distribution of Strata Consistency—Inconsistency Degree Scores Scores Number of Communities Per Cent ion—813% 100 59 10.95 92 133 24.68 86 64 11.87 84 54 10.02 Inconsistent 76 29 5.38 72 13 2.41 71 40 7.42 67 17 3.15 64 21 3.90 63 11 2.04 59 6 1.11 57 15 2.78 55 3 .56 49 1 .19 43 1 .19 40 l '19 35 l .19 189 Table 39. Distribution of Consistency—Inconsistency Types for the Nation and by Region Nation West North Central South Northeast Strata consistency Rank Extremely high 4 16.7% 17.9% 20.5% 7.5% 20.0% (90) (17) (32) (10) (31) High 3 17.8 33.7 13.5 11.3 18.1 (96) (32) (21) (15) (28) Low 2 13.2 4.2 16.7 12.8 15.5 (71) ( 4) (26) (17) (24) Extremely low 1 9.8 . 3.2 23.3 9.7 (53) ( 2) C 5) (31) (15) Per cent consistent 57.5 57.9 53.9 54.8 63.3 Strata inconsistency Income high 10.4% 2.1% 14.1% 2.3% 18.7% (56) ( 2) (22) ( 3) (29) Income low 13.5 11.6 12.8 26.3 4.5 (73) (ll) (20) (35) ( 7) Occupation high 4.6 1.0 3.2 12.8 1.3 (25) ( l) ( 5) (17) C 2) Occupation low 8.2 21.0 11.5 0.8 3.2 (44) (20) (18) < l) ( 5) Education high 2.2 6.3 .3 .3 0.6 (12) ( 6) ( 2) ( 3) ( 1) Education low 3.5 0.0 3.2 . 8.4 (19) ( 0) ( 5) ( l) (13) Per cent inconsistent 42.4 42.0 46.1 45.3 36.7 (N) (539) (95) (156) (133) ( 155) . Via“? .mv . . ; . . . I... .......«L..:. 2,; .1. . 7L3“ , . 1.27.3133. afrvk. . _ J.......r....;é..m . 4.2;: .435». , .::_..:.._.».. .. . ..: 2.. .. . _ 2.5.1., .5. ., r. .1 .3. ......_. .v .3 3.3.1.......\..,,... .V. ..\\.~\.... 4.... 1...... .3... w..........x.\ :13. 1...I..«.\I. 2.. if... 3...... c 4.22....