Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Economic Differentiation And Social Organization Of Standard Metropolitanareas In The United States: 1950
(USC Thesis Other)
Economic Differentiation And Social Organization Of Standard Metropolitanareas In The United States: 1950
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
T h is d is se r ta tio n h as b een 62— 3717 m ic r o film e d e x a c tly a s r e c e iv e d AM EM IYA, E iji C annon, 1 9 2 7 - ECONOMIC D IFFER EN TIA TIO N AND SOCIAL ORGANIZATION O F STANDARD M ETRO PO LITAN AREAS IN THE UNITED STATES: 1950. U n iv e r sity of Southern C a lifo rn ia , P h .D ., 1962 S o c io lo g y , g e n e r a l University Microfilms, Inc., Ann Arbor, Michigan C o p y r i g h t by •-TJI CANNON AMi-lMIYA lQtv.’ ECONOMIC DIFFERENTIATION AND SOCIAL ORGANIZATION OF STANDARD METROPOLITAN AREAS IN THE UNITED STATES: 19D0 by Fiji Cannon Amemiya A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Decree DOCTOR OF PHILOSOPHY (Sociology) June 19<o 2 UNIVERSITY O F S OU TH ERN CALIFORNIA GRADUATE S C H O O L UNIVERSITY PARK LOS A N G E L E S 7. CALIFORNIA This dissertation, written by i i j i a i t j i i y a under the direction of h~^. Dissertation Com mittee, and approved by all its members, has been presented to an d accepted by the D ean of the (Iraduate School, tn partial fulfill m eat of requirements for the deqree of D O C T O R O F P H I L O S O P H Y Dean D ate J u n e 1 9 6 2 DISSERTATION C O M M I/IX E o-ehairman ACKNOWLEDGMENTS I should like to thank the members of my committee for their time and guidance in bringing this work to com pletion, especially Dr. Georges Sabagh and Dr. Maurice D. Van Arsdol, Jr. for many valuable suggestions and ideas in corporated into this dissertation. I am deeply indebted to my wife, Frances Louise Campbell and her parents, Mr. and Mr3. George W. Campbell for their constant encouragement and their effort to pro vide the most favorable conditions to enable me to finish this work. TABLE OF CONTENTS Page ACKNOWLEDGMENTS ii LIST OF TABLES v LIST OF ILLUSTRATIONS Chapter I. INTRODUCTION 1 Statement of the Problem Definitions of Basic Terms Organization of the Remainder of the Dissertation Human Ecology and Economic Differentiation Size of Place Growth Rate Population Composition Type of Production Economic Differentiation Summary of the Review Variables of the Study The index of economic differentiation Percentage increase of total population, 19L vO to 1950 The ratio of persons under twenty and over forty-four to those between twenty and forty-four Fertility ratio Ratio of married couples to the total population Percentage of the foreign-born Percentage of female labor force Percentage of clerical and kindred workers Median number of school years completed Median income II. REVIEW OF THE LITERATURE 11 ITT. METHODOLOGY ill Chapter Page Population potential Manufacturing specialization classi fication by Duncan and Reiss The index of dissimilarity The "hierarchical” functional classi fication by Duncan and his associates The Universe of Study The Sources of Information Hypotheses and Statistics Used IV. FINDINGS ON THE GEOGRAPHICAL AND FREQUENCY DISTRIBUTIONS OF THE INDEX OF ECONOMIC DIFFERENTIATION ............................. 87 V. TESTS OF HYPOTHESES..............................102 VI. SUMMARY AND CONCLUSIONS......................... l^U APPENDIX.................................................... 173 BIBLIOGRAPHY................................................176 iv LIST OF TABLES Table Page 1. Relations between the "Ecological Complex" and Selected Components of a Given Social System ......................................... 5 2. Relationship between the Four Main Referential Concepts of the "Ecological Complex" and Their Empirical Indicators.................. I 4.8 3 . Illustration of the Index of Economic Differentiation When an Equal Distribution of Workers Is Obtained in 11 Industry Categories.................................... 52 I 4.. Illustration I of the Index of Economic Differentiation When an Unequal Distribution of Workers Is Obtained in 11 Industry Categories.................................... 53 5. Illustration II of the Index of Economic Differentiation When an Unequal Distribution of Workers Is Obtained In 11 Industry Categories.................................... Sh 6. Illustration III of the Index of Economic Differentiation When an Unequal Distribution of Workers Is Obtained in 11 Industry Categories.................................... 55 7. Illustration IV of the Index of Economic Differentiation When an Unequal Distribution of Workers Is Obtained In 11 Industry Categories.................................... 56 8. Illustration V of the Index of Economic Differentiation When an Unequal Distribution of Workers Is Obtained in 11 Industry Categories.......................... 57 9. Illustration of the Index of Economic Differentiation When Three Varying Degrees of the Distribution of Workers Are Obtained in 3 Industry Categories.................... 58 v Table 10. 11. 12. 13. 11+. 15. 16. 17. Page Industry Categories in Relation to Resource- Use and Market for Ou tp u t s ................... 60 Probable Relationships between the Index of Economic Differentiation and a Selected Group of Socio-Economic Variables ............ 75 Distribution of the Numbers and Proportions of High and Low Indexes of Economic Differentiation for All Three Categories of D, C, and A by Regions for the 56 Standard Metropolitan Areas: 1950 ........... 93 Distribution of the Numbers and Proportions of High and Low Indexes of Economic Differentiation for All Three Categories of D, C, and A by Regions for the 19 Standard Metropolitan Areas: 1950 ........... 97 Frequency Distribution of the Indexes of Economic Differentiation both of the 56 Standard Metropolitan Areas and of the 19 Sample Standard Metropolitan Areas by Categories......................................100 Pearsonian Coefficients of Correlation and Rho's between the Index of Economic Differentiation and a Selected Group of Socio-Economic Variables by Categories . . . 103 Employment In Manufacturing and Per Capita Value Added by Manufactures for Selected Standard Metropolitan Areas by Metro politan Status, Size, and 191+9 Income Level...............................................1?U Distribution of the Numbers and Proportions of High and Low Indexes of Economic Differentiation for All Three Categories of D, C, and A by Levels of Population Potential for the 56 Standard Metro politan Areas: 1950 ............................ 129 vi Table Page 18. Distribution of the Numbers and Proportions of High and Low Indexes of Economic Differentiation for All Three Categories of D, C, and A by Levels of Population Potential for the 19 Standard Metro politan Areas: 1950 ............................ 133 19. Numbers and Percentages of High and Low Manufacturing Standard Metropolitan Areas In Relation to High and Low Indexes of Economic Differentiation for the D Categories of the 58 Standard Metropolitan Areas...............................................138 20. Numbers and Percentages of High and Low Manufacturing Standard Metropolitan Areas in Relation to High and Low Indexes of Economic Differentiation for the C Categories of the 58 Standard Metropolitan Areas...............................................138 21. Numbers and Percentages of High and Low Manufacturing Standard Metropolitan Areas in Relation to High and Low Indexes of Economic Differentiation for the A Categories of the 58 Standard Metropolitan Areas...............................................137 22. Comparison of the Sizes of the Indexes of Economic Differentiation among the Various Types of the 58 Standard Metropolitan Areas on the Basis of Metropolitan Functions and Regional Relationships and t Ratios for the D Categories..................................... li+5 23. Comparison of the Sizes of the Indexes of Economic Differentiation among the Various Types of the 58 Standard Metropolitan Areas on the Basis of Metropolitan Functions and Regional Relationships and t Ratios for the C Categories..................................... II4.6 v 11 Table Page 21±. Comparison of the Sizes of the Indexes of Economic Differentiation among the Various Types of the 56 Standard Metropolitan Areas on the Basis of Metropolitan Functions and Regional Relationships and t Ratios for the A Categories.....................................11+7 25. Size Rank Order of the Average Indexes of Economic Differentiation for Each of the D, C, and A categories.....................150 26. Size Rank Order of the Average Indexes of Economic Differentiation for Each of the D, C, and A Categories.....................152 27. Distribution of the Indexes of Economic Differentiation for the Sample 30 Cities with 50,000-100,000 Inhabitants by Manufacturing and Trade Specialization . . . 171 v l i l LIST OF ILLUSTRATIONS Figure Page 1. Distribution of High and Low Indexes of Economic Differentiation by D Categories for 56 Standard Metropolitan Areas by Geographical Regions: 1950 90 2. Distribution of High and Low Indexes of Economic Differentiation by C Categories for 56 Standard Metropolitan Areas by Geographical Regions: 1950 91 3. Distribution of High and Low Indexes of Economic Differentiation by A Categories for 56 Standard Metropolitan Areas by Geographical Regions: 1950 92 Ij. Distribution of High and Low Indexes of Economic Differentiation by D Categories for 19 Sample Standard Metropolitan Areas by Geographical Regions: 1950 ................. 9U 5. Distribution of High and Low Indexes of Economic Differentiation by C Categories for 19 Sample Standard Metropolitan Areas by Geographical Regions: 1950 ................. 95 6. Distribution of High and Low Indexes of Economic Differentiation by A Categories for 19 Sample Standard Metropolitan Areas by Geographical Regions: 1950 . . . . , . . . . 96 7. Distribution of High and Low Indexes of Economic Differentiation by D Categories for 56 Standard Metropolitan Areas by Levels of Population Potential: 1950 .... 126 8. DIstrloutlon of High and Low Indexes of Economic Differentiation by C Categories for 56 Standard Metropolitan Areas by Levels of Population Potential: 1950 .... 127 Figure 9. 10. 11. 12. Distribution of High and Low Indexes of Economic Differentiation by A Categories for 56 Standard Metropolitan Areas by Levels of Population Potential: 1950 • • • Distribution of High and Low Indexes of Economic Differentiation by D Categories for 19 Sample Standard Metropolitan Areas by Levels of Population Potential: 19^0 . . Distribution of High and Low Indexes of Economic Differentiation by C Categories for 19 Sample Standard Metropolitan Areas by Levels of Population Potential: 1950 . . Distribution of High and Low Indexes of Economic Differentiation by A Categories for 19 Sample Standard Metropolitan Areas by Levels of Population Potential: 1950 . . Page . 128 . 130 . 131 . 132 x CHAPTER I INTRODUCTION Statement of the problem.--Previous studies of the relationship of social organization and urbanization indi cate that within urban societies differences in the degree of urbanization are associated with differences in the degree of economic differentiation.^ Urbanization has been shown to gradually free a society1s labor force from undifferentiated agricultural activity for specialized activities in urban communities. Population concentration in urban communities is associated normally with new needs and demands, which, in turn, are related to economic dif ferentiation. Economic differentiation appears to have further impact upon organization of urban societies. This research, which is ecological in orientation, explores the role of economic differentiation In the dif ferentiation of urban communities. Otis Dudley Duncan has described human ecology as being concerned with a complex ^Kingsley Davis and Hilda Hertz Golden, "Urbani zation and the Development of Pre-Industrial Areas," Econ omic Development and Cultural Change. Ill (1951+), 6-25; Colin Clark. The Conditions of Economic Progress (London: Macmillan and Co., Li mite cl. 191+0); Wilbert E. fooore, In dustrialization and Labor (New York: Cornell UniversTFy Press, 195l7; Wilbert E. Moore, Economy and Society (Garden City, New York: Doubleday and Company, Inc., 1955). 1 2 of four general referential concepts: population, environ ment, technology, and organization.2 This study focuses on the concept of social organization, and particularly economic differentiation, as related to other indicators of social organization and the other aspects of the "eco logical complex." It is not the concern of this work to examine in detail the theoretical utility of the concepts of this ecological system. The concepts will be used as a frame of reference in an investigation of the differentia tion of urban communities. A number of studies have been made of the economic differentiation of urban communities and societies, but their focus has been largely upon types^ of functional specialization rather than degrees of functional ^Otis Dudley Duncan, "Human Ecology and Population Studies," The Study of Population, ed. Philip M. Hauser and Otis Dudley buncan (Chicago: The University of Chicago Press, 1959), pp. 678-716. ^Chauncy Harris, "A Functional Classification of Cities in the United States," Geographical Review. XXXIII (191+3), 86-99; Grace M. Kneedler, "Functional 'Types of Cities," Public Management. XXVII (191+5), 197-203; Victor Jones, "Economic Classification of Cities and Metropolitan Areas," Municipal Year Book. XX (1953), 1+9-57; Charles S. Liebraan, "Functional Differentiation and Politioal Charac teristics of Suburbs," American Journal of Sociology. LXVI (1961), 1 4. 85-1 +90; Walter Isard, ftobert A. Kavesh, and Robert E. Kuenne, "The Economic Base and Structure of the Urban-Metro polltan Region," American Sociological Review. XVIII (1953), 317-321; Otis Dudley Duncan and Albert J. Reiss, Jr., Social Characteristics of Urban and Rural Communities. 1950 (New York: John Wiley and Sons, Inc., 195l) . 3 specialization. One of the few attempts to consider de grees of functional specialization has been made by Duncan and his associates in their use of the concept of dissimi larity of occupational structure. The Duncan work has made use of an index of dissimilarity to indicate the percentage of the labor force in an urban community which has to be shifted to different categories so as to make its distri bution equivalent to that of the United States as a whole. The concept of economic differentiation as used in this study is related to, but different from, the dissimilarity concept. Economic differentiation is here described in terms of an index that delineates the degree of equality with which the percentages of the labor force in the vari ous industries within an urban community are distributed in relation to each other. Indexes of both economic dif ferentiation and dissimilarity use the labor force in various industries as the basis of calculation, but phe nomena described by these indexes are conceptually inde pendent. Duncan and his associates did not relate the index of dissimilarity to additional indexes of socio-economic characteristics of communities and societies. In this study, however, economic differentiation, as measured by the index of economic differentiation, will be related to indexes of other socio-economic aspects of society. A u selected group of Interrelated socio-economic variables that have been found in previous research to be associated with industrialization, urbanization, and economic differ entiation will be examined in relation to the index of economic differentiation within the framework of the eco logical complex. Selected components of a given social system and their relationships to the ecological complex.--Within a given social system the four general concepts of the eco logical complex may be represented by more concrete refer ents. There are many components in a given social system under each general concept of the ecological complex, but only a limited number of components that have been con sidered important in previous studies^" will be selected for analysis in this particular study. Table 1 schematically presents the relationships between selected components of a given social system and the four general referential concepts of the ecological complex. The components of size, growth rate, and composition are selected under the referential concept of population. Size is one of the most important characteristics of urban communities; in creasing size is normally associated with the demands and ^This point will be elaborated in the review of literature. 5 needs which provide for the specialized activities char acteristic of oltles. Size of place has been found to be related to a number of soolo-econoalc aspects of society.^ TABLE I RELATIONS BETWEEN THE "ECOLOGICAL COMPLEX" AND SELECTED COMPONENTS OP A GIVEN SOCIAL SYSTEM Referential Concepts of Selected Components of the "Ecological Complex" a Given Social System (1) Size 1. Population (2) Growth rate (3) Composition 2. Environment (1) Location 3. Technology (1) Type of production I j - . Organization (1) Economic differentiation Growth rate has been found to be related to types of func tional specialization.^ The needs and demands of an ur ban community are greatly affected by Its population com position. For example, the needs and demands of an urban community with a large proportion of older population are quite different from another urban community with a large '’Duncan and Reiss, op. clt.. p. 5* 6Ibld.. p. 197. 6 proportion of young people. An urban community with a large proportion of the people In the labor force la better able to deal with community problems than one with a large proportion of dependent population. Population composition la known to be related to size of community. For example, the percentage In the agea between twenty and forty-four la related to community size.7 The component of location la selected under the referential concept of environment. Location la vital In the development of urban communities due to lta connection with resources, transportation, and markets. For example, the first cities In the United States to attain what Is regarded as metropolitan size were located at junctions of a water with land transportation. The component of type of production Is chosen under the referential concept of technology. Types of production are associated with the industrial technology In urban communities, and have been found to be related to level of income and age composition.^ The component of economic differentiation is 7Ibld.. p. 1+7. ®Leo F. Schnore and David W. Varley, "Some Concom itants of Metropolitan Size," American Sociological Review. XX (1955), pp. I+08-I+1I+. ^Duncan and Reiss, op. clt.. pp. 256-269. 7 selected under the referential concept of organization. Due to slzef density, and multitude of human contacts, urban communities are differentiated internally. One of the more Important aspects of urban differentiation la eoonomle differentiation. There Is an Indication that urbanization as such Is associated with economic differ entiation.^® More detailed discussion of the Indexes of economic differentiation and dissimilarity and their Interrelations will be found In Chapter III. DEFINITIONS OF BASIC TERMS Standard Metropolitan Areas (SMA*s).--An urban oommunlty has been defined by Louis Wlrth In general terms as "a relatively large, dense, and permanent settlement of heterogeneous individuals."^ In this study the urban com munities Investigated are SMA*s as defined by the United States Bureau of the Census In 1950* An 3MA Is a county or group of contiguous counties which contain at least one 10Colin Clark, The Conditions of Economic Progress (London: Macmillan and Company, Limited, 1914-0). ^Louls Wlrth, ”Urbanlsm as a Way of Life," Cities and Boolety. ed. Paul K. Hatt and Albert J. Reiss, Jr. (Glencoe7 111lnoIs: The Free Press, 1957), p. 50. 8 oity of 50,000 inhabitants or more.12 In addition to the county, or counties, containing such a oity, or cities, contiguous counties are included in an SMA if, according to certain criteria, they are essentially Metropolitan in oharaeter, sooially and economically Integrated with the central city.13 Social organization.--This concept is defined by Robin M. Williams as "the actual regularity of human inter action, no matter what speoiflo form the interaction may assume.”^ On a conceptual level social organization will be used to refer to the regularities found among the com ponents of the ecological complex; population, environ ment, technology, and organization.1^ Empirically, soolal organization will designate the regularities found among the empirical indicators of size, growth rate, population composition, location, type of production, and economic differentiation which constitute seleoted components of the four general referential concepts of the ecological complex. 12U.S., Bureau of the Census, Seventeenth Census of the United States: 1950. Population. II. 5-v. 13Ibid. -^Robin M. Williams, American Society (Hew York: Alfred Knopf and Company, Inc., 1^51 J* P* 33* i c ^Otls Dudley Duncan, loc. olt. 9 Economic differentiation.--This concept refers to differences In the numbers and kinds of activities in the field of production, distribution, and consumption of material goods and services. In this study, however, it will be used in a narrower sense to mean differences among urban communities in the distribution of various industries as measured by the index of economic differentiation. ORGANIZATION OF THE REMAINDER OF THE DISSERTATION Literature relevant to this study is discussed in Chapter II. Theories related to human ecology and economic differentiation are presented at the beginning of this chapter. The review is organized around the selected components of society under the major referential concepts used to delimit the ecological complex: size of place, growth rate, population composition, location, type of production, and economic differentiation. Chapter III deals with the methodology of the study. A series of variables are defined In relation to the selected components of society within the framework of the four referential concepts of the ecological complex. The Index of economic differentiation is described by the use of hypothetical distributions of various Industries. The derivation of the general formula is given in Appendix I. The universe of study is defined and the procedure for 10 obtaining a sample specified. A number of hypotheses are formulated concerning the relations of economic differen tiation and different aspects of the ecological complex, as well as null hypotheses and their alternative hypothe ses. Statistical tools used to test the null hypotheses are also described In this chapter. Findings are discussed in Chapters IV and V. Chapter TV deals with the geographical and frequency dis tributions of the index of economic differentiation and Chapter V with the findings related to the hypotheses. The summary and conclusions are given In Chapter VI. CHAPTER II REVIEW OF THE LITERATURE The material presented here will be organized mainly around the selected components of a given social system which constitute more concrete aspects of the refer ential concepts of the ecological complex. These compo nents are size of place, growth rate, population composi tion, location, type of production, and economic differ entiation. They have been treated extensively in previous studies and must be considered in any analysis of urban communities. Since this work is concerned largely with the study of economic differentiation within the framework of human ecology some of the important theories of human ecology and economic differentiation will be reviewed first. Human ecology and economic differentiation.--One of the early studies related to economic differentiation is Emile Durkheim*s The Division of Labor in Society.^ Unlike other sociologists of his time, who were inclined toward speculation and intuition as methods of inquiry, Durkheim sought to dea? with empirical data and endeavored l-Emile Durkheim, The Division of Labor in Society, trans. George Simpson (New York: 'the Macmillan Comp any, 1933). 11 12 to avoid value-Judgments. He maintained that increasing population density made necessary specialization in social organization. This specialization, in turn, led to a division of interests among different social categories which partly contributed to industrial conflicts. Durkheim, however, distinguished the division of labor from "social disintegration."^ The latter concept re ferred to social disorganization. The division of labor was thought to result in a decline of "mechanical solidarity" based on similar interests and a rise of "organic solidarity" based on dissimilar interests. Historically, human ecology is comparatively young as an academic discipline. It owes its conceptual frame work and much of its method to plant and animal ecology. The term "ecology" was first used by Ernest Haeckel in 1869. He called attention to the fact that the structure and behavior of organisms are greatly affected by their habitat and by their living together with other organisms of their own and other species.^ Robert E. Park and Roderick D. McKenzie were responsible largely for the de velopment of a theoretical frame of reference for human ^Emory S. Bogardus, The Development of Social Thought (New York: Longmans. Green and Company"! 19HBY. p T W T ■^Louis Wirth, Community Life and Social Policy (Chicago: The University of Chicago Press, 195>&) , p. 133. 13 ecology. The term "human ecology" first appeared In An Introduction to the Science of Sociology by Robert E. Park and Ernest E. Burgess In 1 9 2 1 . In his paper, "The City: Suggestions for the Investigation of Human Behavior in City Environment," Park called attention to the opportuni ties for empirical research, and supplied a coherent body of concepts which would help formulate problems and select, describe, and systematically Interpret facts.^ The city was viewed as a constellation of natural areas, and these units as the product of a selection process operating within It. Roderick D. McKenzie described human ecology as a study of the spatial and temporal relations of human beings as affected by the selective, distributive, and accomodatlve forces of the environment.^ The human com munity was considered to develop in a cyclic manner. Under a given state of natural resources and a given condition of arts the community was assumed to Increase in size and ^Robert E. Park and Ernest E. Burgess, An Intro duction to the Science of Sociology (Chicago: The University of cKicago'Press, 1921), pp. 161-216. ^Robert E. Park, "The City: Suggestions for the Investigation of Human Behavior In the Urban Environment," pie City, ed. Robert E. Park, Ernest W. Burgess, and ftb&'erlck D. McKenzie (Chicago: The University of Chicago Press, 1925) , pp. l-ij-6. ^Robert E. Park, Ernest W. Burgess, and Roderick D. McKenzie. The City (Chicago: The University of Chicago Press, 1925), p. 63. 34 structure until It reaches the point of population adjust ment to the economic base. The recent trends in the United States toward decentralization of some economic activities, and suburbanization of population would appear to support these notions. A perspeotivo of the city relating to economic differentiation and human ecology was provided by Louis 7 Wirth. He suggested a limited number of basic concepts in terms of which the complicated and many-sided phenomena of the city can be analyzed. The three basic concepts used by Wirth were size, density, and heterogeneity. An in crease in the number of people of a settlement beyond a certain point was seen as bringing about changes In inter action and in city character. Density reinforces the effect of size in diversifying peoples and their activi ties, and in increasing the structural complexity of society. Cities were considered as products of migration of people of diverse origins and heterogeneous occupations. Differentiation and specialization reinforce heterogeneity. Wirth suggested that urbanism as a characteristic mode of life may be approached empirically from three interrelated perspectives: (1) as a physical structure comprising a ^Louis Wirth, "Urbanism as a Way of Life," Cities and Society, ed. Paul K. Hatt and Albert J. Reiss, Jr. (Glencoe", Illinois: The Free Press, 1957)* PP« U6-63. 15 population base, a technology, and an ecological order; (2) as a system of soolal organization Involving a charac teristic social structure, a series of social Institutions, and a typical pattern of social relationships; and (3) aa a set of attitudes and Ideas, and a constellation of per sonalities engaging In typical forms of collective behavior and subject to characteristic mechanisms of social control A more recent attempt to systematize the field of human ecology, which is in the tradition of the ecological o orientation of this study, was made by Amos H. Hawley.7 Human ecology was viewed as the study both of the form and of the development of the community. The unit of observa tion is not the individual but is the population aggregate which Is either organized or In process of becoming organ ized. Ecology is not concerned with how habits are ac quired but rather with the functions they serve and the relationships they involve. Hawley distinguished demogra phy from human ecology by indicating that the former deals with vital processes in the communal population while the latter is interested In the organization of the population constituting the community. Hawley made a significant 8Ibid.. p. 58. 9 Amos H. Hawley, Human Ecology (New York: The Ronald Press Company, 1950)• 16 contribution with hia systematic account of human ecology as the study of social organization, a statement of Its basic assumptions, and many hypotheses for further re search. Size of plaoe.— Size of place is one of the most important characteristics of metropolitan communities. The larger the place, the greater will be the potential organizational diff©rentiation. In general the inhabitants may have contacts with a large number of other people, but they are likely not to know each other well. Secondary contacts will increase at the expense of primary contacts. Size contributes to more freedom and anonymity but weakens the personal and emotional control of intimate groups. New needs and demands promote specialized activities. Many students have noted that the number of communities of a given size found In a nation is inversely related to the size of community. One of the attempts to specify this relationship In a more precise form Is ZIpf's "rank-size rule.*'*'0 This rule states that the product of a com munity's population size times its rank is a constant. Duncan and Reiss found in their study of urban communities that city size was related to fertility, ■^George Kingsley ZIpf, Human Behavior and the Principle of Least Effort (Cambridge: Addison" Wesley Press, Inc'., 19^9^ , part 2. 17 dependency, percentage of married couples, percentage of females In the labor force, percentage of the foreign-born, median age, median Income, the percentage of clerical and kindred workers to the total employed, and median school years completed. M City size is Important from the point of view of city planners. The question, "What Is the optimum size of cities?" is an important one. Duncan points to the many problems that confront those who attempt an answer. The concept of optimum size varies with the 12 criteria upon which it Is based. One of the studies of population concentration that take community size into consideration was made by Georges Sabagh, Maurice D. Van Arsdol, Jr., and Hamid Zahedi.^3 They utilized an index of urban concentration based upon the Lorenz curve model. The Lorenze curve was obtained through the joint distributions of the cumulated proportions of communities of different sizes and of ^Otis Dudley Duncan and Albert J. Reiss, Jr., Social Characteristics of Urban and Rural Communities. 19i?0 (New York: John' Wiley and iSohs, Inc., 19^6) . l^otis Dudley Duncan, "Optimum Size of Cities," Cities and Society. Paul K. Hatt and Albert J. Reiss, Jr. (Glencoe, Illinois: The Free Press, 1957), pp. 759-772. 13oe orges Sabagh, Maurice D. Van Arsdol, Jr., and Hamid Zahedi, "Urban Population Concentration and Occupa tional Structure In the United States: 1900-1950," The I960 Proceedings of the Social Statistics Section. American Statistical Association, pp^ 16^-17l. 18 cumulated proportions of these communities. Coordinates on any of the points on the curve specify the proportions of communities and of populations below a given size level.^ One of the indexes obtained from the Lorenz model delimits the area between the Lorenz curve and the diago nal of minimum population concentration. This index, de noted as "S" in their study, was analyzed in relation to measures of economic structure. It was found that for 1900, 191+0, and 1950> the proportion of the labor force in primary occupations by states was inversely related to the degree of urban concentration, without any change over time. On the other hand, the correlations of secondary and tertiary occupations with urban concentration were positive for all years under study. However, there was a decreasing correlation of tertiary occupation and urban concentration over time. The Investigators Interpreted this as meaning that, with the passage of time, as the ll+The extent of urban concentration is proportional to the area between the curve formed by points and the diagonal. If the cumulated community and population pro portions are equal, all of the observed points fall on the diagonal. In this case all communities are of equal size and minimum urban concentration exists. Complete concen tration of population into a single city exists if the points of the coordinates delimit the total area of the triangle below the diagonal. Whenever the cumulated pro portions of communities are greater than the cumulated pro portions of populations the observed points fall below the diagonal and a degree of urban concentration Is obtained. 19 total society beoones more concentrated and as techniques of transportation and communication are further developed, service activities may have become more rapidly dissem inated. Since the urban concentration of population has a great Impact upon economic differentiation this type of Index serves as a useful tool In the analysis of social change. Economic analysis may take city size meaningfully into account. Clarence Schettler studied city size in relation to the distribution of eoonomlc servicesCo efficients of correlation were claculated between popula tion and the number of each kind of economic service In the community. It was found that with the exception of dry goods stores, doctors, filling stations, and possibly hotels, the coefficients were sufficiently high to Justify the prediction of the number of each kind of eoonomlc servloe In a small, growing community on the basis of the frequency of lowest ratio of persons per single economic servloe. Population growth rate.--It is known that different types of communities grow at different rates. Duncan and Reiss found that growing and declining urban communities ^Clarence Schettler, "Relation of City-Slze to Economlo Services," American Soclologloal Review. VIII (191*3), pp. 60-62. 20 stood differentially In relation to mobility, age composi tion, fertility ratio, sex ratio, the proportion of non white population, the proportion of persons in families, and the percentage of males married.1^ Urban places which increased markedly in size during the 1930-19U0 «*d I9i|.0- 1950 decades seem to have grown to a great extent through net immigration. Declining SMAfs, central cities, suburbs, and independent cities, regardless of size, had lower residential mobility rates than rapidly growing places. There was a larger proportion of persons over twenty-one years of age in the declining places than in the rapidly growing places of each size and metropolitan status group. In most cases the rapidly growing places had a signifi cantly higher fertility ratio than the group of places with low growth and suburbs. The sex ratio was signifi cantly higher in rapidly growing than in declining places in the case of independent cities and the smaller central cities and SMA’s. Rapidly growing SMA's, central cities, and independent cities of each size group show a much higher proportion of nonwhite residents than do declining SMA's, central cities, and Independent cities. There was evidence that the proportion of persons in families is smaller for rapidly growing than for declining places. ■^Duncan and Reiss, op. clt.. pp. 188-195. 21 The percentage of malea married was somewhat higher In rapidly growing than In declining places, regardless of metropolitan status or size of place. One of the difficulties in comparing growth rates in the study of the urban communities In different coun tries is that the concept of city Is not the same all over the world and the data are not based on the same units. Efforts to solve this problem have been made by Jack P. 17 Gibbs, Kingsley Davis, and Leo P. Schnore. They delin eated standard "metropolitan areas" throughout the world beginning with the listing of all administratively defined cities or continuous urban areas of fifty thousand or more inhabitants in each country or territory. These localities were designated as "principal cities." Administrative territorial units surrounding each principal city were then considered with respect to the percentage of the economi cally active population engaged in agriculture and the distance from the city. The following criteria must be met in order for a territorial unit to be included in the metropolitan area: ■^Jack P. Gibbs and Kingsley Davis, "Conventional Versus Metropolitan Data in the International Study of Urbanization," American Sociological Review. XXIII (1958), 5014.-511*.; Jack P. Gibbs'and Leo P. Schnore, "Metropolitan Growth: An International Study," American Journal of Sociology. LXVI (1960), 160-170. 22 (1) to touch upon the principal city or territorial unit already included, (2) to have at least 65 per cent of ita economically active population engaged in non-agricultural industries, and (3) to be close enough to the principal city to make commuting feasible. Per capita consumption of energy is used as an indicator of levels of industrial ization. The major finding was that higher rates of metro politan growth are found in the underdeveloped areas, and lower rates of growth are found in the industrialized por tion of the world. It was also noted that while the metropolitan areas in less industrialized parts of the world claim only a small share of total growth, the amount is far more disproportionate than in metropolitan areas of countries and territories with high levels of energy con sumption. By using comparable and functional units of analysis this type of study makes the study of world popu lation trends and their concomitants more meaningful. Population composition.--Louis Wirth mentions het erogeneity as one of the chief characteristics of urban community along with the other two characteristics of size l8 and density. Urban communities are inhabited typically ■ j Q Louis Wirth, "Urbanism as a Way of Life," American Journal of Sociology. XLIV (1938), pp. 1 —21+. 23 by migrants of diverse origins. The heterogeneity of popu lation origins along with heterogeneity of urban occupa tions contribute to the economic differentiation of urban communities. It has already been mentioned that urban com munities stand differentially in relation to age composi tion, the fertility ratio, the sex ratio, the proportion of nonwhite population, the proportion of persons in 19 families, and the percentage of males married. Location.--The importance of location is obvious from the standpoint of accessibility to resources and markets. Charles H. Cooley’s early view that population and wealth tend to collect wherever there is a break in transportation is given some empirical support by Leo F. PD Schnore. Although a little oversimplified, a theoreti cal framework for study of the distribution of settlement was provided by the "central place" concept of city loca tion by Christaller and Loesch. Christaller’s view was expounded by Edward Ullman in "A Theory of Location for 21 Cities." Ullman’s main point was that a certain amount ^Duncan and Reiss, op. cit. . pp. 188-195. ^°Clted by Otis Dudley Duncan, W. Richard Scott, Stanley Lieberson, Beverly Duncan, and Hal H. Winsborough, Metropolis and Region (Baltimore: The Johns Hopkins Press, 1^0}~ P* 2I4.. ^Edward Ullman, "A Theory of Location for Cities," American Journal of Sociology. XLVI (19U1) pp. 853-86I 4.. 21* of productive land supports an urban center and that the center exists because essential services must be performed for the surrounding area. The services were called "cen tral” functions by Christaller, and the settlements "cen tral" places. Loesch conceived a system of town locations as functions of distance, mass production, and competition and a hierarchy of central places In terms of size and pp differentiation with respect to function. One of the useful concepts of location in the sense of accessibility of urban community to the population in general is popula tion potential.^ An extensive study using this concept was made by Duncan and his associates.2^ They found that the proportion of the labor force employed in manufactur ing in hinterland SMA1s Increased with rising values of population potential. Activities closely related to re sources were more highly developed In the less accessible of the SMA's, while second stage resource using Industries were more conspicuous in the more accessible of the SMA's. It was also found that high population potential was con ducive to a high proportion of land in farms. 22August Loesch, The Economics of Location, trans. William H. Woglora with the assistance of Wolfgang P. Stolper (New Haven: Yale University Press, 1951*). p-i ^This will be explained more fully in Chapter III. ^■Duncan et. al. , op. clt. 25 Ttpo of production.--This is a component of society selected under the referential concept of technology. One of the chief characteristics of modern urban communities is the application of advanced technology to the production of material goods and services. The Gibbs-Schnore work indicated that higher rates of metropolitan growth were in the underdeveloped areas and lower rates of growth were in the industrialized 2C> areas. Colin Clark found that in every country the pro portion of working population engaged in secondary industry appears to rise to a maximum and then to decrease, which indicates that,after a stage of maximum industrialization is reached,industry begins to decline relative to tertiary production. It was also found by Duncan and his asso ciates that manufacturing specialization is associated with 27 areas of high population potential, ' It Is therefore apparent that production type is associated with population growth rate, occupational structure, and location. Economic differentiation.--In this study the com ponent of economic differentiation is emphasized within ^Gibbs and Schnore, loc. olt. 26 Colin Clark, The Conditions of Economic Progress (London: Macmillan and Company, Limited, l9U°)• ^Duncan et. al. . op. clt.. p. 12. 26 the referential ecological concept of organization. As previously mentioned, differences In the degree of urban ization are associated with differences in the degree of economic differentiation. Economic differentiation has great impact upon other socio-economic aspects of society. While part of the study of community structure but in the tradition of empirical economics, the work of Clark provides one of the industry classifications upon O Q which the index of economic differentiation is based. After an investigation of conditions in England and else where, Clark reached a number of conclusions that seem to have provided a starting point for additional analyses by sociologists. He classified industries into three cate gories: (1) primary, (2) secondary, and (3) tertiary. Agriculture, forestry, and fishing are included in primary industry; manufacturing, mining, and building in secondary industry; and commerce, transport, services, and other economic activities in tertiary industry. Clark estab lished that in the industrial countries where an upward trend of population is advantageous, fertility rates have fallen, while in the agricultural countries where an in creasing population is economically disadvantageous, fertility remains high. A high average level of real ^®Clark, op. clt. 27 Income per capita was associated with a high proportion of working population engaged in tertiary industries. In the United States, Canada, Great Britain, Australia, and New Zealand, the proportion of the working population engaged in secondary industry is declining, and the proportion engaged In tertiary Industry is increasing. In every country the proportion of the working population engaged in secondary industry appears to rise to a maximum and then begins to fall, which indicates that after a stage of maximum industrialization is reached industry begins to decline relative to tertiary production. Substantial occupational changes were noted in Clark's analysis: gradual elimination of the manual worker, particularly the unskilled workers; the rapid growth of the number of clerical and professional workers; and substantial movements of relative earnings accompany these large-scale movements of labor between industries and occupations. The volume of production per unit of labor in primary industry ha3 grown rapidly in all countries. In careful comparisons of production statis tics, industry by industry, in five countries no correla tion was found between average size of plant and high out put per capita. Increases in production per capita in any Industry were largely dependent upon the relative rate of growth of the Industry as a whole. 28 Studies of functional specialization have tended to emphasize types of functional specialization rather than degrees of functional specialization. One of the first attempts in the study of types of functional spe cialization was made through the classification of cities according to types of economic function as based on census data. The original effort in this field was Chauncy Harris' "A Functional Classification of Cities in the pq United States," Harris' classification was revised later by Grace M. Kneedler.-^0 Her technique, in turn, with some changes, was applied to 1950 census data by Victor Jones.'1 Jones classified SMA's and cities of 10,000 or more inhabitants according to the following character istics: (1) metropolitan status of the city, (2) rent level of central cities and suburbs, (3) economic base, and (5) percentage of dwelling units in structures built between 191*0 and 1950. Nine types of cities were ^Chauncy Harris, "A Functional Classification of Cities in the United States," Geographical Review. XXXIII (19U3), PP. 86-09. ^°Grace M. Kneedler, "Functional Types of Cities," Public Management. XXVII (191*5), pp. 197-203. -^Victor Jones, "Economic Classification of Cities and Metropolitan Areas," Municipal Year Book. XX (1953), pp. 1*9-57. 29 distinguished by metropolitan status. The city is inde pendent (I) if it is located outside a SMA. It is central (C) if the Bureau of the Census labels it as a central city of a SMA. It is defined as a suburb (S) if it is located in a SMA but is not a central city. Each of these three types of cities is further classified as a dormitory or residential city (D), a balanced city (B), or an employing city (E). Suburbs and central cities are further classified according to whether or not the median monthly rental was considered to be low, intermediate, high, or exclusive. Classification by major economic base is made by using the data on employment in the city derived from the 191+7 census of manufactures and the 191+8 census of business. The six categories used in this classification are not mutually exclusive. (1) Manufacturing (Mm): employment in manu factures is 50 per cent or more of aggregate employment in manufactures, trade and service establishments (excluding hotels and amusements), and retail trade employment is less than 30 per cent of aggregate employment. (2) Industrial (M): the dominance of manufacturing is balanced by a sub stantial retail trade and the manufacturing ratio is over 50 per cent but retail trade is more than 30 per cent of aggregate employment. (3) Wholesale (W): the number employed in wholesale trade is at least 25 per cent of 30 aggregate employment. (I4.) Retail (Rr): If the number employed In retail trade Is greater than employment In any other category,and employment In manufacturing Is less than 20 per cent, the city Is classified as a retail trade center. (5) Diversified (Mr) and (Rm): In one type of diversified city (Mr) manufacturing Is predominant and retail trade Is second in Importance; this group Includes all cities in which employment In manufacturing is less than 50 per cent of the aggregate employment but greater than employment in retail trade; the other type Is a diversified city (Rm) with retail trade predominant and with employment in manufacturing more than 20 per cent but less than $0 per cent. (6) Dormitory (D): this category has already appeared in the section on metropolitan status. The same criterion is used for establishing metro politan status and for determining the economic base of residential cities. On the basis of the data on occupation of residents in the census of population, a number of classes are established. (1) Education centers (Ed): if total 1950 college enrollments exceeded 20 per cent of the total 1950 population, the city is classified as an educa tion center. (2) Government centers (G): if 15 per cent or more of the labor force resident in a city is employed in government service, the city is classified as a govern ment center. (3) Mining towns (Mg): if 1$ per cent of 31 those In the resident labor force Indicated raining as an occupation, that city Is classified as a raining town. ( l | . ) Transportation centers (T): if more than one-fourth of the resident labor force is engaged in transportation and related functions, the city is classified as a trans portation center. (5) Amusement or health resorts (X): If the cities In which more than 10 per cent of the resi dents had occupations classified In the following industry group--eatlng and drinking places, hotels and lodging places, and amusements, recreation, and related services-- these cities are classified as amusement or health resorts. It was found that in 1950 almost one-third of the 992 cities over 10,000 population in the United States were manufacturing cities; one-fifth were industrial cities or diversified cities; slightly more than one-third were either retail cities or cities In which retail trade pre dominated. In fifteen cities, employment in retail trade and in manufacturing were about equal. The remainder were predominantly mining, transportation, education, govern ment, or resort cities. Of the 992 cities, 521 were in dependent cities; l8L | _ were central cities of SMA*s; and 287 were suburbs of central cities. A smaller proportion of independent cities than of central and suburban cities were manufacturing or industrial cities or diversified cities In which manufacturing predominated. 32 On the other hand, a slightly larger proportion of independent cities were retail or predominantly retail cities. Over one-fifth of the independent cities, as compared with one-seventeenth of the central cities and almost one-twelfth of the suburban cities, were classified as raining, transportation, education, government, or resort cities. A larger proportion of SMA's were manufacturing, industrial, or predominantly industrial than were the central cities of these areas, and a larger proportion of the central cities than of the SMA's were retail or pre dominantly retail. Over one-half were employing cities. Almost one-fifth were dormitory cities. Over one-fourth were balanced cities. Four out of every ten SMA's were classified as employing metropolises. This fact suggests that the SMA may not coincide with the commuting metropoli tan labor market area. The functional classification ap proach by Harris, Kneedler, and Jones used very arbitrary criteria as the basis of classification, and no attempt was made to indicate how this functional classification may be used profitably in relation to other variables. Duncan and Reiss gave a more definite rationale to their classification of functional specialization by com bining the economic base approach and the functional 33 classification approach.They followed Richard Andrew*a definition of economic baae aa the export actlvltlea of a community that brlnga In Its net earnings and enables It to continue aa an Independent economic entity. Recognizing that the export function cannot be measured directly or precisely with census data, It was assumed that when a community has a high proportion employed or a high per capita output In a given industry relative to other com munities in the economy, It probably exports the products of that industry. Rough Inferences as to export functions were made on the basis of statistical data on the Indus trial composition of the labor force or on per capita volume of activity. Duncan and Reiss made more profitable use of their classification of functional specialization than Harris, Kneedler, and Jones, by analyzing its relationship to other variables. Their study indicated ways in which current concepts and hypotheses of community structure may be made specific and operational. Community was defined as the territorially oriented complex of human relationship through which a more or less localized population meets its sustenance and residence requirements.33 The export 3^Duncan and Reiss, op. clt. 33Ibld.. p. xiii. 3k activities considered were as follows: (1) manufacturing, (2) trade, (3) higher education, (1 +) public administration, (5) transportation, (6) entertainment and recreation, and (7) services rendered to military establishment. In addition to the use of export activities as criteria, the procedures of their study for classification of functional specialization differ from those of Harris, Kneedler, and Jones on two points: (1) the criteria of functional spe cialization are varied according to community size and metropolitan location; and (2) it is possible for a given community to appear in more than one functional class. The nine classes of functional specialization utilized by Duncan and Reiss were as follows. (1) Whole sale trade center: a high per capita value of sales in wholesale trade but a low per capita value of retail sales. (2) Retail trade center: a high per capita value of retail sales with a low per capita value of sales in wholesale. (3) Trade center: a high per capita value of sales in both wholesale and retail trade. ( l | . ) Trade center, wholesale: a trade center with some dominance of wholesale over retail trade. (5) Trade center, retail: a trade center with some dominance of retail over wholesale trade. (6) Non trade center: a low per capita value of sales in both wholesale and retail trade. (7) Nontrade center, whole sale: a nontrade center with somewhat more wholesale than 35 retail trade. (8) Nontrade center, retail: a nontrade center with somewhat more retail than wholesale trade. (9) Maintenance trade center: the per capita value of wholesale trade and retail sales is about average for that of all cities, and it is therefore roughly considered to be a level of trade necessary to maintain the local popu lation. One of the important findings which supports Colin Clark*s observation that urbanization tends to increase proportions of workers in tertiary industry was that there is a direct correlation between community size and the pro portion of the labor force engaged in white collar occupa tions. Another was that the socio-economic status pattern which characterizes a functionally specialized type of place tends to hold for all metropolitan status and size of place groups, thus opposing current thinking that suburbs are a special category of community. Suburbs seem to be only corporate extensions of a community showing differences in functionally specialized groups, much as the community areas of a central city would show with a more detailed analysis of the central city*s internal structure. The study by Duncan and Reiss is especially significant in that it has suggested important socio-economic variables in relation to which the index of economic differentiation will bo analyzed. 36 The work by Duncan and Reiss and similar studies indicate that not only functionally specialized central cities but also functionally specialized suburbs vary with respect to socio-economic characteristics. To test this relationship further Liebman made a survey of twenty-one suburbs in Cook County. The question raised was: Do suburbs, distinguished on the basis of their economic functions or on the basis of their employment-residence ratios,vary with respect to their political character istics? The classification method employed by Jones and Collver^ was utilized in classifying these suburbs by economic function. The following were the functional specialization included: (1) retail only, (2) retail with some manufacturing, (3) total retail, () manufacturing with "substantial” retail activity, (5) manufacturing ^Charles S. Liebman, "Functional Differentiation and Political Characteristics of Suburbs," American Journal of Sociology. LXVI (1961), 1+85-1+90. ^The employment-residence ratio is the ratio be tween workers in manufacturing and trades located within a city and the total resident labor force in that city em ployed in manufacturing and trades. Cities whose ratio is less than 85 are called "dormitory cities"; those whose ratio Is between 85 and 115 are "balanced cities"; and cities whose ratio exceeds 115 are "employing cities." ^Victor Jones and Andrew Collver, "Economic Clas sification of Cities and Metropolitan Areas," The Municipal Year Book. 1959 (Chicago: International City Managers* Association, 1959), pp. 67-77. 37 only, and (6) total manufacturing. Total retail and total manufacturing correspond to Schnore1a non-industrial sub urbs and industrial suburbs respectively.^7 The political variables considered were the cityfs form of government, the incidence of partisan or non-partisan election in selection of city councilman, the partisan vote in national elections, the number of full-time city employees per capita, the Increase in the number of full-time city em ployees per capita over the last few years, the number of municipally owned off-street parking spaces per capita, general revenue per capita, and general expenditures per capita. These variables were selected from sources such as The Municipal Year Book-^ and census data. It was found that there was no significant variation between functional types of suburbs in Cook County with respect to political characteristics. Liebman made clear that the dominance of different groups as measured by their quanti tative relationship to one another does not always imply anything about political outcomes. The service classification of American cities by 3?Leo P. Schnore, "Satellites and Suburbs," Social Forces. XXXVI (1957), 121-127. ■j O Jones and Collver, loc. clt. 38 Howard J. Nelson^ is a modification of the methods used by Harris, Kneedler, and Jones. It attempts to escape from the arbitrariness of the Harris, Kneedler, and Jones criteria by employing standard deviation from the mean for each of the nine activity groups. It differs from Gillen1s study in that standard deviations are not combined to form any single overall index. Nelson believes that the pro portion of the labor force employed in a service is of much more direct significance to the economy of the city than the value or volume of sales of goods or of service performed, or similar measures for the manufactured pro ducts in a city. Standard deviations (SD*s) from the mean were calculated for each of the nine activity groups. Three degrees of variation from the average were recog nized, and the cities were grouped in their appropriate categories. Cities that are over one SD from the average in manufacturing are given a Mfl rating, over two SD*s an Mf2 rating, over three or more SD*s an Mf3 rating. A similar procedure was followed for each activity. This type of classification is useful more as a reference tool than as an end In Itself. Cities that are similarly out standing in any service can be separated for further ^Howard J. Nelson, "A Service Classification of American Cities," Economic Geography. XJQCI (1955), pp. 189-210. 39 analysis by this method. The study of SMA*s which is most directly related to this research and provides a starting point is Metropolis and Region by Duncan and his associatesThe index of economic differentiation put forth in the writer*s dissertation is analyzed in relation to the typology of SMA's, the index of dissimilarity, and population potential which are presented in their study. The Duncan work was a comprehensive and systematic ecological treatment of metro polises, their interdependence, and their relationship to hinterlands. The fifty-six SMA*s in the United States with over 300,000 population in 1950 were studied. In inter preting the metropolis heavy emphasis was placed on loca tion and function. Industry analysis was based on the detailed labor force industry statistics of the population census. Per capita volume of wholesale sales and business services receipts were used for analysis of commercial functions. Banking activities were examined on the basis of a sample of the Federal Reserve System weekly records on Interdistrict Settlement Fund clearings. Starting with a comprehensive summary of pertinent theoretical concepts, the Investigators outlined the major structural characteristics of the United States ^-°Duncan et. al. . op. clt. ^0 metropolitan economy and conducted a systematic survey of the industrial composition and regional relationships of the large cities. One of the improvements of this study over those dealing with industry classifications is its more logical and less arbitrary basis of determining functional specialization. The industry categories of the detailed classification of the population census were re grouped into twelve broader groups on the basis of their relationship to resource-use and market for outputs. The degree of functional specialization was determined by the use of location quotients. These were derived by dividing the percentages in the SMA distribution of industry by corresponding percentages in the United States distribu tion. Metropolitan and economic functions were examined in relation to other important variables such as the index of dissimilarity, the location quotient, population poten tial, city size, local urbanization, and distance to the nearest metropolis. A unique aspect of this work was the construction of a hierarchical typology of SMA1s on the basis of metropolitan functions and regional relationships. The seven classes included were: (1) National Metropo lises, (2) Regional Metropolises, (3) Regional Capitals, Submetropolitan, (10 Diversified Manufacturing Centers with Metropolitan Functions, ($) Diversified Manufacturing Centers with Few Metropolitan Functions, (6) Specialized kl Manufacturing Centers, and (7) Special Cases. A more detailed discussion of this typology will be given in Chapter IV in relation to the index of economic differ entiation. Summary of the review.--The review of the litera ture was organized around the selected components of a given social system used to stand for the referential con cepts of the ecological complex. These components are size of place, growth rate, population composition, location, types of production, and economic differentiation. Some of the Important theories concerning human ecology and economic differentiation were reviewed. Emile Durkheim emphasized the role of Increasing density of population In facilitating specialization in social organization. It was noted that human ecology has developed in close con nection with plant and animal ecology and is a compara tively young academically recognized discipline. One of the chief points emphasized in human ecology Is the Inter dependence of human population. In the United States Robert E. Park developed a theoretical frame of reference for the study of human ecology. Louis Wirth called attention to the three Important aspects of the city— size, density, and heterogeneity. An attempt to systematize the field of human ecology was made recently by Amos Hawley who viewed human ecology as the U2 study of both the form and the development of the com munity. He emphasized that the unit of observation is not the Individual but the aggregate which is either organized or in process of becoming organized. The fact that size of place is an important vari able in the study of urban communities is shown by many investigations which found city size to be related to other socio-economic aspects of society. It has been shown that a study of population concentration which takes size of community into account is facilitated by the use of an index of population concentration based on the Lorenz curve model. Size of place may be used to predict the number of each kind of economic service in a small, growing community. Duncan and Reiss1 study indicated that urban popu lation growth seems to be attributed largely to migration rather than natural increase. Growing and declining urban communities stand differentially in relation to mobility, age composition, fertility ratio, sex ratio, the propor tion of nonwhite population, the proportion of persons in families, and the percentage of males married. Location was described as an important factor from the standpoint of accessibility to resources and markets. The concept of population potential has been shown to be a useful tool in the study of location or population K3 concentrat ion. Various studies indicated that type of production is associated with population growth rate, occupational structure, and location. One of the early empirical studies of the relation ship of industrialization and urbanization to eoonomic differentiation was by Colin Clark. His major contribu tions lie In pointing out that industrialization and urbanization seem to be related to increase in tertiary industry. So far in the study of economic differentiation emphasis has been placed upon types of functional special ization rather than degrees of functional specialization. One of the first attempts in the study of types of func tional specialization was the classification of cities according to types of economic function using census data. One shortcoming of this type of classification was the arbitrariness of the criteria selected in the classifica tion. Another was the lack of effort to relate the clas sification to other socio-economic aspects of society. These deficiencies were overcome by Duncan and Reiss. They used a better rationale in the classification of functional specialization and tried to relate the types of functional specialization to other socio-economic aspects of society. Most closely related to this research and one of the few attempts to foous upon degrees of functional spe cialization is Metropolis and Region by Dunoan and his associates. The detailed Industry classifications used In the 1950 United States Census of Population were regrouped into twelve meaningful categories in terms of resource-use and market for outputs. Their index of dissimilarity and the "hierarchical” typology of SMA*a will be analyzed in relation to the economic differentiation as measured by the Index of economic differentiation. CHAPTER III METHODOLOGY As previously stated, this study uses economic differentiation as an analytical tool In the Investigation of the differentiation of urban communities within the framework of the ecological complex. The ecological com plex Includes four general referential concepts: popula tion, environment, technology, and organization. These concepts may be tied together In the following view of human society. A human population exists in an environ ment , both physical and socio-cultural, and copes with the problems posed by this environment by a set of techniques to gain sustenance from this environment and to facilitate the organization of sustenance-producing activities.^- In a given social system these four concepts may be broken down to less abstract components of society. The following components of society have been chosen because they have been emphasized in traditional ecological work and have been found to be important in the study of urban communi ties, especially with respect to economic differentiation. -^This statement was developed by this writer on the basis of the main Ideas set forth by Otis Dudley Duncan in "Human Ecology and Population Studies," The Study of Population, ed. Philip M. Hauser and Otis Dudley f)uncan (Chicago: The University of Chicago Press, 1959), pp. 670-716. U5 1*6 A detailed discussion of these components has already been given in Chapter I. The components of size, growth rate, and composition have been selected under the concept of population; the component of location under the concept of environment; the component of type of production under the concept of technology; and the component of economic dif ferentiation under the concept of organization. These components, in turn, may be further represented by "empirical indicators" as they cannot be measured directly. The focus of attention will be upon the concept of organization, especially the economic differentiation as measured by the index of economic differentiation newly developed in this study. This index will be treated as an independent variable and all other socio-economic vari ables as dependent variables. Variables of study.--The selected empirical indi cators of the components of society have been chosen be cause previous investigations have indicated their impor tance in various ways in the study of urban communities and because they are comparatively readily available. Data for describing these variables are obtained from the following: U.S. Census of Population: 1950. Vol. II, kl Characteristics of the Population.2 Social Characteristics of Urban and Rural Communities. 1950 by Duncan and Relss^ and Metropolis and Region by Dunoan and his associates.k It Is emphasized that the variables selected are not used to examine In detail the theoretical utility of the concept of the eoologlcal complex. Rather, the eco logical complex notion is regarded as frame of referenoe within which economic differentiation, as measured by the Index of economic differentiation, Is evaluated In rela tion to a selected group of socio-economic variables. Table 2 shows these selected empirical indicators under each component of society and the selected components of society under each general referential concept of the ecological complex. The Index of economic differentiation is treated as an Independent variable and all other variables as de pendent variables. These are the dependent variables: (1) size rank of SMA, (2) percentage Increase of total ^U.S. Bureau of the Census, U.S. Census of Popula tion: 1950. Vol. II. Characteristics of tfae Population (Washing^ on, D.C.: $.S. Government Printing Office, 1952). 3otis Dudley Duncan and Albert J. Reiss, Jr., Soolal Characteristics of Urban and Rural Communities. 1950 (frew frork: Jokn Wiley and Sons, Inc., 1956). ^-Otis Dudley Duncan, W. Rlohard Scott, Stanley Lleberson, Beverly Duncan, and Hal H. Winsborough, Metropo lis and Region (Baltimore: The Johns Hopkins Press, I960). TABLE 2 A SCHEMATIC PRESENTATION OF THE REFERENTIAL CONCEPTS, COMPONENTS OF A GIVEN SOCIAL SYSTEM, AND EMPIRICAL INDICATORS Referential Concepts Components of a Given Social System Empirical Indicators 1. Population 2. Environment 9. Technology 4. Organization (1) Size (2) Growth rate (3) Composition (1) Location (l) Type of production (l) Economic differen tiation fa) Size rank (a) Increase of population in 1940-1950 (a) Median age (b) Ratio of persons under 20 and over 44 to persons between 20 and 44 fc) Fertility ratio (d) Ratio of married couples to the total population fe) Percentage of the foreign-born if) Percentage of female labor force (g) Percentage of clerical and kindred workers fhj Median number of school years completed (i) Median income (a) Population potential (a) Manufacturing specialization classifica tion by Duncan and Reiss fa) The index of economic differentiation (b) The "hierarchical" functional classifica tion of Standard Metropolitan Areas by Duncan and his associates (c) The index of dissimilarity ■F' CD 1+9 population, 191+0-1950, (3) median age, (1+) ratio of persons under twenty and over forty-four to those between twenty and forty-four, (5) fertility ratio, (6) ratio of married couples to the total population, (7) percentage of the foreign-born, (8) percentage of female labor force, (9) percentage of clerioal and kindred workers, (10) median number of school years completed, (11) median income, (12) population potential, (13) manufacturing specializa tion classification by Duncan and Reiss, (11+.) the index of dissimilarity, and (15) the "hierarchical" functional classification by Duncan and his associates. The index of economic differentiation (IED).--This is the index in terms of which economic differentiation is measured. It describes how the various industries are distributed within a particular area in relation to each other when the labor force in each industry is expressed in proportional terms. Minimum economic differentiation is obtained when all industry categories are evenly dis- tributed--the proportion of each industry category Is the same. The IED value will be 0 In this case. Maximum economic differentiation is obtained when the labor force Is completely concentrated in one industry only and no others--the proportion of one Industry category is 1 and all others 0. Then the IED value will be 1. All other combinations of industry categories will result in IED 50 values ranging between the minimum of 0 and the maximum of n Is the number of Industry categories and p Is the pro portion of each Industry category. The derivation of this general formula Is given In Appendix I. applied to any number of variables when the subcategories of the variable are expressed in proportions. It always varies between a minimum value of 0 and a maximum value of 1. Hypothetical distributions of industries are given in Tables 3, 1+, 5, 6, 7, 8, and 9 to illustrate how varying degrees of economic differentiation result in varying values of the IED. Eleven and three subcategories are used in the illustration since they are the numbers of categories used In this research. bution of workers is obtained in eleven Industry sub categories. The IED value is 0. Table 8 Illustrates the IED when a given distribution of workers is found in one industry category only. The IED value is 1. All other combinations of industry distributions result in IED values ranging between the minimum of 0 and the maximum of 1. Tables I j . , 5» 8, and 7 illustrate the IED when unequal distributions of workers are obtained in eleven industry categories. The distributions of workers Illustrated in 1. The general formula One of the advantages of the TED is that it can be Table 3 illustrates the IED when an equal distri- 51 Tables 1+ and 5 are more evenly spread than those In Tables 6 and 7, and the IED values of the former two groups are smaller than those of the latter two groups. In other words, the more evenly the various Industries are distri buted within a particular area, the smaller the IED value. Table 9 Illustrates the IED when varying degrees of the distribution of workers are obtained In three Industry categories. The even distribution of Industries result in the minimum IED value of 0 and for a centratlon of workers in one industry only the maximum IED value of 1 Is ob tained. The middle group has a degree of economic differ entiation between the top group and the bottom group and has an IED value between those of the other two groups. For the purpose of comparison three different groupings of industry categories will be used for the study of the economic differentiation as measured by the IED. The first grouping of Industry categories Is by Duncan and his associates;^ the second grouping by Colin Clark;^ the third grouping by this writer. For the sake of convenience these three groupings of industry categories will be referred to as D, C, and A categories, respectively. 5lbld. ^Colin Clark, The Conditions of Economic Progress (London: Macmillan and Company, Limited, 19^6}. 52 TABLE 3 ILLUSTRATION OF THE INDEX OF ECONOMIC DIFFERENTIATION WHEN AN E'.JJAL DISTRIBUTION OF WORKERS IS OBTAINED IN 11 INDUSTRY CATEGORIES n 1 , 11 , 1.2 Category P P - — (P - — ) 10^P " 1. . . . 1 0 0 0 ‘ > c . . • » • 11 1 0 0 0 3 — 11 1 11 0 0 0 4___ 1 IT 0 0 0 0--- l 11 0 n 0 l IT 0 0 0 7--- 1 11 - i n 0 0 8___ 1 IT 0 0 0 u.... 1 IT 0 0 0 10.... 1 IT 0 0 0 11.... l TT 0 0 0 The index of economic differentiation-^y^(P - = 0 P denote;; proportion . 53 TABLE U ILLUSTRATION I OP THE INDEX OP ECONOMIC DIFFERENTIATION WHEN AN UNEQUAL DISTRIBUTION OF WORKERS IS OBTAINED IN 11 INDUSTRY CATEGORIES Category Pa P - _i n ^ - — f ir 11 - -I)2 V5 ir 1____ 2 l ( 1)2 Tt i 11 IT H o 2___ 1 TT 0 0 0 3--- 1 11 0 0 0 I*. . .. 1 IT 0 0 0 5--- 1 11 0 0 0 6___ l TT 0 0 0 7---- 1 IT 0 0 0 8___ 1 IT 0 0 0 9---- l TT 0 0 0 * • • 4 c H 1 22 -1 22 1 ii____ 1 -l i 22 22 The index of* economic differentiatlon *» ^ll/p _ ]\2 * .011+ ®p denotes proportion. 5k TABLE 5 ILLUSTRATION II OP THE INDEX OP ECONOMIC DIFFERENTIATION WHEN AN UNEQUAL DISTRIBUTION OP WORKERS IS OBTAINED IN 11 INDUSTRY CATEGORIES Category Pa p - 1 TT u(p - _y2 io' ir 1____ 2 l 1 IT IT no 2__ 2 1 ( l)2 i IT IT 'ir 110 3-- l IT 0 0 0 I*.... i IT 0 0 0 5.... 1 11 0 0 0 6___ l 11 0 0 0 7--- l IT 0 0 0 8__ l 22 -1 22 dio 9___ l -l (-1)2 1 22 22 y227 10___ l -l (-1)2 '22' l 22 22 11___ l -l (-1)2 227 i 22 22 d+o The index of economic differentiation * <ll(p - -I)2 <lo' ll7 a* .027 *p denotes proportion. 55 TABLE 6 ILLUSTRATION III OP THE INDEX OP ECONOMIC DIFFERENTIATION WHEN AN UNEO.UAL DISTRIBUTION OF WORKERS IS OBTAINED IN 11 INDUSTRY CATEGORIES Category P* p - JL (P - -i)2 11(p - -i)2 n 11 10 11 1____ 8 7 <T?>2 ^7 11 11 110 B . . . . 1 (tK ' 1 11 11 110 7--- 1 0 0 0 4 11 0 -1 (^)2 11 1 11 110 0. . . . 0 -1 {=1)2 1 11 11 110 6... . 0 -1 1 11 11 110 7--- 0 -1 (=±V' 1 8___ 0 11 -1 li (zl)2 110 1 11 n 110 9... . 0 -1 (2±)2 1 11 11 „ 110 10___ 0 -1 11 't t 5 1 TTT3 11___ 0 -1 o-i)2 l 11 11 110 The index of economic differentiation=/2A(p - — )2 \10 11 z . 809 ap denotes proportion. 56 TABLE 7 ILLUSTRATION IV OP THE INDEX OF ECONOMIC DIFFERENTIATION WHEN AN UNEQUAL DISTRIBUTION OF WORKERS IS OBTAINED IN 11 INDUSTRY CATEGORIES Category Pa P - JL (P - I)2 ii(P - -i)2 11 TT 10 11 1___ _2 (-9)2 81 11 11 110 2___ 0 0 0 3--- -1 (ii)2 1 11 110 4___ -1 (=1,2 1 II 110 l j... . -1 (TT) 1 ? 1 11 110 -1 'n 1 ^ # * • * 11 rro 7--- -1 (^i)? 11 i 11 110 8 -1 (Zi)2 i • * * * 11 110 -1 1 j.... 11 110 10___ -1 11 (=1,* 1 TIC 11___ -1 l 11 rro The index of economic differentiationz/Ai(p - _J.)2 \10 11 z .818 ap denotes proportion 57 TABLE 8 ILLUSTRATION V OP THE INDEX OP ECONOMIC DIPPERENTIATION WHEN AN UNEQUAL DISTRIBUTION OP WORKERS IS OBTAINED IN 11 INDUSTRY CATEGORIES Category Pa p - Jk (p - -i)2 li(p - -I)2 11 11 10 11 1____ 1 10 11 (XO)2 10 11 2___ 0 -1 (— )2 11 1 11 TIo A 0 -1 (zl)2 11 1 J . . . . 11 110 4___ 0 -1 11 (zl)2 11 1 rro D . . . • 0 -1 (zl)2 1 11 11 110 6. . . . 0 -1 (zl)2 1 11 11 110 7--- 0 -1 (zi)2 1 11 11 110 8___ 0 -1 11 ( ^) 2 vll' 1 110 9--- 0 -1 11 (-1)2 TT 1 TTo 10___ 0 -_1 (zl)2 l 11 11 0 110 11___ 0 -1 (zi)‘ l 11 11 110 The index of economic dif ferentiation=/H(p - __i)2 <10 11 - 1 ap denotes proportion. 58 TABLE 9 ILLUSTRATION OP THE INDEX OP ECONOMIC DIFFERENTIATION WHEN THREE VARYING DEGREES OF THE DISTRIBUTION OF WORKERS ARE OBTAINED IN 3 INDUSTRY CATEGORIES Category a P l P " 3 TJ I V > 4 ’ ro |(p - 3 1--- 1 3 0 0 0 2--- 1 3 0 0 0 3--- The index of 1 3 economic 0 0 differentiations^ (p -i 0 )2- - 0 1 . . . . 2 3 1 3 <3>2 1 Z 2 l 3 0 0 0 0 3 . ^1 3 1 6 / - 5 1 2 The index of economic differentiation=^2(P ~ v) - -333 3' l. .. . 1 2 ---- 0 3 . 0 2 2 2 2 3 3 4 <i>: * 1 <-;> j O, 1 , 2 The index of economic differentiations ^ (P - z 1 aP denotes proportion. 59 For the D categories the detailed Industry classi fication In the U.S. Census of Population: 1950. Vol. II. Characteristics of the Population was regrouped Into twelve broader categories In terms of their resource and market orientation as follows:? 1. Primary resource users; production for non-final marks t. 2. First stage resource users; production for non-final market. 3. First stage resource users; production for final marke t. I 4.. Second stage resource users; production for non-final marke t. 5. Second stage resource users; production for final marke t. 6. Resources of indirect significance; production for non-final market. 7. Resources of indirect significance; production for final market. 8. Service industries; local. 9. Service Industries; non-local. 10. Service Industries; may be local or non-local. 11. Construction. 12. Industry not reported. As the Industry classification of the 1950 Census of Population Is not logically arranged In terms of ?Dunean et. al.. op. clt.. pp. 200-209. 60 resource or market orientation, It is not conducive to Industry analysis. The basis provided by Duncan and his associates furnishes a rationale for regrouping the de tailed Industry classification to reach a meaningful analysis. Table 10 shows the set of Industry categories that are formed according to the two principles of classI- fication--type of resource-use and market for outputs-- dlsregarding service industries and construction. TABLE 10 INDUSTRY CATEGORIES IN RELATION TO RESOURCE- USE AND MARKET FOR OUTPUTSa Resource-Use Type of Non-final Market Pinal Primary resource extractors......... (1) ___b First stage resource users.......... (2) (3) Second stage resource users......... Industries for which resources have (k) (5) the most indirect significance.... (6) (7) aAdopted from: Otis Dudley Duncan, W. Richard Scott, Stanley Lleberson, Beverly Duncan, and Hal H. Winsborough, Metropolis and Region (Baltimore: The Johns Hopkins Press^ i960) , p. '203', ^The numbers correspond to those In the twelve industry categories. With respect to the C categories Colin Clark class Ified industries as primary, secondary, and tertiary (service). Agriculture, forestry, and fishing were in cluded in primary industry; mining, construction, and 61 manufacturing In secondary Industry; transportation, trade, finance, services, and public administration In tertiary Industry. The C and the A categories are Identical except that for the A categories mining Is Included In primary Industry rather than In secondary Industry on the assump tion that mining may be better considered as an extractive industry. The different types of economic differentiation as measured on the basis of the three groupings of indus try categories are expected to be associated differentially with other socio-economic aspects of society. Size rank.--Size rank refers to the ranks of SMA's when they are arranged in order of size. The higher the rank, the larger the size. Size Is one of the most Impor tant variables In the analysis of urban communities. One of the most Important attributes of a city is Its large size. The larger the number of people, the greater the potential differentiation. Contacts among people become more segmental rather than diffused. Size creates new demands and problems which Intensify differentiation within and among urban communities. Percentage Increase of total population. 196.0 to 1950.--This is the percentage Increase of total population of the SMA from 19l(.0 to 1950. Types of industry have been found to be related to this variable, which suggests the desirability of Investigating the relationship between 62 this variable and the IED. Median age.--This is the age of the person above whom there are SO per cent of the population and below whom there are SO per cent of the population when the people are ordered according to age from young to old. Age distribu tion is related to size of place and types of functional specialization. It is very likely that this variable is affected by degree of economic differentiation. The ratio of persons under twenty and over forty- four to those between twenty and forty-four.— This ratio is obtained by dividing the number of persons under twenty and over forty-four by the number of those between twenty and forty-four. This ratio is the proportion of persons at their maximum productivity to the rest of the popula tion. It is similar to the dependency ratio. The differ ence between this ratio and the dependency ratio is that the latter is the ratio of persons under twenty and over sixty-four to those between twenty and sixty-four. Fertility ratio.--This refers to the number of persons aged under five per one hundred females aged twenty to forty-four. The fertility ratio has been found to be related inversely to the degree of urbanization, the level of educational attainment, and socio-economic status. Since the degree of economic differentiation seems to be influenced by these factors the relationship of the 63 fertility ratio to economic differentiation aleo should be Investigated. Ratio of married couples to the total population.-* The number of married couples is divided by the number of total population and multiplied by one hundred. It is the number of married couples per one hundred persons in the total population. A married couple is defined by the United States Bureau of the Census as a husband and his wife enumerated as members of the same household.® The percentage of married couples in the total population is known to be associated with such things as the degree of urbanization and types of functional specialization. For example, the percentage of married couples is smaller in college communities than in some other types of community. Different industries attract differing proportions of men and women, which results in differences in the percentage of married couples among urban communities. Percentage of the foreign-born.— The percentage of the foreign-born is obtained by dividing the number of the foreign-born aged twenty-one and over by the number of the total population and by multiplying the result by one hundred. Historically, the foreign-born have clustered in large cities. They tend to be attracted to certain types O U.S., Bureau of the Census, Seventeenth Census of the United States: 1950* Population. II-5. vlll. &k of Industry and may be associated with degree of economic differentiation. Percentage of female labor force.— This is the per centage of females In the labor force to the total labor force. The United States Bureau of Census in 1950 defined the labor force as all individuals classified as employed or unemployed and members of the armed forces.^ The female labor force participation is more associated with tertiary Industry than primary or secondary industry. Tertiary industry is, In turn, strongly associated with the extent of urbanization. These relations suggest some sort of association between female labor force participation and degree of economic differentiation. Percentage of clerical and kindred workers,--This percentage Is obtained by dividing the number of clerical and kindred workers by the total employed and multiplying the result by one hundred. The category of clerical and kindred workers, as defined by the U .S. Bureau of the Census In 1950, included bank tellers, bill and account collectors, bookkeepers, cashiers, dentists' assistants, express agents, express messengers, library assistants and attendants, mail carriers, messengers, office boys, office machine operators, physician's office attendants, railway 9Ibid.. p. x. 65 mail clerks, receiving clerks, secretaries, shipping clerks, station agents, stenographers, telegraph messengers, tele graph messengers, telegraph operators, telephone operators, ticket agents, and typists. The percentage of clerical and kindred workers among employed persons appears to be positively and highly related to the degree of urbaniza tion. Duncan and Reiss believe that it is the best single occupational Indicator of the complexity of the division of labor.10 Its relationship to tertiary industry is expected to be high and it would appear to strongly influ ence and be influenced by other socio-economic aspects of society. Median number of school years completed .--This is the median number of school years completed by the popula tion twenty-five and more years of age. This variable is a measure of the general level of educational attainment. The level of educational attainment is related to such things as size of place and types of functional specializa tion. It may be affected by degrees of economic differ entiation. Median Income.--This Index is taken from the 1950 U.S. Census of Population. It is the income of the person who Is located in the middle of the population when the 10Duncan and Reiss, op. clt.. p. 5. 66 population ia arranged according to income from low to high. Median Income la aaaociated with the types of functional specialization such as manufacturing and is also associated with increase or decrease of population. Degrees of economic differentiation are likely to be asso ciated with types of functional specialization as well as with population growth. If this follows, then median in come will be also related to economic differentiation. Population potential.— This variable is used in the same sense as by Duncan and his associates.^ It Is an indicator of the generalized accessibility in terms of straight line distance of the SMA to the total population of a larger area. In this study crude Population potential is obtained Indirectly for SMA1s by superimposing the SMA1s on the population potential chart by Duncan and his 1 p associates. It Is assumed that this method will throw some light on the relationship between economic differ entiation and population potential. Strictly speaking, population potential is defined In terras of a point In space, although we speak of the population potential of an areal unit such as a city or a metropolitan area. It varies from point to point over ■^Duncan et. al.. op. clt.. pp. 561-553* 12Ibld. . p. I 4. 3. space. When a single population potential is used to characterize a SMA, it refers to a rough average of the potential values over the area. The population potential at a point la obtained by finding the distance from that point of each individual in the population living in the universe of territory under investigation, calculating the reciprocal of each distance, and summing the reciprocals. In practice, another method is used. The universe of ter ritory under investigation is subdivided into manageable number of areal units. The population of each unit is de limited on the assumption that this population is concen trated at a single central point within the areal unit. Distances are measured to L (the central point) from each such point. Population potential at L is the sum of Pi/Dl, where Pi is the population of the i'th areal unit and Di is the distance of its central point from L, and the summation is over all areal units. Duncan and his associates obtained from the U.S. Bureau of the Census an unpublished work map of the United States, showing boundaries of state economic areas. The United States was divided into iSb subareas: 116 of these subareas contained one or more metropolitan state economic areas. For the 116 subareas with metropolitan state eco nomic areas, it was assumed that the entire population of the subarea was concentrated at the center of the most 68 populous metropolitan area. For the thirty-eight subareas which did not Include any metropolitan area, a point of concentration was selected after examining a map showing the distribution of population over the subarea. Some arbitrary decisions were involved in the de lineation of the 1Sh subareas and the selection of con centration points. An acetate overlay, with a series of concentric zones each 1-3/8 inches (150 miles) in width, was prepared. The center of the set of concentric zones was placed on the point for which potential was to be determined and the population falling in each of the con centric zones was aggregated. In this way it was possible to obtain a distribution of the United States population by distance zone from the point for which population potential was to be calculated. The aggregate population in each distance zone was divided by the mid-point of the distance interval; the sum of the quotients over all zones represented the population potential for the given point. Population potential is a measure of generalized accessibility to population and an Important variable from the standpoint of market and labor supply. It seems reasonable to believe that Industrial diversification will develop in areas of high population potential. Certain Industries are likely to be more population potential oriented than others. The result would be that population 69 potential is associated with economic differentiation as measured by the IED. Manufacturing specialization classification by Duncan and Reiss,— This variable is based on classification of SMA*s into the high and the low manufacturing category by Duncan and Reiss.^ They use the qulntile limits for percentage employed in manufacturing specialization. SMA*s in the upper quintile are classified as high in man ufacturing specialization, while those in the lower quin tile are classified as low in manufacturing specialization. Since most urban communities are most specialized in ter tiary industry, high or low specialization in manufactur ing is likely either to counterbalance or to accentuate tertiary industry and to influence the pattern of indus trial distribution. Therefore, degrees of economic differ entiation can be expected to be associated with manufactur ing specialization. The index of dissimilarity.--Up to a certain point the steps in obtaining the Duncan index of dissimilarity and the TED are the same. Both use the twelve categories derived from the detailed industry classification of the U.S. Census of Population. The index of dissimilarity is an overall measure of the divergence of the labor force ■^Duncan and Reiss, op. clt.. p. 253* 70 distribution of the SMA from that of the United States as a whole. It is calculated by taking differences in per centage between the two distributions, category by cate gory, and summing either the positive or the negative dif ferences. The index of dissimilarity indicates the per centage of workers in occupational categories that must be shifted to other categories in order to make the distribu tion of a particular area the same as that of the United States as a whole. The index of dissimilarity indicates for particular SMA*s the extent to which various indus tries deviate from the national averages of their corres ponding industries; it does not, however, show how equally various industries are distributed in relation to each other within a particular area. The IED Indicates how evenly or unevenly the various industries are distributed In relation to each other within a particular SMA. The "hierarchical" functional classification by Duncan and his associates.— This refers to the classes in the classification of SMA1s with 300,000 inhabitants or more in terms of metropolitan functions and regional rela tionships. There are seven classes in this "hierarchical" typology:1^ (1) National Metropolis, (2) Regional Metro polis, (3) Regional Capital, Submetropolitan, • ’ ■^■Duncan et. al.. op. clt.. pp. 259-275. 71 (I4.) Diversified Manufacturing Center with Metropolitan Functions, (5) Diversified Manufacturing Center with Few Metropolitan Functions, (6) Specialized Manufacturing Center, and (7) Special Case. A more detailed discussion will be given in Chapter V. Since each class is related to certain industries and metropolitan functions it can easily be assumed that the seven classes, on the average, stand differentially in terms of degrees of economic differentiation. The universe of study.— The universe of study con sists of all of the fifty-six SMA*s with 300,000 or more inhabitants as defined by the United States Bureau of the Census in 1950.^ All of the fifty-six SMA1s will be studied rather than a sample. The selection of all areas will maintain comparability with the fifty-six SMA1s reported in Metropolis and Region. Another factor in Theso SMA's are New York, Chicago, Los Angeles, Philadelphia, Detroit, San Francisoo, Minneapolis-St. Paul, Kansas City, Seattle, Portland, Atlanta, Dallas, Denver, Houston, New Orleans, Louisville, Birmingham, Indianapolis, Columbus, Memphis, Omaha, Fort Worth, Richmond, Oklahoma City, Nashville, Jacksonville, Boston, Pittsburgh, St. Louis, Cleveland, Buffalo, Cincinnati, Baltimore, Milwaukee, Albany-Schenectady-Troy, Toledo, Hartford, Syracuse, Providence, Youngstown, Rochester, Dayton, Allentown-Bethlehem-Easton, Akron, Springfield-Holyoke, Wheeling-Steubenville, Charleston, W. Va., Washington, San Diego, San Antonio, Miami, Norfolk-Portsmouth, Wilkes- Barre -Hazleton, Tampa-St. Petersburg, Knoxville, and Phoenix. 72 choosing SMA!s rather than incorporated cities Is the assumption that the SMA is more suitable as unit of eco nomic analysis. Many workers in the cities do not live within city limits but in outlying areas and commute to work. Since the basic data for economic analysis in this study are obtained from the 1950 U.S. Census of Population where industry classification is based upon the occupation of residents, it is important that workers and residents match within the unit. The study by Duncan-^ illustrates this point. It was found that residents of the Chicago urbanized area number 1^,921,000 in 1950. The 2,1[(.3,000 employed workers comprised I 4I4 . per cent of the resident population. Most of the workers were involved in daily movements as workplace and residence were infrequently located in the same urbanized area. In the metropolitan United States in 195U, 18.7 per cent of the workers living In outlying counties traveled to the core county of the metropolitan area for work; although crossing a county line on the Journey to work does not necessarily mean travelling long distances between residence and workplace. The SMA Is chosen as unit of economic analysis rather than the incorporated city or the urbanized area because it Is ^Beverly Duncan, "Intra-Urban Population Move ments," Cities and Society, ed. Paul K. Hatt and Albert J. Reiss, JFI (Glencoe', Illinois: The Free Press, 1957), pp. 297-31^. 73 reasonable to suppose that most labor Inputs In the Indus tries of a SMA come from the area Itself. This point Is especially Important In view of the fact that economic analysis will be made on the basis of Industry classifi cations which are derived from the occupations of resi dents. In order to examine the Influence of the size of the SMAf a systematic random sample of nineteen SMA1s were selected from the SMA1a with Inhabitants between 100,000 and 300,000.^7 The SMA*s with Inhabitants between 100,000 and 300,000 were arranged in order of size; the first SMA was selected at random and then every following sixth SMA. The sources of Informatlon.--The major sources of data were taken from the 1950 United States Census of Population. The IED Is constructed on the basis of both the detailed and Intermediate Industry classification In the Census of Population. Most socio-economic variables that are to be investigated In relation to the IED are obtained from the Census of Population. The Index of dissimilarity and population potential are derived from the study of Metropolis and Region by Duncan and his associates. 17The se cities are Austin, Little Rock-North Little Rock, Lincoln, Montgomery, Mobile, Harrisburg, Tacoma, Bridgport, Sioux City, Lowell, Madison, Brockton, Rockford, Canton, Binghamton, Erie, Winston-Salem, Peoria, and Jackson. Duncan et. al., op. clt. Ik One reason for using the 1950 U.S. Census of Population la to maintain comparability with Metropolis and Region and Soolal Characteristics of Urban and Rural Communities. 1950. which analyses were based largely on the 1950 U.S. Census of Population. Hypotheses and Statistics used.— Table 11 shows the probable relationships between economic differentia tion as measured by the IED and the selected empirical Indicators of the components of society discussed on pages 1*5-71. As the IED is newly derived t the probable relationships between It and the selected socio-economic variables had to be deduced from the findings on other variables which seem to be related to economic differen tiation. The relationships presented in Table 11 will be more fully discussed when formulating the hypotheses. The null hypotheses will be tested against their alternative hypotheses. Previous research related to economic differentiation Indicates the directions of asso- 19 elation among the selected variables to be studied. 7 One 19 Clark, op. clt.; Duncan and Reiss, op. clt.: Duncan et. al., op. ci¥. TABLE 11 PROBABLE RELATIONSHIPS BETWEEN THE INDEX OF ECONOMIC DIFFERENTIATION AND A SELECTED GROUP OF SOCIO-ECONOMIC VARIABLES Variables That Are Likely to Be Positively Related with the Index of Economic Differentiation Size rank Increase of population in 1940-1950 Percentage of female labor force Percentage of clerical and kindred workers Median number of school years completed The index of dissimilarity Rank In the ’ ’ hierarchical'’ typology by Duncan and his associates Variables That Are Likely to Be Negatively Related with the Index of Economic Differentiation Median age Ratio of persons under 20 and over 44 to persons between 20 and 44 Fertility ratio Ratio of married couples to the total population Percentage of the foreign-born Median income Population potential Manufacturing specialization classification by Duncan and Reiss -U \n 76 tail tests will be conducted with which to compare the null hypotheses against their alternatives. Linear associations will be assumed between inde pendent and dependent variables. The 5 P®** cent level of significance is arbitrarily chosen as the point at which to reject the null hypotheses. Spearman's rank-difference correlation of coefficient will be used in order to test null hypothesis 1. Pearson's product-moment of coeffi cients will be used to test null hypotheses 2 through 12 and lit.. Chi square and the contingency coefficient will be used to test null hypothesis 13 and the t ratio will be utilized for a test of null hypothesis 15. The Fisher exact probability test will be used to test differences in the geographical distribution of the IED. Hypothesis 1: The size rank of the SMA in 1950 increases with an increase in the value of the IED. In most urban communities tertiary industry occupies the largest proportion of the employed labor force. It Is known that urbanization Is positively related to the development of tertiary industry. If it Is assumed that the larger the urban community, the higher the proportion of tertiary industry, then the proportion of tertiary industry which is already large will be all the more accentuated In large urban communities in relation to primary and secondary industries. This would result In 77 large IED values. A large urban community would, there fore, expect a large IED value. Null hypothesis 1: The association between values of the IED for SMA*s In 1950 and the size rank Is zero. Alternative hypothesis 1: The association between values of the IED for SMA* s in 1950 and the size rank is positive. Hypothesis 2: The increase of population as measured by the percentage increase of total population between 19U0 and 1950 of the SMA in 1950 increases with an increase in the value of the IED. If it is assumed that when some industries have not kept pace with other industries in an urban community a high potential growth is expected in the undeveloped industries, then an urban community with an uneven distribution of industries will expect a high population growth. If it is assumed that the Industries have kept pace with each other in an urban community with an even distribution of industries, then a low population growth Is expected In this community. Therefore, the past Increase of population in an urban community with an even distribution of Industries is not likely to be great. Null hypothesis 2: The association between values of the IED for SMA1s in 1950 and an In crease of population as measured by the percentage in crease of total population between 19l|0 and 1950 is zero. Alternative hypothesis 2: The association between values 78 of the IED for SMA* a in 1950 and an increase of population as measured by the percentage increase of total population between 19l|0 and 1950 is positive. Hypothesis 3: The median age of the SMA in 1950 decreases with an increase in the value of the IED. Manu facturing specialization is known to be associated with a somewhat high median age and is expected to have a small IED value because It would expect to bring an even distri bution of industries by counterbalancing tertiary Industry, which occupies the largest proportion In most urban com munities. Therefore, urban communities with a large IED value will tend to have a low median age and vice versa. Null hypothesis 3: The association between values of the IED for SMA*s in 1950 and the median age is zero. Alternative hypothesis The association between values of the IED for SMA*s in 1950 and the median age Is nega tive . Hypothesis U* The ratio of persons under twenty and over forty-four years of age to those between twenty and forty-four of the SMA in 1950 decreases with an in crease in the value of the IED. There is some indication that the population of an urban community specializing In manufacturing is on the average somewhat older than other types of urban community. Since manufacturing specializa tion is likely to result In a low IED value because It 79 would bring about an even distribution of industries by- counterbalancing tertiary industry which occupies a large proportion in most urban communities, it would seem reason able to expect a high ratio of persons under twenty and over forty-four years of age to those between twenty and forty-four from an urban community with a low IED value. Null hypothesis U: The association between values of the IED for SMA's in 1950 and the ratio of persons under twenty and over forty-four years of age to those between twenty and forty-four Is zero. Alternative hypothesis U: The association between values of the IED for SMA1s In 1950 and the ratio of persons under twenty and over forty-four years of age to those between twenty and forty-four is negative. Hypothesis 5: The fertility ratio of the SMA in 1950 decreases with an Increase In the value of the IED. The degree of urbanization Is in general inversely related to the fertility ratio and positively related to the pro portion of tertiary industry. As an urban community spe cializing in tertiary Industry Is expected to have a high IED value because tertiary industry oocupies the largest proportion in most urban communities, and it would be all the more accentuated In an urban community specializing In tertiary Industry bringing about very uneven distribu tion of various Industries, a low fertility ratio is 80 expected to be aaaociated with large IED values. Null hypothesis 5t The association between values of the IED for SMA*s in 1950 and the fertility ratio is zero. Alternative hypothesis 5: The association between values of the IED for SMA*s In 1950 and the fertility ratio is negative. Hypothesis 6: The ratio of married couples to the total population of the SMA in 1950 decreases with an In crease in the value of the IED. The percentage of the female labor force Is known to be associated more with tertiary industry than with any other. High specialization in tertiary Industry may result In uneven distribution of Industries and a large IED value. The larger the percent age of females working, the more unfavorable the sex ratio to female marriages. An urban community with a large IED value would therefore be expected to have a low ratio of married couples to the total population. Null hypothesis 6: The association between values of the IED for SMA's in 1950 and the ratio of married couples to the total population is zero. Alternative hypothesis 6: The asso ciation between values of the IED for SMA* s in 1950 and the ratio of married couples to the total population Is negative. Hypothesis 7: The percentage of the foreign-born aged twenty-one and over to the total population decreases 81 with an Increase In the value of the IED. Historically the foreign-born have tended to congregate In urban communi ties. It appears reasonable to assume that a large per centage of the foreign-born Is found In those urban com munities with an even distribution of industries because such communities would provide a wide range of Job oppor tunities. Thus, an urban community with a large IED value would expect a low percentage of the foreign-born aged twenty-one and over to the total population. Null hypo thesis 7: The association between values of the IED for SMA*s in 1950 and the percentage of the foreign-born aged twenty-one and over to the total population Is negative. Alternative hypothesis 7: The association between values of the IED for SMA1s in 1950 and the percentage of the foreign-born aged twenty-one and over to the total popula tion Is negative. Hypothesis 8: The percentage of female labor force to the persons fourteen years old and over of the SMA In 1950 Increases with an Increase in the value of the IED. The percentage of female workers Is known to be related positively to the development of tertiary Industry. High specialization In tertiary Industry would result In an uneven distribution of Industries and a large IED value. Therefore, an urban community with a large IED value would be expected to have a high percentage of female 82 workers to the persons fourteen years old and over. Null hypothesis 8: The association between values of the IED for SMA* a In 1950 and the percentage of female labor force to the persons fourteen years old and over Is zero. Alternative hypothesis 8: The association between values of the IED for SMA* s In 1950 and the percentage of female labor force to the persons fourteen years old and over Is pos itive. Hypothesis 9: The percentage of clerical and kindred workers to the total employed of the SMA In 1950 Increases with an Increase In the value of the IED. The percentage of clerical and kindred workers to the total employed Is related positively to tertiary industry. An urban community specializing in tertiary Industry is ex pected to have an uneven distribution of Industries and a consequent large IED value. This would suggest that large IED values are associated with a high percentage of clerical and kindred workers to the total employed. Null hypothesis 9: The association between values of the IED for SMA*s In 1950 and the percentage of clerical and kindred workers to the total employed Is zero. Alternative hypothesis 9: The association between values of the TED for SMA* s in 1950 and the percentage of clerical and kin dred workers to the total employed Is positive. Hypothesis 10; The median number of school years 83 completed for persons twenty-five years old and over of the SMA In 1950 Increases with an Increase In the value of the IED. Non-manufacturing specialization is asso ciated with a higher number of school years completed than manufacturing specialization. If the latter results in a small IED value because it counterbalances service indus try which occupies the largest proportion in most urban communities, then it is likely that a small IED is asso ciated with a low median number of school years completed for persons twenty-five years old and over. Null hypo thesis 10: The association between values of the IED for SMA1s In 1950 and the median number of school years com pleted for persons twenty-five years old and over is zero. Alternative hypothesis 10: The association between values of the IED for SMA* s in 1950 and the median number of school years completed for persons twenty-five years old and over Is positive. Hypothesis 11: The median Income of the SMA In 1950 decreases with an Increase in the value of the IED. Manufacturing specialization Is associated with a fairly high Income level and with a small IED value. If the reverse Is true, then a large IED value would be related to a low median Income. Null hypothesis 11: The asso ciation between values of the IED for SMA*s In 1950 and the median Income Is zero. Alternative hypothesis 11: &k The association between values of the IED for SMA1a in 1950 and the median income is negative. Hypothesis 12: The population potential of the SMA in 1950 decreases with an Increase in the value of the IED. Population potential Is a measure of accessibility of an urban community to the population in general. The higher the value of population potential, the higher the accessibility to labor resources and markets. An urban community with a high value of population potential is likely to develop all types of industries and to have a small IED value. Null hypothesis 12: The association be tween values of the IED for SMA*3 in 1950 and values of population potential is zero. Alternative hypothesis 12: The association between values of the IED for SMA’s in 1950 and values of population potential is negative. Hypothesis 13: The degree of specialization In manufacturing of the SMA In 1950 decreases with an In crease in the value of the IED. Urban communities In general are highly developed In tertiary Industry. High specialization in manufacturing would counterbalance this tendency and bring an even distribution of industries. Therefore, an urban community with a large IED value Is likely to have low specialization In manufacturing. Null hypothesis 13: The association between values of the IED for SMA's in 1950 and degrees of specialization In 65 manufacturing is zero. Alternative hypothesis 13: The association between values of the IED for SMA*s in 1950 and • degrees of specialization in manufacturing is negative. Hyp o the sis liq,: The value of the index of dissimi larity of the SMA in 1950 increases with an increase in the value of the IED. A high index of dissimilarity in dicates that the percentage distribution of the various industries of an urban community deviates greatly from that of the United States as a whole. It is likely that the percentage distribution of Industries of an urban community whoso industries are unevenly distributed deviates more greatly from that of the United States as a whole than another urban community whose industries are evenly distri buted. Null hypothesis II4; The association between values of the IED for SMA* s In 1950 and values of the Index of dissimilarity is zero. Alternative hypothesis 1U: The association between values of the IED for SMA1s in 1950 and values of the index of dissimilarity is positive. Hypothesis 15: The rank of class In the "hierar chical" typology of Duncan and his associates of the SMA In 1950 Increases with an Increase In the value of the IED. Since the lower the class, the more it Is associated with manufacturing specialization (except for the National Metropolises and Special Cases), and manufacturing special ization Is expected to have a small IED value; It is 86 reasonable to anticipate a small IED value from a class low In the "hierarchical" typology. Null hypothesis 15: The association between values of the IED for SMA*s In 1950 and ranks of classes in the "hierarchical" typology of Duncan and his associates is zero. Alternative hypothesis 15: The association between values of the IED for SMA*s in 1950 and ranks of classes In the "hierarchical" typology of Duncan and his associates is negative. In this chapter on methodology the variables were specified in relation to the four general referential con cepts of the ecological complex. Special emphasis was placed on the description of the economic differentiation as measured by the IED. The three types of categories used for the construction of the IED were also described. Several hypothetical distributions of industries were given to illustrate how varying degrees of economic differ entiation result in varying values of the IED. The uni verse of study was specified and the procedure for the selection of a sample described. The source of Informa tion was also given. Finally the hypotheses were formu lated and the statistical tools described. The null hypo theses were presented along with their alternatives. CHAPTER IV FINDINGS ON THE GEOGRAPHICAL AND FREQUENCY DISTRIBUTION OF THE INDEX OF ECONOMIC DIFFERENTIATION The section on findings will be treated in this and the following chapter. Geographical and frequency distributions of the IED will be dealt with here. Chapter V will analyze the findings with reference to the hypo theses. It should be remembered that the IED Is evaluated on the basis of three different groupings of industry cate gories characteristic of certain occupations: the D, C, and A categories. These three classifications of economic differentiation are applied both to the fifty-six SMA1s with populations over 300,000 and to the nineteen sample SMA*s with populations between 100,000 and 300,000 for comparison. The geographical distribution of the IBP.— There Is a clear pattern in the geographical distribution of the IED values for the fifty-six SMA*s. Except for the geo graphical distribution of the IED on the basis of the D categories for one region--the West North Central--, the New England, the Middle Atlantic, the East South Central, and the East North Central Regions are characterized by low DSD values. The South Atlantic, the West South Central, the Mountain, and the Pacific Regions are 87 88 characterized by the predominance of high TKD values for all the D, C, and A categories. The regions with low ZED values are located in the areas of high population poten tial, and the regions with high IED values are located in low population potential areas. SMA*s in the former regions tend to have an even distribution of the various industries, while SMA1s in the latter regions are likely to have an uneven distribution. The West North Central Region does not have a consistent pattern in that for the D categories it is characterized by the predominance of low IED values, but for the A and C categories it is characterized by high IED values. In the Fisher exact probability test the obtained value indicates the degree of probability of occurrence for a given combination of events. The smaller the value, the more significant it Is. The Fisher exact probability test of the D categories both with the C and A categories results in the same high probability value of .U5» which Indicates that the differences are not significant. Figures 1, 2, and 3 show the geographical distri bution of dichotomized high and low IED values for the fifty-six SMA's by Regions for the D, C, and A categories respectively. Triangles indicate those areas with high IED values and circles those with low IED values. The triangles are scattered in the western section of the 89 United States, while the circles are more concentrated in the eastern section. Table 12 shows the distribution of the numbers and the proportions of high and low IED values in each Region for the fifty-six SMA*s. All the Regions except the West North Central have consistently high proportions of either high or low IED values in all three categories of D, C, and A. In the West North Central Region Minneapolis and Omaha have high IED values for the C and A categories, but they have low IED values for the D categories. The high IED values of the SMA1s for both C and A categories are due to their high development in non-local service industries and in indus tries that may be local or non-local. However, when the service industries are broken down into smaller subcate gories for the D categories, the comparatively even dis tribution of these subcategories within the service in dustry has the effect of reducing the IED value for the D categories. Figures I 4., 5, and 6 show the geographical distri bution of the high and low IED values according to Regions for the D, C, and A categories for the nineteen sample SMAfs. A pattern similar to that found in the fifty-six SMA’s can be seen. In both cases, the Middle Atlantic and the East North Central Regions consistently have pre dominantly a higher proportion of low IED values, while FIGURE 1.— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERENTIATION BY D CATEGORIES FCR 56 STANDARD METROPOLITAN AREAS BT GEOGRAPHICAL REGIONS: 1950 A high indexes of economic differentiation o low indexes of economic differentiation Note: See table 12 vO o FIGURE .— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERENTIATION BT C CATEGORIES FOR 56 STANDARD METROPOLITAN AREAS BY GEOGRAPHICAL REGIONS: 1950 fi high indexes of economic differentiation low indexes of economic differentiation Eote: See table 12. FIGURE 3.— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERENTIATION BY A CATEGORIES FOR 56 STANDARD METROPOLITAN AREAS BY GEOGRAPHICAL REGIONS: 1950 A\ o o ° °^ Aq O A high indexes of economic differentiation O low indexes of economic differentiation Note: See table 12. TABLE 12 DISTRIBUTION OF THE NUMBERS AND FROPCRTIONS OF HICK AND LOW INDEXES OF ECONOMIC DIFFERENTIATION FOB ALL THREE CATEGORIES OF D, C, AND A BT REGIONS FOR THE 56 STANDARD METROPOLITAN AREAS: 1950 Division Categories D C A Hi^ Low High Low HiAh Low N P N P N P N P N P N p United States... 28 .50 28 .50 28 .50 28 .50 28 .50 28 .50 New Bigland....... 1 .25 3 .75 1 .25 3 .75 1 .25 3 .75 Middle Atlantic... 1 .11 8 .89 1 .11 8 .89 2 .22 7 .78 South Atlantic.... 7 .78 2 .22 7 .78 2 .22 6 .67 3 .33 East South Central.. 1 .20 4 .80 2 .40 3 .60 2 .40 3 .60 East North Central.. 6 .50 6 .50 1 .08 11 .92 1 .08 11 .92 West North Central.. 1 .25 3 .75 3 .75 1 .25 3 .75 1 .25 West South Central.. 5 .83 1 .17 6 1,00 0 .00 4 .67 2 .33 Mountain.......... 2 1.00 0 .00 2 1.00 0 .00 2 1.00 0 .00 Pacific.......... 4 .80 1 .20 5 1.00 0 .00 4 .80 1 .20 sO u> FIGURE ▲ O .— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERUITIATION BY D CATEGORIES FCR 19 SAMPLE STANDARD METROPOLITAN AREAS BY GEOGRAPHICAL REGIONS: 1950 high indexes of economic differentiation low indexes of economic differentiation Note: See table 13. sO -F" FIGURE 5.— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERENTIATION BY C CATEGORIES FOR 19 SAIIPLE STANDARD METROPOLITAN AREAS BY GEOGRAPHICAL REGIONS: 1950 i !! i , i; 11 * i i! ? i 1 f ) ! I! * !! ...j i A higfr indexes of economic differentiation O low indexes of economic differentiation Rote: See table 13. FIGURE ▲ O 6.— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERENTIATION BY A CATEGORIES PCR 19 SAMPLE STANDARD METROPOLITAN AREAS BY GEOGRAPHICAL REGIONS: 1950 \\ i i r C " - Y \ J I L. H high indexes of economic differentiation low indexes of economic differentiation Note: See table 13. TABLE 13 DISTRIBUTION OF THE NUMBERS AND PROPORTIONS OF HIGH AND LOW INDEXES OF ECONOMIC DIFFHtafTIATION FOR ALL THREE CATEGORIES OF D, C, AND A HI REGIONS FOR TOE 19 STANDARD METROPOLITAN AREAS: 1950 Division Categories D C A High .. Low High Low High Low N P N p N P N p N P N P United States... 9 .47 10 .53 10 .53 9 .47 9 .47 10 .53 New England....... 1 .33 2 .67 2 .67 1 .33 1 .33 2 .67 Middle Atlantic... 0 •00 3 1.00 1 .33 2 .67 1 .33 2 .67 South Atlantic.... 0 .00 1 1.00 0 .00 1 1.00 0 .00 1 1.00 East South Central.. 2 1.00 0 .00 2 1.00 0 .00 2 1.00 0 .00 East North Central.. 2 .40 3 .60 0 .00 5 1,00 0 .00 5 1.00 West North Central.. 2 1.00 0 .00 2 1.00 0 .00 2 1.00 0 .00 West South Central.. 2 1.00 0 .00 2 1.00 0 .00 2 1.00 0 .00 Mountain.......... 0 .00 0 .00 0 .00 0 .00 0 .00 0 .00 Pacific........... 0 .00 1 1.00 1 1.00 0 .00 1 1.00 0 .00 sO -o 98 the West South Central Region has a higher proportion of high IED values. It may be said on the basis of the nine teen sample SMA*s that, generally speaking, the eastern part of the United States are characterized by low IED values, while the western part Is dominated by high IED values. However, there Is also some dissimilarity between the regional pattern of the IED for the fifty-six SMA*s and that for the nineteen sample SMA*s. The South Atlantic Region has a higher proportion of high IED values for the fifty-six SMA*s, while it tends to have low IED values for the nineteen sample SMA1s. The opposite Is true for the East South Central Region. No explanation Is attempted for this phenomena as only one SMA is Included for the nineteen sample SMA1s. The Pacific Region has a higher proportion of high IED values for all the categories of both the fifty-six SMA's and the nineteen sample SMA1s except for the D cate gories of the nineteen sample SMA's in which the only one SMA included has a low IED value. It can be seen that the New England Region has a higher proportion of low IED values for the D and A categories, but it has a higher proportion of high IED for the C categories. This re sults from the fact that Lowell, Massachusetts, in the New England Region has low IED values for the D and A 99 categories, but it has a high IED value for the G cate gories. The Pisher exact probability test of the C cate gories both with the D and A categories results in both cases in a high probability value of .50, which indicates that the differences are not significant. Table 11+ shows the frequency distribution of the IED values by categories for both the fifty-six SMA*s and the nineteen sample SMA*s. All the distributions are positively skewed, with a larger proportion of IED values at the lower end of the range. For every category the average value for the fifty-six SMA*s is greater than that for the nineteen sample SMA*s. However, the differences are not statistically significant. For the D categories, the average IED value Is .1^3 for the fifty-six SMA*s and .132 for the nineteen sample SMA*s. For the G categories, the average IED value Is .292 for the fifty-six SMA*s and ,2Sb for the nineteen sample SMA*s. For the A categories, the mean IED value is .277 for the fifty-six SMA*s and .252 for the nineteen sample SMA*s. The t ratios of 1.22, 1.90, and 1.11+, respectively, do not reach the 5 per cent level of significance. For both the fifty-six SMA's and the nineteen sample SMA's the mean IED value for the G categories is larger than for the other categories. For the D categories the mean IED value Is the smallest. The TABLE 14 FRBCJJENCY DISTRIHJ TION OF THE INDEXES OF ECONOMIC DIFFERENTIATION BOTH OF THE 56 STANDARD METROPOLITAN AREAS AND OF THE 19 SAMPLE STANDARD METROPOLITAN AREAS BY CATEGORIES 56 Standard Metropolitan Areas 19 Standard Metropolitan Areas D Categories C Categories A Categories D Categories C Categories A Categories Intervals 1 Intervals f Intervals f Intervals t Intervals t Intervals f .295— .319 1 .542— .562 1 .526— .581 1 .189— .199 1 .413— .439 1 .408— .434 1 .270— .2% 0 .501— .541 0 .470— .525 1 .178— .188 1 •386— .412 0 .381— .407 0 .245— .269 1 .460— .500 1 .414— .469 3 .167— .177 0 .359— .385 1 .354— .380 1 . 220— .244 1 .419— .459 3 .358— 413 2 .156— .166 2 .332— .358 2 .327— .353 3 .195— .219 2 .378— .418 2 .302— 357 13 .145— .155 3 .305— .331 1 .300— .326 0 .170— .194 6 .337— .377 5 . 246— .301 13 .134— .144 1 .278— .304 0 .273— .299 0 .145— .169 11 .296— .336 10 .190— .245 15 .123— .133 2 .251— .277 2 .246— .272 3 .120— .144 17 .255— .295 10 .134— .189 5 .112— .122 3 .224^-.250 3 .219— .245 2 .095— .119 13 .214— .254 20 .078— .133 1 .101— .111 3 .197— .223 5 .192— .218 5 .070— .094 4 .173— .213 4 .022— 077 2 . 090— .100 3 .170— .196 4 .165— .191 4 M .143 .291 .277 .132 .254 .252 SD .042 .086 .099 .029 .071 .072 100 101 difference between the average IED values for the C cate gories and for the A categories Is not significant statls tioally for both the flfty-slx SMA's and the nineteen sample SMA's.1 However, the difference between the mean IED values for the C categories and for the D categories Is statistically significant for both the fifty-six SMA's and the nineteen sample SMA's.2 Furthermore, the differ ence between the average IED values for the A categories and for the D categories Is statistically significant for both the fifty-six SMA's and the nineteen sample SMA's.3 The large absolute IED values for the C and A categories may be due largely to the high proportions of workers In tertiary industry. The small absolute IED values for the D categories appear to be due to the disappearance of this pattern when tertiary Industry is broken down Into smaller categories. ^The t ratios of .78 and .09, respectively, do not reach the 5 P®** cent level of significance. The t ratios of 11.38 and 6.78, respectively, are significant at the 1 per cent level. ^The t ratios of 7.1*14- *nd 6.67, respectively, are significant at the 1 per cent level. CHAPTER V TESTS OP HYPOTHESES Each specific hypothesis will be discussed first on the basis of the findings, and then, when analysing the fifty-six SMA's, a brief explanation will be given of those which deviate greatly from the general pattern in a case where the null hypothesis is rejected. Table 15 indicates the distribution of the Pearson- lan coefficients of correlation and Rho's between the IED and a selected group of socio-economic variables by D, C, and A categories for the fifty-six SMA's and for the nine teen sample SMA's. Five of the variables show high and consistent relationships with the IED. The index of dis similarity, Increase of population, median number of school years completed, population potential, and the ratio of persons under twenty and over forty-four years of age to those between twenty and forty-four are all correlated sig nificantly with the IED values for the D, C, and A cate gories of the fifty-six SMA's. This pattern does not apply to the nineteen sample SMA's except for the population In crease variable and population potential variables. The proportion of clerical and kindred workers and the proportion of female labor force are not significantly related to the IED value on the basis of the D categories 102 103 TABLE 15 PEARSON IAN COEFFICIENTS OF CORRELATION AND RHO'S BETWEEN THE INDEX OF ECONOMIC DIFFERENTIATION AND A SELECTED GROUP OF SOCIO-ECONOMIC VARIABLES BY CATEGORIES 56 Standard Metropolitan Areas Variables Index of dissimilarity....... Increase of population....... Median school years.......... Population potential......... Ratio of under 20 and over 44 to 20-44..................... Ratio of married couples to the total population....... P. of clerical workers....... P. of female labor force..... Fertility ratio............... Median age..................... P. of the foreign-born....... Median income.................. Index of economic diff....... City size rank®............... D C A .60** .41** .36** .51** .53** .54** ,39** .39** .42** -.35** -.53** -.47** -.27* -.48** -.51** -.26* . -.05 .01 .21 .51** .61** .17 .40** .55** .14 -.02 .04 -.11 -.22 -.12 -.10 -.13 .11 -.09 -.18 -.19 ,62** .53** -.04 .09 .15 19 Sample Standard Metropolitan Areas Varlable s Index of dissimilarity....... Increase of population....... Median school yearB........... Population potential......... Ratio of under 20 and over 44 to 20-44..................... Ratio of married couples to the total population....... P. of clerical workers....... P. of female labor force..... Fertility ratio............... Median age..................... P. of the foreign-born....... Median income.................. Index of economic diff....... City size rank®............... 0 C A .25 - .27 .01 .47* .51* .54* - .03 .29 .23 - .44* - . 55* - . 56* - .44* - .45* - .37 - .08 - .44* - .42* . 36 .62** . bl** . 13 .29 .31 . 10 .09 - .24 - .42 - .53* - .54* - .32 .37 - .39* .37 - . 73** - . 70** -- .70** . 88** - .07 .03 . 16 ®Rho is used. P denotes percentage. * Significant at the 5 per cent level. ** Significant at the 1 per cent level. The levels of significance of r were determined by the use of Fisher's table. 1014- of the fifty-six SMA*s, but they show & high correlation with the IED values on the basis of the C and A categories. This nay be explained by the fact that both the proportion of clerical and kindred workers and the proportion of female labor foroe are associated with tertiary industry, which, in turn, is related to the large IED value. This relationship disappears, however, when tertiary Industry is subdivided into smaller units used with respect to the D categories. Median income is not related significantly to the IED values of all D, C, and A categories for the fifty-six SMA*s and to the IED value of the D categories for the nineteen sample SMA*s, but is significantly related to the IED values of the C and A categories for the nineteen sample SMA*s. The reason for this phenomenon is not clear. Null hypothesis 1 is that the association between values of the IED for SMA's in 1950 and the size rank is zero. Alternative hypothesis 1 is that the association between values of the IED for SMA's in 1950 and the size rank is positive. When the size-of-place distribution is taken as the standard of relative urbanization, the four major Regions may be ranked according to the degree of ur banization. The Northeast Region is the most urbanized, followed by the West Region, the North Central Region, and the South Region, in that order. In oarrying out this 105 study, no significant relationship between the IED value and size of place has been found. When both the IED values and the numbers of population of SMA's are ranked according to size, no significant relationship Is found between the two variables. The rank-dlfferenoe coefficients of corre lation for the D, C, and A categories of the fifty-six SMA's are -.0I 4., .09, and .15 respectively, and they do not reach the 5 P©** cent level of significance. Those for the respective categories of the nineteen sample SMA's are -.07, .03, and .16, none of which reaches the 5 per cent level of significance. The null hypotheses are not re jected for both the flfty-slx SMA's and the nineteen sample SMA's. With respect to the United States SMA's constitu ting the universe of study there Is no apparent relation ship between size of place and how evenly or unevenly Industries are distributed within a given SMA. Null hypothesis 2 Is that the association between values of the IED for SMA* s in 1950 and the Increase of population as measured by the percentage Increase of total population between 191+0 and 1950 is zero. Alternative hy pothesis 2 Is that the association between values of the IED for SMA's In 1950 and the Increase of population as measured by the percentage Increase of total population between 19i|0 and 1950 Is positive. The null hypothesis assumes no relationship between the IED and the degree of 106 population increase. The null hypotheses are rejected in favor of their alternatives for all D, Ct and A categories both for the flfty-slx SMA*s and for the nineteen sample SMA*s. All the correlation coefficients between the IED value and population increase are positive. For the fifty- six SMA*s the coefficients are .51, .53, and all of which are significant at the 1 per cent level. The corre lation coefficients for the nineteen sample SMA's are .i+7, .51, and .51+ for the D, C, and A categories, respectively, all of which are significant at the 5 P®r cent level. This means that SMA's with an uneven distribution of industries in 1950 are the ones that tended to have a higher rate of population Increase between 191+0 and 1950 than those with an even distribution. It Is of Interest to note that many of the SMA's with an uneven distribution of industries are In the low manufacturing category as defined by Duncan and Reiss.* This seems to be in accord with their finding that there is less manufacturing in the communities of rapid growth than in those of relative stability or decline. It may be that population Increase is more common in those SMA's where manufacturing is in the process of developing. Akron, Phoenix,and Ban Diego are deviates In the ^Otls Dudley Duncan and Albert J. Reiss, Jr., Soolal Characteristics of Urban and Rural Communities. 1950 (Hew 'KrKs""ToKf Wiley and 33ns",' Tbb.; 1$55T7 PpT"^?-^9. 107 general linear relationship that holds for the fifty-six SMA's: Akron has a high TED value and a low increase of population; Phoenix and San Diego have low IED values and high population inoreases. It may be recalled that in crease of population is not great in manufacturing metro politan areas. The fact that Akron Is one of the metro politan manufacturing areas may be one of the reasons why the population does not increase to the degree expected from the metropolitan area with a high IED value. Phoenix and San Diego have a high increase of population which is not commensurate with the size of their IED values. Phoenix Is a relatively new city attracting people from all over the United States for health reasons because of Its climate and because it has irrigation systems for the surrounding productive Salt River Valley thereby providing 2 many Job opportunities. The eoonomy of San Diego has been based largely on aircraft production and military Installa tion, both of which aocount for about 30 per cent of its wage and salary employment.3 In late years Its mild cli mate seems to be attracting people from many parts of this 2 Otis Dudley Duncan, W. Richard Scott, Stanley Lleberson, Beverly Dunoan, and Hal H. Wlnsborough, Metro- 5 oils and Region (Baltimore: The Johns Hopkins Press, 960), pp. 5S-B-550. 3Ibid.. p. 525. 108 countjry.^ Null hypothesis 3 la that the association between ▼aluea of the IED for SMA's in 1950 and the median age la aero. Alternative hypothesis 3 la that the association between values of the IED for SMA's and the median age Is negative. No significant relationship Is found between economic differentiation and fcedlan age for all categories of the flfty-slx SMA's. The null hypothesis Is not re jected. The coefficients of correlation for the D, C, and A categories are -.11, -.22, and -.12, respectively, none of which are significant at the 5 P«r cent level. A sig nificant association is found, however, for the C and A categories of the nineteen sample SMA's although no sig nificant relationship is found for the D categories. The null hypotheses are rejected in favor of the alternative hypotheses for the C and A categories, but the null hypo thesis is not rejected for the D categories. The coeffi cients of correlation for the D, C, and A categories are -.1j.2, -.53# *nd -.51*-, respectively. The last two coeffi cients of correlation are significant at the 5 P«r oent level, while the first one does not reach the 5 P®** cent level of significance. All the TED values of the nineteen sample SMA's are larger than those of the flfty-slx SMA's. ^Ibld.. p. 52k. 109 There appears to be an association between IED values and median age for smaller SMA* s which is not found for larger SMA1s. For the smaller SMA's which are represented by the nineteen sample SMA'8, an even distribution of industries is associated with high median age when the IED is con structed on the basis of three broad Industry categories. Null hypothesis h. is that the association between values of the IED for SMA's In 1950 and the ratio of per sons under twenty and over forty-four years of age to those between twenty and fomty-four is aero. Alternative hypothesis h- is that the association between values of the IED for SMA's in 1950 and the ratio of persons under twenty and over forty-four years of age to those between twenty and forty-four is negative. The null hypotheses of no relationship between dependency and economic differen tiation are rejected in favor of the alternative hypotheses for all D, C, and A categories of the fifty-six SMA's and for the D and C categories of the nineteen sample SMA's, while the null hypothesis Is not rejected for the A cate gories of the nineteen sample SMA's. The coefficients of correlation of the fifty-six SMA's for the D, C, and A categories are -.27, and “»5l, respectively. The first coefficient of correlation is significant at the 5 per cent level, while the last two correlation coeffi cients are significant at the 1 per cent level. The 110 coefficients of correlation of the nineteen sample SMA’s for the D, C, and A categories are -.I4J4., -.hS* and ”«37, respectively. The first two coefficients of correlation are significant at the 5 pax* cent level, while the last one fails to reaoh the significance level of 5 P®** cent. Both the fifty-six SMA’s and the nineteen sample SMA’s with an uneven distribution of industries tend to have a higher proportion of people in the prime of life, while the reverse is true with those with an uneven distribution of Industries. Akron is a deviate from the general pattern found in this study. It has a relatively high ratio of persons under twenty and over forty-four years of age to those between twenty and forty-four for the D categories. While Akron has a low IED value on the basis of the three broad categories used for the C and A categories, Its IED value Is the highest when calculated on the basis of the D cate gories. When calculated on the basis of the C and A cate gories, both IED values come down to the fortieth In the group of fifty-six. This seems to indicate that the ratio of persons under twenty and over forty-four years of age to those between twenty and forty-four Is associated more with how evenly the three broad Industries--primary, secondary, and tertiary— are distributed than with how evenly the twelve smaller subcategories are distributed. Ill Nall hypothesis 5 is that the association between values of the IED for SMA1s in 1950 and the fertility ratio is zero. Alternative hypothesis 5 is that the association between values of the IED for SMA*s In 1950 and the fer tility ratio is negative. According to the findings of this study the null hypothesis of no relationship between economic differentiation and the fertility ratio is not rejected. The coefficients of correlation for the D, C, and A categories of the fifty-six SMA's are .lk, .02, and .OI4., respectively, while those for the D, C, and A cate gories of the nineteen sample SMA's are .10, .09, and -.2L|_. None of these correlation coefficients reaches the signi ficance level of 5 Per cent. Whether or not the various industries are evenly or unevenly distributed is not re lated to the fertility ratio. This is understandable from the fact that the fertility ratio has been found to be related to size of place, but not the IED. The latter will be discussed more in detail later. Null hypothesis 6 is that the association between values of the IED for SMA's In 1950 and the ratio of married couples to the total population Is zero. Alter native hypothesis 6 Is that the association between values of the IED for SMA* s in 1950 and the ratio of married couples to the total population Is negative. The coeffi cients of correlation for the three categories of D, C, 112 and A of the fifty-8lx SMA*a are -.26, -.05, and .01, respectively. The first coefficient of correlation is significant at the 5 P®r cent level, while the last two do not reach the significance level of 5 per cent. The co efficients of correlation for the D, C, and A categories of the nineteen sample SMA* a are -.06, -.I4 J4 ., and -.l|2, respectively. The first coefficient of correlation does not reaoh the significance level of 5 par cent, but the last two are significant at the 5 P®r cent level. The null hypotheses are not rejected for the C and A categories of the fifty-six SMA*s and the D categories of the nineteen sample SMA*s, but they are rejected for the D categories of the fifty-six SMA*s and for the C and A categories of the nineteen sample SMA*s. When the three broad categories of primary, second ary, and tertiary are used as the basis for the construc tion of the IED no significant relationship is found be tween the economlo differentiation as measured by the IED and the ratio of married couples to the total population for the C and A categories of the fifty-six SMA*s. How ever, a significant relationship is found between these two variables for the C and A categories of the nineteen sample SMA*a. When twelve broader categories are used for the construction of the IED a significant relationship is found between these two variables for the D categories of 113 the fifty-six SMA's, but no relationship is found for the D categories of the nineteen sample SMA's. When the D categories are used an even distribution of Industries tends to be aocompanled by a high ratio of married couples to the total population for the fifty-six SMA's, whereas this relationship does not hold for the nineteen sample SMA's. When the C and A categories are used the above relationship holds for the nineteen sample SMA's, but not for the flfty-slx SMA's. The reason for this situation is not dear. Null hypothesis 7 Is that the association between values of the IED for SMA* s in 1950 and the percentage of the foreign-born aged twenty-one and over to the total population is zero. Alternative hypothesis 7 Is that the association between values of the IED for SMA's In 1950 and the percentage of the foreign-born aged twenty-one and over to the total population is negative. The null hypo thesis of no relationship between the IED and the percent age of the foreign-born Is not rejected for all categories both of the fifty-six SMA's and for the D and C categories of the nineteen sample SMA's, while it is rejected for the A categories of the nineteen sample SMA's. The coeffi cients of correlation for the D, C, and A categories of the fifty-six SMA's are -.10, -.13# *nd .11, respectively. They do not reach the significance level of 5 per cent. 1114. The coefficients of correlation for the D, C, and A cate gories of the nineteen sample SMA's are -.32, -.37, and -.39, respectively. The first two correlation coefficients do not reach the significance level of 5 P©r oent, although they approach the size of the last one whioh Is significant at the 5 pax* cent level. The evenness or unevenness of the distribution of industries within a SMA may be said on the whole not to be significantly associated with the percentage of the foreign-born in that area. A high degree of association between size of place and the percentage of the foreign-born and no significant relationship between size of place and the IED are in accord with this finding. It appears that the foreign-born cluster in large cities of almost any type of functional specialization. Null hypothesis 8 is that the association between values of the IED for SMA* s in 1950 and the percentage of female labor force to the persons fourteen years old and over is zero. Alternative hypothesis 8 is that the asso ciation between values of the IED for SMA's In 1950 and the percentage of female labor force to the persons four teen years old and over is positive. The null hypotheses of no relationship between the IED and the percentage of female workers to the total labor force are rejected in favor of the alternative hypotheses for the C and A categories of the fifty-six SMA's, while the null 115 hypothesis Is not rejected for the D categories of the fifty-six SMA*s. The coefficients of correlation for the D, C, and A categories of the fifty-six SMA*s are .17* -U-O* and .55* respectively. The first ooefficlent of correla tion does not reach the significance level of 5 per cent, while the last two coefficients of correlation are signi ficant at the 1 per cent level. The coefficients of corre lation for the D, C, and A categories of the nineteen sample SMA1s are .13* *29, and .31, respectively. None of these obtain the significance level of 5 per oent. When three broad industry categories of primary, secondary, and tertiary are used, an uneven distribution among them is associated with a high proportion of female workers. This relationship does not occur when the twelve smaller subcategories of industries are used. The sizes and direction of the coefficients of correlation for the nineteen sample SMA*s are In the same pattern as that found for the fifty-six SMA*s except for the fact that they do not reach the significance level of 5 P®** cent. Some explanation is necessary for the lack of significant relationship for the D categories of the fifty- six SMA*s, while there is a significant relationship for the C and A categories. The finding by Duncan and Reiss that the percentages of women in the labor force are greater in trade centers than nontraie centers gives a 116 clue to thla phenomenon. There seems to be an association between the high IED values for the C and A categories of the fIfty-six SMA* a and trade centers and the low IED values with nontrade oenters on the basis of very crude examination. Chicago, Minneapolis-St. Paul, Dallas, Denver, Indianapolis, Kansas City, Memphis, and Omaha are trade centers, and their IED values are higher than those of Youngstown, Allentown-Bethlehem-Easton, Wheeling- Steubenvllle, and Wllkes-Barre-Hazleton which are non trade centers. Slnoe those metropolitan areas which specialize in trade seem likely to have high IED values, the significant positive association between the IED and the female labor force may be due to the closeness of the association of female workers with trade centers. Appar ently this relationship breaks down when the twelve smaller subcategories are used for the construction of the IED. San Antonio and New Orleans deviate from the ex pected pattern in that the female labor force in these SMA's is very low as compared with the size of their IED values for the C and A categories. The explanation for San Antonio is that this SMA Is a military base with a large proportion of males.^ In addition. It is a nontrade center, wholesale which is also characterized by a ^Ibid.. p. 526. 117 comparatively large proportion of male workers.^ New Orleans la essentially a wholesale center, and also a center of water transportation, whloh does not use many female workers.7 Null hypothesis 9 is that the association between values of the IED for SMA* s in 1950 and the percentage of clerical and kindred workers to the total employed is zero. Alternative hypothesis 9 is that the association between values of the IED for SMA* s in 1950 and the percentage of clerical and kindred workers to the total employed is positive. The null hypotheses of no relationship between the IED and the percentage of clerical and kindred workers to the total employed are rejected in favor of the alter native hypotheses for the C and A categories both of the fifty-six SMA's and of the nineteen sample SMA's. The co efficients of correlation are .21, .51, and .61 for the D, C, and A categories of the fifty-six SMA's, respectively. The first coefficient of correlation does not reaoh the significance level of 5 par oent, while the last two co efficients of correlation are significant at the 1 per cent level. The coefficients of correlation for the D, C, and A categories of the nineteen sample SMA's are .36, .62, £ Duncan and Reiss, op. clt.. p. 388 7Ibid. 118 and .61, respectively. The first coefficient of correla tion does not obtain the significance level of 5 per oent, while the last two coefficients of correlation are signi ficant at the 1 per oent level. It is of interest to note that the coefficients of correlation for the D categories both of the fifty-six SMA*s and of the nineteen sample SMA's are not significant, while the coefficients of correlation for the C and A categories are significant at the 1 per cent level for both groups of SMA's. This means that when the Industries are grouped together in three broad categories of primary, secondary, and tertiary, the SMA's with an uneven distri bution of industries are likely to have a higher propor tion of clerloal and kindred workers than those with an even distribution of industries. This relationship dis appears when the IED is constructed on the basis of a larger number of smaller industry categories. The reason for this phenomenon may be that clerical and kindred workers are associated more with tertiary than secondary industry, and that those SMA's specializing in tertiary industry have higher IED values than those specializing in secondary industry when the IED is constructed on the basis of three broad categories of primary, secondary, and tertiary because of the lack of a high level of seoondary industry to counteract the high level of tertiary industry. 119 However, when the twelve smaller oategorles (the D cate gories) are used for the constructIon of the IED, those SMA1s specialising In secondary Industry may obtain high IED values because of the hbsenoe of high levels of ter tiary Industry categories to counterbalance secondary industry, tertiary Industry having been broken down Into several smaller subcategories. Miami Is a deviate from the general pattern found for the fifty-six SMA's with respeot to the relationship between economic differentiation and the percentage of clerical and kindred workers to the total employed. It has a high IED value but has a low percentage of clerical and kindred workers. This phenomenon Is probably due to the fact that Miami la a resort town specializing In hotels and lodging places and entertainment and recreation services In which clerical and kindred workers are not as essential as, for example, In centers of public adrainistra- O tion. Miami is also an air transportation center, which does not require many clerical and kindred workers.^ These factors undoubtedly result in the unexpected low percentage of clerical and kindred workers considering the high IED value. ®Duncan et. al.. op. clt.. p. 531. 9Ibid.. p. 530. 120 Null hypothecla 10 Is that the association between values of the IED for SMA's In 1950 and the median number of school years completed for persons twenty-five years old and over Is zero. Alternative hypothesis 10 Is that the association between values of the IED for SMA's In 1950 and the median number of school years completed for persona twenty-five years old and over Is positive. The null hypotheses are rejected in favor of the alternative hypotheses for the fifty-six SMA's. The coefficients of correlation between economic differentiation and.median number of school years completed are .39, .39, and ,1+2 for the D, C, and A categories, respectively. They are significant at the 1 per cent level. This relationship is not found for the nineteen sample SMA's. The coeffi cients of correlation between economic differentiation and median number of school years completed are -.03, *29, and .23 for the D, C, and A categories, respectively, and do not reach the significance level of 5 par cent. The null hypotheses are not rejected for these SMA's. For the flfty-slx SMA's the finding means that those SMA's with an uneven distribution of Industries tend to have a popula tion with a higher median number of sohool years completed than those SMA's with a more even distribution of indus tries. Differences In occupational composition which are related to an even or uneven distribution of Industries 121 may partly explain the difference. The finding by Duncan and Reiss mentioned that the educational level In growing places is much higher than in declining places, which is consistent with the findings of this study that an uneven distribution of industries is associated with high level of educational attainment.10 New Orleans and San Antonio among the fifty-six SMA*s deviate from the linear regression pattern in that the median number of school years completed by their in habitants is lower than expected from their IED values. Areas specializing in wholesale, transportation, and military establishments have an educational achievement lower than the United States median.11 New Orleans is a wholesale trade center which is highly specialized in 1 P transportation. San Antonio is a military base whose specialization is nontrade center, w h o l e s a l e . ^ These factors account for their unexpected relatively low edu cational achievement considering their high IED values. Null hypothesis 11 is that the association between values of the IED for SMA's in 1950 and the median income 10Duncan and Reiss, op. cit.. p. 207. 11Ibid., p. 329. ^Duncan et. al. . op. cit. . pp. 389-391. ■^Duncan and Reiss, op. c it. . p. 388. 122 Is zero. Alternative hypothesis 11 Is that the association between values of the IED for SMA*a in 1950 *nd the median Income la negative. The null hypotheses for all D, C, and A categories of the fifty-six SMA*s are not rejected. The coefficients of correlation for the respective categories are -.09, -.18, and -.19, respectively, none of which reaches the significance level of 5 per cent. Evenness or unevenness of industry distribution is not related to income level. The picture is quite different with the nineteen sample SMA*s. The null hypothesis is not rejected for the D categories, but the null hypotheses are rejected in favor of the alternative hypotheses for the C and A cate gories. The coefficients of correlation for the respective categories are -.37, “»73, and -.70. The first coefficient of correlation does not reach the significance level of 5 per cent, but the last two coefficients of correlation are significant at the 1 per cent level. Some explanation is necessary for these differences. Why are the relation ships for the C and A categories of the nineteen sample SMA*s significant, while those of all categories of the fifty-six SMA*s and of the D categories of the nineteen sample SMA*s are not significant? An examination of types of functional specialization of the nineteen sample SMA*s reveals that six out of nineteen are highly specialized in 123 manufacturing, while two show very little suoh specializa tion.^ One of the reasons for the differences, therefore, could be that SMA's baaed on manufacturing are overrepre sented in the nineteen sanqple SMA's. It has been already mentioned that manufacturing la associated with SMA's of high Income. Another Inter pretation is that manufacturing plays a different role in large and In small SMA's. As shown in Table 16, this Interpretation has some support in the findings by Duncan and Reiss-*-^ They divided SMA's Into high and low income categories and compared these two groups In relation to the proportion of workers in manufacturing and the per capita value added by manufacturing. Places with a high income are defined as those where the median Income Is in the upper quintile for all places of a particular metro politan status and size. Places with a low income are defined as those where the median lnoome is in the lower quintile of the distribution. Table 16 indicates that for smaller SMA's the difference between high and low income SMA's in percentage of employed persons in manufacturing and in per capita value added by manufacturing is greater than that found between high and low income SMA's of ^Ibid., pp. 388-390 l5Ibid., p. 355. 121+ larger size. For the nineteen sample SMA*a an evenness of distribution of Industries la associated with high median Income when three broad Industry categories of primary, secondary, and tertiary are used as the basis of construct ing the IED. TABLE 16 EMPLOYMENT IN MANUFACTURING AND PER CAPITA VALUE ADDED BY MANUFACTURES FOR SELECTED STANDARD METROPOLITAN AREAS BY METROPOLITAN STATUS, SIZE, AND 191+9 INCOME LEVEL Metropolitan Status and Size of Place 191+9 Income Level Workers In Manufactur ing as Per Cent of All Employed Persons 1950 Per Capita Value Added by Manu factures 191+7 (dollars) Standard Metro High 1+2.1* 888* politan Areas: 50,000-250,000 Low 15.1+ 255 250,000 or more High 32.6* 781+* Low 20.9 21+3 *Denotea difference between high and low groups significant at 0.05 level of probability. Null hypothesis 12 Is that the association between values of the IED for SMA*a In 1950 and values of popula tion potential is zero. Alternative hypothesis 12 is that the association between values of the IED for SMA* s In 1950 and values of population potential la negative. The 125 null hypotheses are rejected in favor of the alternative hypotheses for all D, C, and A categories for the fifty-six SMA's. The coefficients of correlation are -.35, -.53, and — .1+7» respectively. They are significant at the 1 per cent level. In other words, the IED is inversely related to the level of population potential. Those SMA's with an even distribution of industries are likely to be located in areas of high population potential. The null hypotheses are also rejected In favor of the alternative hypotheses for the D, C, and A categories of the nineteen sample SMA's. The coefficients of correlation are -.55, and -.56 for the D, C, and A categories, respectively. All three coefficients of correlation are significant at the 5 per cent level. The small size of coefficient of correlation for the D categories both for the fifty-six SMA's and for the nineteen sample SMA's may be due to the fact that population potential is more related to large overall aggregates such as primary, secondary, and tertiary Industry rather than to smaller segments within each one of these Industries. Figures 7, 3, and 9 show how the high and low IED values are distributed in relation to levels of population potential for the fifty-six SMA's for the D, C, and A categories, respectively. Figures 10, 11, and 12 show the same for the nineteen sample SMA's. Circles (o) denote 4 o FIGURE 7.— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERENTIATION BY D CATEGORIES FCR 56 STANDARD IISTKOPOLITAN AREAS BY LEVELS OF POPULATION POTENTIAL: 1950 high indexes of economic differentiation low indexes of economic differentiation Note: See table 17. FIGURE 8.— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERENTIATION BY C CATEGORIES FCR 56 STANDARD METROPOLITAN AREAS BY LEVELS OF POPULATION POTENTIAL: 1950 /T o l! \ L, \ j ! It A high indexes of economic differentiation O low indexes of economic differentiation Note: See table 17, FIGURE 9.— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERENTIATION BY A CATEGORIES FCR 56 STANDARD METROPOLITAN AREAS BY LEVELS OF POPULATION POTENTIAL: 1950 I I ! Zto 1 I I: !! 00 ! O O o AA Jfo l A high indexes of economic differentiation 0 low indexes of economic differentiation Note: See table 17. TABLE 17 DISTRIBUTION OF THE NUMBERS AND PROPORTIONS OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERENTIATION FCR ALL THRBS CATEGORIES OF D, C, AND A BI LEVELS OF PORJLATION POTENTIAL FCR THE 56 STANDARD METROPOLITAN AREAS: 1950 ______________________________ Categories_________________________________ Levels of Population _________ D________________________C_________________________ A__________ Potential High_______Low___________ High_______Low____________ High_______Low ______________________N P N P_______ N P M P________N P N P Over 500.......... 0 .00 6 1.00 1 .17 5 ,83 1 .17 5 .83 450— 500.......... 3 . 30 7 .70 1 .10 9 .90 1 .10 9 .90 400— 450.......... 5 .45 6 .55 2 .18 9 .82 3 #27 8 .73 350—400 3 .50 3 .50 3 .50 3 .50 3 .50 3 .50 300—350 2 .50 2 .50 2 .50 2 .50 2 .50 2 .50 250—300 1 .33 2 .67 3 1.00 0 . 00 3 1.00 0 .00 200—250 .......... 5 . 83 1 .17 6 1.00 0 .00 4 .67 2 .33 150— 200.......... 7 1.00 0 .00 7 1.00 0 .00 6 .86 1 .14 Less than 150 2 .67 1 .33 3 1.00 0 .00 2 .67 1 .33 129 FIGURE 10.— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFERENTIATION BY D CATEGORIES FCR 19 SAMPLE STANDARD METROPOLITAN AREAS BY LEVELS OF POHJLhTION POTENTIAL: 1950 A high indexes of economic differentiation O low indexes of economic differentiation Note: See table 18. 130 o ► FIGURE 11.— DISTRIBUTION OF HIGH aND LOW INDEXES OF ECONOMIC DIFFERENTIATION BY C CATEGORIES EXE 19 SAMPLE STANDARD METROPOLITAN AREAS BY LEVELS OF POPULATION POTENTIAL: 1950 I I I ' i i i ! i I II i I I ! V \ ffO /fo / / “s. 1°° X V * * 1*° S t-—- ".IX.-" " | / / / i f ^ o o V - . 2-00 I ^ 2S~° - ' 1 3»o y s') L'-T* ' i f_-— \L+oo-tz y < jrvPr ^fo / ./ / ,4 f-ZEP't* ? ! i ' /' V \a / I a / i K < : ° ‘ /- \ /ro \ / 'J floo n I ! • • t I I I I i w » A 1 \ . i i high indexes of economic differentiation low indexes of economic differentiation Note: See table 18. FIGURE 12.— DISTRIBUTION OF HIGH AND LOW INDEXES OF ECONOMIC DIFFGRENTIATION BY A CATEGORIES FOR 19 SAMPLE STANDARD METROPOLITAN AREAS BY LEVELS OF POPULATION POTENTIAL: 1950 JOO I * ! ' i i 11 4 t t ‘ i ! | /. / S ~ o J 2 . S-O 2. OO * \ Z To ▲ high indexes of economic differentiation O low indexes of economic differentiation Note: See table 18. TABLE 18 DISTRIBUTION OK THE NUMB HIS AND PROPORTIONS OF HIGH AND LOW INDEXES OF ECONOMIC DIFPHIENTLATION FOR ALL THREE CATEGORIES OF D, C, AND A BT LEV5LS OF PORJLATION POTTOTIAL FOR THE 19 STANDARD METROPOLITAN AREAS: 1950 ______________________________Categories________________________________ Levels of Population _________ D C_________________________A_________ Potential High_______Low___________ Hlrti_______Low____________High_______Low _____________________ N P N P_______N P N P________N P N P Over 500 1 .33 2 .67 2 .67 1 .33 2 .67 1 .33 450— 500 0 . 00 2 1.00 0 . 00 2 1.00 0 .00 2 1.00 400— 450.......... 1 .33 2 . 67 0 . 00 3 1.00 0 . 00 3 1.00 350— 400.......... 1 .25 3 .75 0 . 00 4 1.00 0 .00 4 1.00 300— 350 0 .00 0 .00 0 . 00 0 . 00 0 . 00 0 .00 250— 300 3 1.00 0 . 00 3 1.00 0 .00 3 1.00 0 .00 200— 250 3 1.00 0 .00 3 1.00 0 .00 3 1.00 0 .00 150— 200 0 . 00 0 .00 0 .00 0 .00 0 . 00 0 .00 Less than 150 ..... 0 . 00 1 1.00 1 1.00 0 . 00 0 . 00 1 1.00 131* low IED values, while triangles (A) denote high IED values. Generally speaking, high IED values are In the areas of high population potential and low IED values are located In the areas of low population potential. Tables 17 and 18 show the numbers and proportions of high and low IED values by different levels of population potential. Within each category the proportion of high IED values tends to increase with a decrease In the level of population potential. The finding by Duncan and his associates that SMA's specializing in manufacturing tend to be found in areas of high population potential,^ and the finding of this study that SMA's with an even distribution of indus tries are likely to be located In areas of high population potential are consistent with another finding that the metropolitan areas with an even distribution of industries In this study have high manufacturing activity as defined by Duncan and ReIss.^ Washington, D.C., Is a deviate within the general pattern of the inverse relationship between the IED and population potential that holds for the fifty-six SMA's. This SMA is high both in the IED value and In the level of ■^Duncan et. al. , op. cit.. p. lj.2. ^Duncan and Reiss, op. cit.. pp. 388-390. 135 population potential. The reason for this deviation is obvious. It is the capital of the country and with its specialization in federal administration, employment is not market oriented. More than 31 per cent of its popu lation is employed in federal public administration and 1 f t printing, publishing, and allied industries. Null hypothesis 13 is that the association between values of the IED for SMA's in 1950 and degrees of special ization in manufacturing is zero. Alternative hypothesis 13 is that the association between values of the IED for SMA’s in 1950 and degrees of specialization in manufactur ing is negative. It has already been mentioned that manu facturing specialization is associated with socio-economic factors such as level of income and degree of labor force participation. It is important to find out whether or not evenness or unevenness of the distribution of industries Is correlated with one of the major industry groups such as manufacturing. If any relationship is found, It will further illuminate the relationship of the IED to other socio-economic factors and Increase its practical useful ness. In testing the null hypothesis the classification of high and low manufacturing SMA's by Duncan and Reiss was used. They employ the quintile limits for percentage “ 1 ft Duncan et. al.. op. cit.. p. 519. 136 TABLE 19 N U M B E R S A N D PE R C E N T A G E S O P H IG H A N D L O W M A N U F A C T U R IN G S T A N D A R D M E T R O PO L IT A N A R E A S IN R ELA TIO N T O H IG H A N D L O W IN D EX ES O F E C O N O M IC DIFFERENTIATION F O R T H E D C A TEG O R IES O F T H E 56 S T A N D A R D M E T R O PO L IT A N A R E A S Index of Manufacturing S pecialization Economic D ifferen tiatio n High L ow Total Per Per Per No. Cent No. Cent No. Cent High 5 k$ 11 100 16 73 Low 6 55 0 0 6 2? Total 11 100 11 100 22 100 T A B L E 20 N U M B E R S A N D P E R C E N T A G E S O F H IG H A N D L O W M A N U F A C T U R IN G S T A N D A R D M E T R O P O L IT A N A R E A S IN R E LA T IO N T O H IG H A N D L O W IN D EX E S O F E C O N O M IC DIFFERENTIATION F O R T H E C C A T E G O R IE S O F T H E 56 S T A N D A R D M E T R O P O L IT A N A R E A S Index of Manufacturing Specialization Economic Differentiation High _ Low Total Per No. Cent No Per Cent Per No. Cent High 0 0 11 100 li 50 Low 11 100 0 0 n 50 Total 11 100 11 100 22 100 137 TABLE 21 NUMBERS AND PERCENT AGES OP HIGH AND LOW MANUFACTURING STANDARD METROPOLITAN AREAS IN RELATION TO HIGH AND LOW INDEXES OP ECONOMIC DIFFERENTIATION FOR THE A CATEGORIES OF THE 56 STANDARD METROPOLITAN AREAS Index of Economic Differentiation Manufacturing Specialization High Low Total No. Per Cent No Per Cent Per No. Cent High 1 9 10 91 li 50 Low 10 91 1 9 n 5o Total 11 100 11 100 22 100 in manufacturing specialization. SMA's in the upper quintile are classified as high in manufacturing speciali zation, while those in the lower quintile are classified as low In manufacturing s p e c i a l i z a t i o n . ^ Metropolitan areas are classified according to high and low IED values as divided at the median. Tables 19, 20, and 21 show the numbers and percentages of high and low manufacturing SMA's In relation to high and low IED values. Twenty-two out of the fifty-six SMA's are classified either as high or as low manufacturing SMA's. As can be seen from the tables, for all three categories of D, C, and A, a larger ^Duncan and Reiss, op. cit. . p. 253. 138 proportion of low manufacturing SMA*s are In the high category of the IED, while a larger proportion of high manufacturing SMA1s are In the low category of the IED. The chi squares for the respective categories of D, C, and A are 7.2\±, 22.00, and Hj.,72. All of these are significant at better than 1 per cent, which Indicates that the ob served frequencies deviate significantly from the expected frequencies. Contingency coefficients are used to examine the degree of association between the IED and manufacturing specialization. The contingency coefficients for the re spective categories of D, C, and A are .U9, .71, and .63. For a 2 x 2 table as in this case a "perfect" correlation yields C = 0.71. The correlation between the IED and manufacturing for the C categories Is Mperfeot.l f The correlations for the other two categories are substantial considering the maximum value of contingency coefficient obtainable from a 2 by 2 table. The null hypotheses are rejected in favor of the alternative hypotheses for all categories of the fifty-six SMA*s. It may be concluded, therefore, that those SMA*s with low IED values are associated with high specialization in manufacturing. Or, In other words, those SMA*s highly specializing in manufacturing are likely to have an even distribution of Industries, while those very little specializing In manufacturing tend to have an uneven 139 distribution of industries. Null hypothesis lit is that the association between values of the IED for SMA1s in 1950 and values of the index of dissimilarity is zero. Alternative hypothesis lit is that the association between values of the IED for SMA1s in 1950 and values of the Index of dissimilarity la positive. The null hypotheses are rejected in favor of the alterna tive hypotheses for all categories of the fifty-six SMA*a. The null hypotheses are not rejected for all categories of the nineteen sample SMA*s. The correlation coefficients are .60, .L|.l, and .36, respectively, for the D, C, and A categories of the fifty-six SMA1s. All these are signi ficant at the 1 per cent level. The comparatively large coefficient of correlation of .60 for the D categories may be due to the use of the same categories for the construc tion both of the index of dissimilarity and of the IED. This finding does not indicate that the IED measures com pletely different phenomena from the index of dissimilar ity. It means that the SMA*s which deviate from the national average in the distribution of people in the various Industry categories are likely to have an uneven distribution of these categories. Akron, Youngstown, Allentown-Bethlehem-Easton, Wilkes-Barre-Hazleton, and Detroit are deviates from the general pattern found for the fifty-six SMA1s. The 1 1 4 . 0 comparatively low IED value of Akron for the C and A cate gories in relation to its index of dissimilarity Is due to its large proportion of workers in secondary Industry when the 0 and A categories are used. It is important to remember that either a high IED value or a low IED value may represent a type of specialisation in the sense of deviation from the national average. For Instance, if the national distribution in the three categories Is 5 per cent, 35 per cent, and 60 per cent, respectively, an SMA whose distribution is 1 per cent, 53 per cent, and I 4.6 per cent has a lower IED value than the national average be cause of the more even distribution among the three cate gories, but it is highly specialized in the second category. The index of dissimilarity of Youngstown is high because of Its specialization in blast furnaces, steel works, and rolling mills, other primary iron and steel industries, fabricated metal industries, electrical machinery, equipment and supplies, railroad and miscellan eous transportation equipment, pottery and related pro ducts, and leather.20 The high proportions of workers in these categories contribute to the even distribution of all the industry categories, which result in comparatively ^°Duncan et. al.. op. clt. . pp. I 4 . 75-I4 . 78. 114-1 low IED values for all categories of D, C, and A. The high IED value of Allentown-Bethlehem-Easton Is due to Its specialization in the categories of second stage resource users for production for non-final market, second stage resource users for production for final market, and re sources of indirect significance for production for non final market.2* Detroit*s high index of dissimilarity Is due to specialization In the category of resources of indirect significance for production for non-final market.22 The high Index of dissimilarity of Wilkes-Barre-Hazleton Is the result of its specialization In the categories of primary resource extractors for production for non-final market, first stage resource users for production for final market, second stage resource users for production for non-final market, and second stage resource users for production for final market.The specialization In the above categories also help produce low Indexes of economic differentiation by contributing to the even distribution of the various industry categories. All the SMA*s with a high index of dissimilarity and a low IED value are highly specialized ^Duncan et. al.. op . clt.. pp. i 4 . 90- Ij .9i 4 . . 22Ibid.. p. 213. 23Ibld.. pp. 533-537. In secondary industry. Null hypothesis 15 is that the association between values of the IED for SMA* s in 1950 and ranks of classes In the "hierarchical" typology of Duncan and his associates is zero. Alternative hypothesis 15 is that the association between values of the IED for SMA*s In 1950 and ranks of classes In the "hierarchical" typology of Duncan and his associates is negative. This null hypothesis aims to test the relationship between the IED and the typology of SMA* s by Duncan and his associates.^ They classified the fifty- six SMA*s into seven classes on the basis of metropolitan functions and regional relationships. The seven classes are the following: (1) National Metropolis, (2) Regional Metropolis, (3) Regional Capital, Submetropolitan, (I4 .) Diversified Manufacturing Center with Metropolitan Func tions, (5) Diversified Manufacturing Center with Few Metropolitan Functions, (6) Specialized Manufacturing Center, and (7) Special Case. The major criteria used In this classification are population size, per capita levels of manufacturing and trade, indicators of "metropolitan" functions, and character of industry profiles and regional relationships. The following is a very simplified explana tion of eaoh class. ^ Ibld. . pp. 259-279. 1. National Metropolis. a) The five largest SMA* a. b) Have greater conqalexities than prepared to handle. 2. Regional Metropolis. a) Highly specialized in Industries representing "metropolitan” functions. b) Trade functions are moderately to highly developed. 3* Regional Capital, Submetropolitan. a) The same as Regional Metropolis except for smaller size and lower per capita wholesale sales. i;. Diversified Manufacturing Center with Metropolitan Functions. a) A moderate to high emphasis on manufacturing. b) A considerable range and strength of metropolitan functions (high levels of trade and financial activities). c) Larger population size than the class 5. 5. Diversified Manufacturing Center with Few Metropolitan Functions. a) High levels of manufacturing activity (per capita value added by manufacturing). b) Not exceptionally high on per capita wholesale trade (though higher than the class 6). c) Do not have a full repertoire of "metropolitan" functions. 6. Specialized Manufacturing Center. a) Moderately high to quite high per capita value added by manufacture. b) Quite low per capita wholesale sales. 7. Special Case. a) Centers of functions other than trade or manufac turing. b) Quite low levels of both per capita wholesale sales and per capita value added by manufacture. c) Heterogeneous among themselves. Though not entirely based upon industry special ization, this typology takes into account many economic indicators. It Is worthwhile to Investigate interclass differences with regard to the IED. The null hypotheses 11* are partially rejected in favor of the alternative hypo theses. Tables 22, 23, and 2k show comparison of the IED values for the subclasses of the typology for the respec tive categories mt D, C, and A. For the D categories, the number of significant interclass differences is not as many as that either for the C categories or for the A cate gories. For the D categories, the average IED value for the Regional Metropolises is significantly larger than that either for the Diversified Manufacturing Centers with Metropolitan Functions or for the Diversified Manufacturing Centers with Few Metropolitan Functions, indicating that the latter two classes have a more even distribution of industries. The Regional Capitals, Submetropolitan have a sig nificantly higher average IED value than the Diversified Manufacturing Centers with Metropolitan Functions. The average IED value for the Special Cases is also larger than that for the Diversified Manufacturing Centers with Metro politan Functions. For the C categories, the National Metropolises have a significantly higher average IED value than the Specialized Manufacturing Centers. The average IED value for the Regional Metropolises Is significantly larger than that for the Regional Capitals, Submetropoli tan, for the National Metropolises, for the Diversified Manufacturing Centers with Metropolitan Functions, for the TABLE 22 COMPARISON OF THE SIZES OF THE INDEXES OF ECONOMIC DIFFERENTIATION AMONG THE VARIOUS TTFES OF THE 56 STANDARD METROPOLITAN AREAS ON THE BASIS OF METROPOLITAN FUNCTIONS AND REGIONAL RELATIONSHIPS AND t RATIOS PCB THE D CATEGORIES N.Metropolis .142 R.Metropolis .151 R.Capital,Sub. .147 D.Mamf.Mf. .116 D.Mamf.f.Mf. .126 Special.Mamf. .132 R,Metropolis .151(t=.60) R.Capital,Sub. D.Manuf.Mf. .147(t=.31) .ll6(t=1.51) D.Mamf.f.Mf. Special.Mamf. Special Cas. ,126(t=.89) .132(t=.30) .172(t=1.07) R.Capital,Sub. D.Mamf.Mf. D.Manuf.f.Mf. Special.Mamf. Special Cas. .147(t=.40) .Il6(t-4.3S»*) .126(t=2.78*) ,132(t*.79) .172(t-1.47) D.Mamf.Mf. .Il6(t=2.58*) D.Manuf.f.Mf. Special.Mamf. Special Cas. .126(t=1.75) .132(t=.75) .172(t=1.47) D.Mamf.f.Mf. Special.Mamf. Special Cas. .126(1=1.00) .132(t=.57) .172(t=2.L2*) Special.Mamf. Special Cas. .132(t=.2l) .172(t=2.00) Special Cas. .172(t=1.38) N.Metropolis— National Metropolis Special Cas.— Special Case R.Metropolis— Regional Metropolis R.Capital,Sub.— Regional Capital, Subnetropolitan D.Manuf.Mf.— Diversified Manufacturing Center with Metropolitan Functions D.Mamf.f.Mf.— Diversified Manufacturing Center with few Metropolitan FNinctions Special.Mamf.— Specialized Manufacturing Center * Significant at the 5 per cent level ** Significant at the 1 per cent level -F* TABLE 23 COMPARISON OF THE SIZES OF THE INDEXES OF ECONOMIC DIFFERENTIATION AMONG THE VARIOUS TYPES OF THE 56 STANDARD METROPOLITAN AREAS ON THE BASIS OF METROPOLITAN FUNCTIONS AND REGIONAL RELATIONSHIPS AND t RATIOS FCR THE C CATEGORIES N.Metropolis .263 R.Metropolis • 3AO R.Capital,Sub. .323 D.Manuf.Mf. .250 D.Mamf.f.Mf, .244 Special.Manuf. .220 R. Metropolis R.Capital,Sub. D.Manuf.Mf. .340(t=6.42**) ,323(t=6.19**) .250(t-.81) D.Manuf,f.Mf. Special.Manuf. Special Cas. .244(t=1.46) .220(t=4.30#*) .351(t=1.90) R.Capital,Sub. D.Mamf.Mf. D.Mamf.f.Mf. Special.Mamf. Special Cas. .323(t=2.2?*) .250(t=8.18#*) .244(t=10.7**) .220(t-17.14**).351(t=.08) D.Mamf.Mf. D.Mamf.f.Mf. Special.Manuf. Special Cas. .250(t=2.6l**) .244(t=2.93**) .220(t=4.68**) .351(ts.68) D.Manuf.f.Mf. Special.Mamf. Special Cas. .244(t=.50) .220(t=3.33**) .351(t=1.91) Special.Mamf. Special Cas. .220(t=4.00**) ,351(t=2.02) Special Cas. ,351(t=3.12) Special Cas.— Special Case N.Metropolis— National Metropolis R.Metropolis-—Regional Metropolis R.Capital,Sub.— Regional Capital, Submetropolitan D.Manuf.Mf.— Diversified Manufacturing Center with Metropolitan Functions D.Manuf.f.Mf.— Diversified Manufacturing Center with few Metropolitan Functions Special.Manuf.— Specialized Manufacturing Center * Significant at the 5 per cent level ** Significant at the 1 per cent level ■ P " O' TABLE 24 COMPARISON OF THE SIZES OF THE INDEXES OF ECONOMIC DIFFERENT!ATTON AMONG THE VARIOUS TYPES OF THE 56 STANDARD METROPOLITAN AREAS ON THE BASIS OF METROPOLITAN JUNCTIONS AND REGIONAL RELATIONSHIPS AND t RATIOS FOR THE A CATEGORIES N.Metropolis .258 R.Metropolis .325 R.Capital,Sub, .312 D .Manuf.Mf. .251 D.Mamf.f.Mf. .243 Special.Manuf, .177 R. Metropolis R.Capital,Sub. D.Manuf.Mf. D.Mamf.f.Mf. Special.Mamf. Special Cas. .325(t=4.53**) .312(t=1.54) .251(t=.35) ,243(td.36) .177(t=9.00**) .315(t=7.12**) R.Capital,Sub. D.Manuf.Mf. D.Manuf.f.Mf. Special.Manuf. Special Cas. .312(1=.46) .251(1=3.70**) .243(t=5.47**) .177(1=5.69*^) .*315(t=.l6) D.Mamf.Mf. D.Mamf.f.Mf. Special.Mamf. Special Cas. .251(t=1.85) .243(t=2.23*) .177(t=4.35**) .315(t-.06) D.Mamf.f.Mf. Special.Mamf. Special Cas. .243(t=.44) .177(1=7.71**) .315(t=3.25**) Special.Manuf. Special Cas. .177(t*2.54*) .315(t».32) Special Cas. .315(t=2.23*) N.Metropolis— National Metropolis Special Cas.— Special Case R.Metropolis— Regional Metropolis R.Capital,Sub.— Regional Capital, Submetropolitan D.Manuf.Mf.— Diversified Manufacturing Center with Metropolitan Functions D.Manuf.f,M£r—Diversified Manufacturing Center with few Metropolitan Functions Special.Manuf.— Specialized Manufacturing Center * Significant at the 5 per cent level ** Significant at the 1 per cent level 11+.© Diversified Manufacturing Centers with Pew Metropolitan Functions, or for the Specialized Manufacturing Centers. The difference between the average IED value for the Regional Capitals, Submetropolitan on one hand and that for the National Metropolises, for the Diversified Manu facturing Centers with Metropolitan Functions, for the Diversified Manufacturing Centers with Few Metropolitan Functions, and for the Specialized Manufacturing Centers on the other is significant. Both the Diversified Manu facturing Centers with Metropolitan Functions and the Diversified Manufacturing Centers with Few Metropolitan Functions have an average IED value significantly higher than the Specialized Manufacturing Centers. The average IED value for the Special Cases is sig nificantly larger than that either for the Specialized Manufacturing Centers or for the National Metropolises. For the A categories, the average IED for the National Metropolises is significantly higher than that for the Specialized Manufacturing Centers. The Regional Metro polises have an average IED value significantly higher than the National Metropolises, the Diversified Manufac turing Centers with Metropolitan Functions, the Diversified Manufacturing Centers with Few Metropolitan Functions, and the Specialized Manufacturing Centers. The Regional Capitals, Submetropolitan have a significantly much larger 11+9 IED value than the Diversified Manufacturing Centers with Few Metropolitan Functions and the Specialized Manufactur ing Centers. Both the Diversified Manufacturing Centers with Metropolitan Functions and the Diversified Manufac turing Centers with Few Metropolitan Functions have an IED significantly higher than the Specialized Manufacturing Centers. The average IED for the Special Cases is signi ficantly larger than that of the National Metropolises, the Diversified Manufacturing Centers with Metropolitan Functions, and Specialized Manufacturing Centers. The Interclass differences mentioned thus far are only those that are large enough to be statistically sig nificant. It is rather difficult to grasp the overall picture of the relationship among all the Interclass dif ferences. The size rank order of the average IED values for each of the D, C, and A categories are Indicated in Table 25- The smaller the size rank, the smaller the average IED value. Not all Interclass differences are significant, however. The Special Cases are excluded from the table because they are centers of functions other than trade and manufacturing which constitute the main criteria for the typology and are heterogeneous among themselves. Table 25 shows a definite pattern among the classes of the typology for all categories of D, C, and A except for the reversed positions among the classes strongly oriented 150 toward manufacturing of the D categories: the Diversified Manufacturing Centers with Metropolitan Functions, the Diversified Manufacturing Centers with Few Metropolitan Functions, and Specialized Manufacturing Centers. For the TABLE 25 SIZE RANK ORDER OF THE AVERAGE INDEXES OF ECONOMIC DIFFERENTIATION FOR EACH OF THE D, C, AND A CATEGORIES Type of SMA* s Types of Cateaories D C A N. Metropolis....... k k R. Metropolis....... 6 6 6 R. Capital, Sub.... 5 5 5 D. Manuf. Mf........ 1 3 3 D. Manuf. F. Mf.... 2 2 2 Special. Manuf..... 3 1 1 categories of C and A, the patterns are Identical. The size rank order Is the Specialized Manufacturing Centers, the Diversified Manufacturing Centers with Few Metropolitan Functions, the Diversified Manufacturing Centers with Met ropolitan Functions, the National Metropolises, the Region al Capitals, Submetropolitan, and the Regional Metropolises in that order from the most even distribution of industries to the least even distribution of industries. For the D 151 categories, the National Capitals, Subraetropolitan follow the pattern of C and A categories, but for the Specialized Manufacturing Centers, the Diversified Manufacturing Cen ters with Pew Metropolitan Functions, and the Diversified Manufacturing Centers with Metropolitan Functions, the positions are reversed. The reason for this ranking is not difficult to find. All these classes are highly related to manufactur ing. The twelve smaller subcategories (the D categories) are used for the construction of the IED. When the three broad categories of primary, secondary, and tertiary are broken down into smaller subcategories, the manufacturing specialization stands out and contributes to the increase of the IED value. As far as the three manufacturing classes of the D categories are conoerned, the higher the size rank, the more specialized in manufacturing they are. The reason for the reversal of this phenomenon in the C and A categories is that the manufacturing specialization is absorbed into secondary industry whose increase counter balances tertiary industry and brings about more even distribution of the three broad industries. It is of in terest to note that for all D, C, and A categories, all the classes highly related to manufacturing have an average IED value lower than those less related to manufacturing. In the above analysis only the Special Cases are 152 excluded. However, the National Metropolises may also be excluded In pattern analysis, for the category of National Metropolis Is based on rather arbitrary criterion as It Is too complex for Investigation considering the limitations of time and resources for the study. When the C and A TABLE 26 SIZE RANK ORDER OF THE AVERAGE INDEXES OF ECONOMIC DIFFERENTIATION FOR EACH OF THE D, C, AND A CATEGORIES Type of SMA* s Types of Categories D C A R. Metropolis..... 5 5 5 R. Capital, Sub... • k k k D. Manuf. Mf...... 1 3 3 D. Manuf. F. Mf... • 2 2 2 Special. Manuf.... 3 1 ] . categories are used in the construction of the IED, the average IED value goes up as the class moves up In the "hierarchy” of the typology as shown in Table 26. Those classes highly associated with manufacturing, namely, the Specialized Manufacturing Centers, the Diversified Manu facturing Centers with Few Metropolitan Functions, the Diversified Manufacturing Centers with Metropolitan Functions, consistently have an average IED value smaller 153 than the other two classes which are not manufacturing oriented, namely, the Regional Metropolises and the Regional Capitals, Submetropolitan. There is a similar consistent pattern for the D categories as well, except for the fact that the positions are reversed for the manufacturing oriented classes. The reason for this phenomenon has already been stated. Even though not all interclass differences of the IED are statistically sig nificant, It may be said that in general the subclasses of the typology by Duncan and his associates are followed by a consistent pattern of evenness and unevenness of the distribution of industries. CHAPTER VI SUMMARY AND CONCLUSIONS Previous studies of the relationship of social organization and urbanization have indicated that, within urban comraunities, differences in the degree of urbaniza tion are associated with differences in the degree of economic differentiation. This research has explored the role of economic differentiation in the differentiation of urban communities within the framework of the "ecological complex," as based on referential concepts consisting of population, environment, technology, and organization. The focus of attention was on the concept of social organ ization, especially, economic differentiation. It was not the concern of this study to examine in detail the theo retical utility of the referential concepts of the eco logical complex but rather to use the concepts as a frame of reference in an investigation of the differentiation of urban communities. A number of studies have been made of the economic differentiation of urban communities, but their focus has been largely upon types rather than degrees of functional specialization. One of the few attempts to consider de grees of functional specialization was made by Duncan and his associates in their use of the concept of dissimilarity 151+ 155 of occupational structure. They utilized an index of dis similarity to indicate the percentage of the labor force in an urban community which has to be shifted to different categories In order to make its distribution the same as that for the United States as a whole. The concept of economic differentiation as used in this study is related to, but different from, the concept of dissimilarity. As used here the economic differentiation was described in terms of an index that illustrates how the percentage of labor force in the various industries within an urban community are distributed in relation to each other. Duncan and his associates did not try to relate the index of dissimilarity to indexes of other socio economic characteristics of urban communities. In this study, however, economic differentiation as measured by an index of economic differentiation was related to indexes of other socio-economic aspects of society. A selected group of socio-economic variables that were found to be related to industrialization, urbanization, and economic differentiation were examined in relation to the economic differentiation as measured by the index of economic dif ferentiation within the framework of the ecological com plex. The four referential concepts of the ecological complex were represented by selected components of society which, in turn, were represented by their selected 156 empirical Indicators. The components of society were chosen because they have been traditionally emphasized and found to be impor tant in the study of urban communities, especially, In the study of economic differentiation. The components of size, growth rate, and composition were selected under the general referential concept of population; the component of location under the general referential concept of en vironment; the component of type of production under the general referential concept of technology; and the com ponent of economic differentiation under the general refer ential concept of organization. The following selected empirical indicators of the components of society were chosen because previous studies of urban communities indicated their importance in various ways and because they were comparatively readily available. Size rank was selected under the component of size; per centage increase of population, 191+0 to 1950, under the component of growth rate; and median age, ratio of persons under twenty and over forty-four years of age to those be tween twenty and forty-four; fertility ratio, ratio of married couples to the total population, percentage of the foreign-born, percentage of female labor force, percentage of clerical and kindred workers, median number of school years completed, and median income under the component of 157 composition; population potential under the component of location; manufacturing specialization as classified by Duncan and Reiss under the component of technology; and the index of economic differentiation, the index of dis similarity, and the "hierarchical" functional classifica tion by Duncan and his associates under the component of economic differentiation. The general formula for describing economic differ entiation was expressed as £ (p - jj) ^ • ln above formula n is the number of industry subcategories and p is the proportion of the labor force in each industry subcategory. The resulting index of economic differentia tion has values that range from the minimum of 0 to the maximum of 1 and indicate how evenly or unevenly the various subcategories of a variable are distributed. When all the subcategories are evenly distributed--each sub category has the same proportlon--there is no differentia tion, and the IED value will be 0. However, if concentra tion Is found In one subcategory only and no others--the proportion of one subcategory is 1 and all the others 0-- then the most extreme form of differentiation is found, and the IED value will be 1. All other combinations of evenness or unevenness will be recorded somewhere in the range between 0 and 1. For the purpose of comparison three different groupings of industry categories were used for 158 the study of the economic differentiation. The first grouping of Industry categories Is by Duncan and his associates;-^ the second grouping by Colin Clark; the third grouping by this writer. For the sake of convenience these groupings of industry categories were called the D, C, and A categories, respectively. For the D categories the detailed industry classification of the U.S. Census of Population: 1950 is regrouped Into twelve broader categories In terms of their resource and market orientation. This grouping of industry categories was used in the belief that It provides an excellent rationale for regrouping the detailed industry classification for meaningful analysis. For the C categories Colin Clark classifies Industries Into three broad categories of primary, secondary, and tertiary. Agriculture, forestry, and fishing are Included In primary Industry; mining, construction, and manufacturing In secondary industry; transportation, trade, finance, services, and public ad ministration in tertiary Industry. The C categories and the A categories are exactly Identical except that mining ^0-tia Dudley Duncan, W. Richard Scott, Stanley Lieberson, Beverly Duncan, and Hal H. Wlnsborough, Metro polis and Region (Baltimore: The Johns Hopkins Press. i960). ^Colin Clark, The Conditions of Economic Progress (London: Macmillan and Company, Limited, . 159 la Included in primary industry rather than in secondary Industry on the assumption that mining may be better con sidered as an extractive Industry. These three broad cate gories were assumed to be important because recently atten tion has been paid to the relationship between industrial ization and urbanization and the relative proportion of workers in primary, secondary, and tertiary industries.^ All of the fifty-six SMA*s with 300,000 or more inhabitants as defined by the U.S. Bureau of the Census in 1950 were compared with a sample of nineteen SMA*s selected by a systematic random sampling from the SMA*s with in habitants between 100,000 and 300,000. SMA*s rather than Incorporated cities were chosen for study on the assumption that the former constitute a more useful unit for economic analysis than the latter. The data used in this study came from the 1950 U.S. Census of Population,^ Social Character istics of Urban and Rural Communities. 1950.^ and -^Clark, op. clt.; Wilbert E. Moore, Economy and Society (Hew Yorlcl Doubleday and Company, Inc., 19^5)» Samuel^Pratt, "Metropolitan Community Development and Change in Sub-Center Eoonomlc Functions," American Socio logical Review. XXII (1957), PP. U3h-hU0. ^U.S. Bureau of the Census, U.S. Census of Popula tion: 1950. Vol. II. Characteristics of the Population (Washington, D.cT: U.S. Government Printing Office, 1952). ^Otis Dudley Duncan and Albert J. Reiss, Jr., Social Characteristics of Urban and Rural Communities. 1950 (hew York: ^Tohn Wiley and Sons , Inc., 1955). 160 Metropolla and Region.^ There were two reasons for the choice of the fifty-six SMA*a for study and for the use of the 1950 U.S. Census of Population for data: first, to maintain comparability with the study of Duncan and his associates,7 and second, to concern itself with the finding of principles rather than with an accurate description of current socio-economic conditions. Fifteen hypotheses pertaining to the relationships of the IED to the variables describing the ecological com plex were formulated. The null hypotheses were tested against their alternative hypotheses. For a lack of a better alternative a linear relationship was assumed. Since the directions of relationships among the variables studied were Indicated by previous studies related to these variables one tail tests were conducted. The statistical level of significance was set at 5 per cent for rejection of any null hypothesis. Spearman's rank-difference co efficient of correlation, Pearson's product-moment of coefficient, t ratio, chi square, the contingency coeffi cient, and the Fisher exact probability test were used to test the null hypotheses. The following were the fifteen hypotheses: ^Duncan et. al., op. clt. 7Ibld. 161 1. The else rank of the SMA in 1950 increases with an increase in the value of the IED. 2. The increase of population as measured by the percent age increase of total population between 19U0 and 1950 of the SMA in 1950 Increases with an increase in the value of the IED. 3. The median age of the SMA In 1950 decreases with an increase In the value of the IED. 1+. The ratio of persons under twenty and over forty-four years of age to those between twenty and forty-four of the SMA in 1950 decreases with an Increase in the value of the IED. 5. The fertility ratio of the SMA in 1950 decreases with an increase In the value of the IED. 6. The ratio of married couples to the total population of the SMA in 1950 decreases with an increase in the value of the TED. 7. The percentage of the foreign-born aged twenty-one and over to the total population decreases with an increase In the value of the IED. 8. The percentage of female labor force to the persons fourteen years old and over of the SMA in 1950 In creases with an increase in the value of the IED. 9. The percentage of clerical and kindred workers to the total employed of the SMA in 1950 increases with an increase In the value of the IED. 10. The median number of school years completed for persons twenty-five years old and over of the SMA In 1950 Increases with an increase in the value of the IED. 11. The median income of the SMA In 1950 decreases with an Increase in the value of the IED. 12. The population potential of the SMA In 1950 decreases with an Increase In the value of the IED. 13. The degree of specialization in manufacturing of the SMA in 1950 decreases with an increase in the value of the IED. 162 11+. The value of the Index of dissimilarity of the SMA In 1950 Increases with an Increase In the value of the IED. l^. The rank of class in the "hierarchical" typology of Duncan and his associates of the SMA in 1950 Increases with an increase in the value of the IED. There was a definite pattern in the geographical distribution of the IED when the IED values were divided into high and low categories at the median. Broadly speak ing, SMA* s with low IED values were located In the New England, the Middle Atlantic, the East South Central, the East North Central, and the West North Central Regions. SMA1s with high IED values were found In the South Atlantic, the West South Central, the Mountain, and the Pacific Regions. The Regions with large proportions of SMA's with low IED values seem to be associated with areas of high population potential, while those with large pro portions of SMA* s with high IED values are related to areas of low population potential. There was no significant difference in the average IED values between the fifty-six SMA*s and the nineteen sample SMA's. However, both within the fifty-six SMA's and within the nineteen sample SMA's, there were differ ences in the average IED value among the D, C, and A cate gories. The average IED values for both C and A categories were significantly larger than the average IED value for the D categories. 163 A larger number of socio-economic variables were significantly associated with economic differentiation as measured by the IED for the flfty-slx SMA1 s than for the nineteen sample SMA*s. Null hypothesis 1 was that the association between values of the IED for SMA's In 1950 and the size rank Is zero. This null hypothesis was not rejected both for the flfty-slx SMA1s and for the nineteen sample SMA»s. There was no association between the IED and size rank. Null hypothesis 2 was that the association between values of the IED for SMA1s In 1950 and the Increase of population as measured by the percentage Increase of total population between 19l(.0 and 1950 Is zero. The null hypo thesis was rejected for all categories both for the fifty- six SMA*s and for the nineteen sample SMA*s In favor of its alternative that the association between the two variables is positive. Null hypothesis 3 was that the association between values of the IED for SMA*s In 1950 and the median age is zero. The null hypothesis was not rejected for the D cate gories of the nineteen sample SMA's, but It was rejected for all categories of the fifty-six SMA*s and for the C and A categories of the nineteen sample SMA's in favor of Its alternative that the association between the two variables is negative. 16U Null hypothesis U. waa that the association between values of the IED for SMA*s In 1950 and the ratio of per sons under twenty and over forty-four years of age to those between twenty and forty-four Is zero. This null hypo thesis was not rejected for the A categories of the nine teen sample SMA*a, but was rejected for all categories of the fifty-six CMA*s and for the D and G categories of the nineteen sample SMA*s in favor of Its alternative hypo thesis that the association between the two variables is negative. Null hypothesis 5 was that the association between values of the IED for SMA*s in 1950 and the fertility ratio is zero. The null hypothesis was not rejected for all categories both of the fifty-six SMA*s and of the nineteen sample SMA* s. Null hypothesis 6 was that the association between values of the IED for SMA* s In 1950 and the ratio of married couples to the total population is zero. This null hypothesis was not rejected for the C and A categories of the fifty-six SMA's and for the D categories of the nine teen sample SMA*s, but was rejected for the D categories of the fifty-six SMA*s and for the C and A categories of the nineteen sample SMA* s In favor of its alternative that the association between the two variables is negative. Null hypothesis 7 was that the association between 165 values of the IED for SMA* s in 1950 and the percentage of the foreign-born aged twenty-one and over to the total population Is zero. The null hypothesis was not rejected for all categories both of the fifty-six SMA's and for the D and C categories of the nineteen sample SMA's, but was rejected for the A categories of the nineteen sample SMA*s in favor of its alternative that the association between the two variables is negative. Null hypothesis 8 was that the association between values of the IED for SMA's in 1950 and the percentage of female labor force to the persons fourteen years old and over is zero. This null hypothesis was not rejected for the D categories of the fifty-six SMA1s and for all cate gories of the nineteen sample SMA's, but was rejected for the C and A categories of the fifty-six SMA*s in favor of its alternative that the association between the two vari ables is positive. Null hypothesis 9 was that the association between values of the IED for SMA* s in 1950 and the percentage of clerical and kindred workers to the total employed is zero. The null hypothesis was not rejected for the D categories both of the fifty-six SMA* s and of the nineteen sample SMA*s, but was rejected for the G and A categories both of the fifty-six SMA*s and of the nineteen sample SMA's in favor of its alternative that the association between the two variables is positive. Null hypothesis 10 was that the association between values of the IED for SMA's in 1950 and the median number of school years completed for persons twenty-five years old and over is zero. This null hypothesis was not re jected for all categories of the nineteen sample SMA's, but was rejected for all categories of the fifty-six SMA's in favor of its alternative that the association between the two variables is positive. Null hypothesis 11 was that the association between values of the IED for SMA's In 1950 and the median income is zero. The null hypothesis was not rejected for all categories of the fifty-six SMA's and for the D categories of the nineteen sample SMA's, but was rejected for the C and A categories of the nineteen sample SMA* s in favor of its alternative hypothesis that the association between the two variables is negative. Null hypothesis 12 was that the association between values of the IED for SMA's In 1950 and values of popula tion potential is zero. This null hypothesis was rejected for all categories of both the fifty-six SMA's and the nineteen sample SMA's In favor of its alternative that the association between the two variables is negative. Null hypothesis 13 was that the association between values of the IED for SMA's In 1950 and degrees of 167 specialization In manufacturing la zero. Only the flfty- slx SMA*s were examined In relation to this null hypothesis since the number of SMA*a classified aa either high or low In manufacturing specialization among the nineteen sample SMA1s was very small. The null hypothesis was rejected in favor of its alternative that the association between the two variables Is negative. Null hypothesis 1U was that the association between values of the IED for SMA* s in 1950 and values of the index of dissimilarity is zero. This null hypothesis was not rejected for all categories of the nineteen sample SMA's, but was rejected for all categories of the fifty-six SMA's in favor of its alternative that the association between the two variables is negative. Null hypothesis 15 was that the association between values of the IED for SMA's in 1950 and ranks of classes In the "hierarchical" typology of Duncan and his associates Is zero. Only the flfty-slx SMA's were investigated in rela tion to the null hypothesis since the nineteen sample SMA's were not included in the "hierarchical" typology of Duncan and his associates. The null hypothesis was "partially" rejected In favor of Its alternative that the association between the two variables is negative In that significant associations existed for some classes, but not for others In the typology. Broadly speaking, however, values of the 168 IED increased with an increase in the rank of class in the "hierarchical” typology. The five socio-economic variables that were most significantly associated with economic differentiation as measured by the IED were: (1) population increase, (2) median number of school years completed, (3) the Index of dissimilarity, (L | . ) population potential, and (5) ratio of persons under twenty and over forty-four years of age to those between twenty and forty-four. The first three variables were positively associated with the IED, and the last two were inversely associated with the IED. The above associations may be Interpreted in the following manner: an uneven distribution of industries is asso ciated with a large proportion of people who must be shifted to different labor force categories In order to make the distribution of the labor force activities of an urban community the same as that of the United States as a whole; with an Increase of population between 191+0 and 1950; and with a high median number of school years com pleted. An even distribution of industries is associated with high generalized accessibility of an urban community to population and a high ratio of persons under twenty and over forty-four years old to those between twenty and forty-four. In spite of the fact that size of place is one of the moat important variables in the analysis of urban com munities, economio differentiation as measured by the IED was not significantly associated with this variable. This independence of evenness or unevenness of the distribution of industries within an urban community from size of place strongly suggests the desirability of considering evenness or unevenness of the distribution of industries as one of the important variables to be taken into account in the analysis of urban communities. Also the fact that so many socio-economic variables were associated with economic dif ferentiation as measured by the IED supports the above idea. With increasing attention paid to the relationship between the relative distribution of workers among the three broad categories of primary, secondary, and tertiary and industrialization and urbanization, the concept of degree of functional specialization which may be measured by the IED becomes especially useful in the study of the differentiation of urban communities. One of the chief advantages of the IED Is that it can be applied to any number of subclasses of a variable. The relative distribution of workers in secondary and ter tiary Industries with primary industry excluded may be considered especially In the analysis of urban communities where the latter plays a relatively small role. An experi ment may be made on the basis of such categories as 170 agriculture, manufacturing, construction, wholesale trade, retail trade, service trade, education service, and gov ernment service. A very crude study of thirty sample cities chosen by a systematic random sampling from the incorporated cities with inhabitants between 5>0,000 and 100,000 in 1950 on the basis of the above categories re veals an interesting result. Table 27 shows that the cities included were differentiated with regard to the IED values, and high IED values were associated with high specialization in manufacturing,and vice versa. This type of Index serves as a useful tool In the analysis of social change If the same categories are used for each of the years studied. Any changes in the pattern of the distribution of Industries will be easily found over the years, and the direction of change will be determined. The index of economic differentiation will be more useful if it is used as a starting point for further detailed analysis of individual cases along with other indexes such as the index of dissimilarity and location quotients. This type of Index of differentiation may be profitably employed not only for industry analysis but also for the analysis of other aspects of rural as well as urban communities. This research strongly indicates that it is important to con sider degrees of functional specialization along with types of functional specialization in the study of the 171 TABLE 27 DISTRIBUTION OF THE INDEXES OF ECONOMIC DIFFHLENTIATION FOR THE SAMPLE 30 CITIES WITH 50,000— 100,000 INHABITANTS BY MANUFACTURING AND TRADE SPECIALIZATION City Index of Economic Differentiation Specialisation* Mamfacturixuc Trade Lorain .52 High NT New Britain .50 High NT Bethlehem .50 High MTC Racine .47 High MTC Niagara Falls .45 High MTC Cicero .44 High MTC Lynn .36 — MTC Bayonne .37 — Ww Dearborn .35 High MTC Saginaw .34 — MTC Lancaster .33 — MTC Covington .32 — NTT Brockton .31 -- NTr Irvington .30 — MTC Waterloo .30 — MTC South Gate .27 ---- NTw New Albany .21 — MTC Quincy .16 — MTC Oak Park .16 — Rr Richmond .16 — Ww Cleveland Hights .14 ---- NTr Alameda .14 Low NT Raleigh .13 Low TCw Greenville .13 — TC Charleston .12 — TC Lincoln .10 — MTC Wichta Falls .09 Low MTC Lubbock .08 Low TC Amarillo .07 Low — Orlando .06 Low TCr •Obtained from: Otis Dudley Duncan and Albert J. Reiss, Jr., Social Characteristics of Urban and Rural Communities. 1950 (New York: John Wiley and Sons, Inc7,~ 195^) # pp. 337-410. **Ww— Wholesale trade centers; Rr— Retail trade centers; TC— Trade centers; TCw— Trade centers, wholesale; TCr— Trade centers, retail; MTC— Maintenance trade centers; NT— Nontrade centers; NTw— Nontrade centers, wholesale; NTr— Nontrade centers, retail. economic differentiation of urban communities. APPENDIX APPENDIX I DERIVATION OP THE INDEX OP ECONOMIC DIFFERENTIATION The index desired is one that obtains the value 0 when all workers are equally distributed with respect to all industry subcategories of an industry classification and attains the maximum value 1 when all workers are con centrated in one industry subcategory only and no others. The variable is defined as the proportion of workers in the i1th industry subcategory of an industry classification, and n is defined as the number of sub categories of this industry classification. By definition, then, the sum of all p.'s is 1, £ p.rl. Since O^pXl, izl " = 2< n 2.C then p, -p. for all p.'s and £ p. -1. Under the condi- i = 1 n n 0 tion ^ b^l, the quantity ^ p.^ can be shown by mathe- i=l 1=1 matical analysis to attain a unique relative minimum value 1 n o when all p 's have the same value p and £ p ^ in i = 1 . ] > 2 i j i ° =n(±) z—. Hence the range for £ p/ is from the minimum n n i=1 i value — to the maximum value 1. n It is desirable for £ p * to range from a i = l minimum value of 0 to a maximum of 1. Hence the following transformations are necessary. 17) Let p.. 'rp-j-i. so that p. '=0 n i = l 175 n and n o T o l n ^ o n £ Pl-2=^ Ipi2 - £P1 t (i)2} = £ Pl2 - |g p ♦ ((ly 1=1 1=1^ n J 1=1 1=1 1 V n n n = £ Pi2 - — 1 ♦ ~ = £ Pi2 - - 1=1 n n i=1 1 n n 2 n hence the maximum value of ^ p, ' = ^ p ‘ - A is 1 - — 1=1 1=1 1 n n -n~l since the range for p 2 Is from a minimum of 1 . to n n a maximum of 1 and the minimum Is 0. Now make the transformation p^ 1 1 = p^ ' or p^'^r p^'2 'n-l " 1 n-l n so that ^ p, ' ' 2 ranges from a minimum vslue of 0 to a i = 1 maximum value of 1 where p,'' = _IL_(p. - — )' 1 n-l 1 n The index of economic differentiation then is defined to n ’ n , be Iffl: £ P,"2: £ A ( P , - ±)' 1:1 1 l=ln 1 1 n BIBL IO GR AP HY Public Documents U.S. Bureau of the Census. Eighteenth Census of the United States: 1950. Population, vol. it. Books Barnes, Harry Elmer (ed.). An Introduction to the History gf Sociology. Chioago: The University of Chioago Bartholomew, Harland. Urban Land Uses, Cambridge: Harvard University fcresa, 1932. Becker, Howard, and Boskoff, Alvin. Modem Sociological Theory. Hew York: The Dryden Press, Bogardus, Emory S. The Development of Social Thought. Hew York: Longmans, Green and bompany, 195b. Bogus, Donald J. The Structure of the Metropolitan Community. Ann Artor. Michigan: University of MichiganPress, 1950. and Harris, Dorothy L. Comparative Population and Urban Research Via Mul - - - - ■nd nptitp nesearoh via Multiple Regression ana Covariance Analysis. Oxford. 6kto: Scrlppa Foundation, Cattell, Raymond B. Factor Analysis. Hew York: Harper and Brothers, 1952. Churchman, C. Vest, Aokoff, Russell L., and Wax, Murray (ed.). Measurement of Consumer Interest. Phila delphia: “TnlversTEyoTTemiiiylvanlTa^ 19^7. Clark, Colin. The Conditions of Economic Progress. London:' “HaomU lan and Company, Limited, I9I 4 .O. Cole, William E. Urban Society. Boston: Houghton Mifflin Company, 1<!>5B. Cuzzort, Raymond P. Suburbanization of Service Industries Within Standyd Metropolitan Areas. Oxford. Ohio: Scrlppa Foundation, 19^57 177 178 Dloklnson, Robert S. City, Region, and Regionalism. London: Kegan Paul, ferencn, Trubner and Company, Limited, 191+7. Dobriner, William M. (ed.). The Suburban Community. Hew York: G. P. Putnam*a Sons, l95o. Duncan, Otla Dudley, and Relaa, Albert J., Jr. Social Characteristica of TTr^f" »nd Rural Comsmnltlea. 1458. Yonti John Wiler and Sons. So.*1956. _______ , Scott, W. Richard, Lleberaon, Stanley, Duncan, Bererly, and Wlnsborough, Hal H. Metroptfij** Region. Baltimore: The Johna Hopkins Press, i960. Durkhelm, Bmlle. The Division of Labor In Society. Trans lated by George simp a on. ttew lork: The Macmillan Company, 1933* Brlcksen, B. Gordon. Urban Behavior. New York: The Maomlllan Company, 145i+. Plrey, Walter. Land Pie In Central Boston. Cambridge, Massachusetts: fearrard University Press, 191+7. Fisher, Robert Moor (ed.). The Metropolis in Modern Life. New York: Doubleday and &o^pany, 1955. Gillen, Paul Bates. The Distribution of Occupations as a City Yardstick, fcew York: king*s Crown Preas. Columbia trnlreralty, 1951. Gist, Noel P., and Halbert, L. A. Urban Society. New York: Thomas Y. Crowell Company, 19*6. Grablll, Wilson H., Kiser, Clyde V., and Whelpton, Pascal K. The Fertility of Amerloan Women. New York: John Wiley and Sons, xho., 1^50. Gutklnd, B. A. Community and Enylronment. New York: , B. A. Community and Enylro Philosophical Library, 195I+. Hallenbeok, W. F. American Urban Society. New York: Van Nostrand, 195^. Hatt, Paul K., and Reiss, Albert J., Jr. (ed.). Cities ^nd^Soolety. Glencoe, Illinois: The Free Press, 179 Hauser, Philip M., and Dunotn, Otia Dudley (ed.). The Study of^Populat^ji. Chioago: The University of Hawley, Amos H. Hunan Boology. Hew York: The Ronald Preaa Company, 1956. Hoover, Bdgar M., Jr. The Location of Boononlc Activity. Hew York: MoCraw-tilll, Hurd, Rlohard M. Principlesof City Land Valuea. Hew York: The hecord and Guide, 1903. Kitagawa, Evelyn M., and Bogue, Donald J. Suburbanization of Hanufaoturlng Aotlvlty within Standard keferopo- lltan Aroaa. Oxford. Ohio: Sorlpps foundation. 1955. Loeaoh, Auguat. The Boononlca of Looation. Hew Haven: Yale University Press, l95h« Mercer, Blaine E. The Anerloan Community. Hew York: Random House, 1^6. Miller, Hernan P. Income of the Anerloan People« New York: John wiley and &ons, Inc., 1955* Moor, Wilbert B. Economy and Sooletv. Carden City, Hew York: Doubledayand Company, Inc., 1955* Park, Robert Bara. Hunan Conmunltlea. Clenooe, Illinois: The Free Preaa, 1952. ________, and Burgeaa, Ernest E. An Introduction to the Solenoe of Soolology. Chioago: The tJniveraity of Chicago fcreea, 1921. ________, Burgeaa, Brneat W., and McKenzie, Roderick D* The City. Chioago: The University of Chioago Press, 1925* Quinn, Janes A. Hunan Ecology. Hew York: Prentloe-Hall, 1950. Roaai, Peter. W&JjM»jUAeg,^yei.. A Study in the Sooj^l Psychology of Urban Residential Mobility.Clenooe, Illinois: The Free Preaa, 1955. 180 Shevky, Bshref, and Ball, Wendell. Social Area Analysis: Theory. Illustrative Application, and fcorouiation frr^ceduree. frtanfor£: Stanford University Press. _______ , and Will lau, Marilyn. The Social Area of Lob Angeles: ij^flf.a^Typplggy. **>• Angeles: University or California press, 19U9. Smith, T. V., and White, Leonard D. (ad.). Chloy o: An grperlnent In Sociy Solsnos Re a ear eh. Chicago: fba University of Cnieago Preaa, 19^9. Sorokin, Pitlrlm. Contenporyy Sociological Theories. Hew York: karper an^i Brothers, 192b. — - Spangler, Joseph J., and Duncan, Otla Dudley (ad.). Population Theory and Policy. Glencoe, Illinois: Wie^ree Press, ±9$1>T Thorndike, E. L. s e ller Cities. How York: Haroourt, Braoa and Company, l^O. _______. Your City. Haw York: Harcourt, Brace and Company, 1939. Thurstons, L. L. Multiple Paotor Analysis. Chioago: Tha Universityof Chioago £ress, 19^7. Tlmasheff, Hlcholas S. Sociological Theory. Garden City, Hew York: Doubleaay ana Company, Inc., 1955* Tryon, Robert C. Cluster Analysis. Ann Arbor, Michigan: Edwards Brothers, 1939* . Identification of Social Areas by Cluster Analysis* Los Angeles: frnlverslty of California Press, 1955* Vance, Rupert B., and Jenerath, Hlcholas J. (ed.). The Urban South. Chapel Hill: The University of Worth Carolina freis, 195U* Williams, Robin M., Jr. Anerloan Society; A Sociological Interpretation. Hew York: Alfred A. Knopf, 1951. Wlrth, Louis. Community Life and Social Policy. Chioago: The University of Chicago >ress, 1956. 181 Zipf, George Kingsley. H a m Behavior and the Prlnolple of Leaat Effort. Cambridge. Maaoachnaetls: Adalson-ttesley Proas, Inc., 191*9. Articles and Periodicals Adcook, C. J. "A lots on Cluster Directed Analysis," Psrchometrlka. XVII (1952), 21*9-251*. Alexander, John V. "Tho Baslo-Honbaslc Conoopt of Urban Econcmio Punotlons," Economlo Geography. XXX (19514.), 214.6-261, Anderson, Nels. "Urbanism and Urbanisation." Amerloan Journal of Sociology. LXV (1959), 66-7TI Arenaberg, Conrad M, "The Comsnmlty-Study Method," American Journal of Soelology. LX (19514-), 109-121*. Baltzell, E. Dlgby. "Urbanisation and Governmental Administration In Lower Bucks County." Soolal Problems. II (1951|-), 38-1*6. Bell, Wendell. "The Soolal Areas of the San Francisco Bay Region," American Sociological Review. XVIII (1953), 39-l*‘ 7. _______, and Boat, Marlon D. "Urban Heighborhoods and Informal Social Relations." Amerloan Journal of Soolology. LXII (1957), 391-5WT Blumenfeld, Hans. "On the Concentrlo-Clrcle Theory of Urban Growth," Land Boonomlos. XXV (19l*9), 209-212. Bourland, D. David. "The Distribution of Professions within Cities in the United States," American Journal of Psychology. LXIII (1950), Calhoun, John B. "Soolal Welfare as a Variable in " "-"— osla on Quantitative Population Dynamics." Symposia Biology. XXII (1957?, 559=3557 Clark, Colin. "The Economic Functions of a City in Relation to Its Sise," Econometrics. XIII (1951*), 97-H3. 182 Clark, P. J., and Evans, F. C. "Dlstanoe to Haarest neighbor as a Maaaura of Spatial Relations," Ecology. XXXV (195M, U45-&53. Conway, Preda. "Tha Induatrlal Structure of Towns Manchester Eoononlo and Soolal Stadias. Darla, Kingslay, and Golden, Hilda Harts, "Urbanisation and tha Development of Pra-Induatrlal Areas," Eoononlo Development and Cultural Change. Ill ( l M f r / g - g S . ' — --------------- Dunoan, Otla Dudlay. "Urbanisation and Ratall Specialisa tion," Soolal Foroas. XXX (1952), 267-271, _______• "Review of Shevky and Ball, Soolal Area Analysis." Amerloan Journal of Soolology. LXI (1955). SU-BT. _______• "Raaaarch on Matropolltan Population: Evaluation of Data." Journal of tha Anerloan Statistical Association.' El *------ , and Dunoan, Beverly Davis. "A Methodological ’ Analysis of Segregation Indexes," Anerloan Soolologloal Review. XX (1955)» 215^217. _______, and Dunoan, Beverly Davis. "Residential Distri bution and Oeoupatlonal Stratification," Anerloan Journal of Soolology. LX (1955), lj.93-503. Erickaen, E. Gordon. "Tha Superhighway and City Planning," Soolal Foroas. XXVIII (1950), lj.29-14.3^. Gibbs, Jaok P., and Sohnore, Lao F. "Matropolltan Growth: An International Study," Anerloan Journal of Soolology. LXVI (I960), 166-176. Gross, Edward, "Tha Role of Density as a Factor In Metro politan Growth In tha United States of Anerloa," Population Studies. VIII (195^), 113-120. Guttman, Louis. "The Cornell Technique for Soale and Intensity Analysis," Eduoational and Psyohologloal Measurement. VII (19l|.7)# Si/-z79- 183 Harris, Chaunoy. "A Functional Classification of Cities In the United States," Geographical Review. XXXIII (191*3). 86-99. Hart, John Fraser. "Funotlons and Occupational Structures of Cities of the Anerloan South." Annals of the Anerloan Geographers. xlv (1955?. 259-285. Isart, Valter, Kayesh, Robert A., and Kuenne, Robert E. "The Eoononlo Base and Structure of the Urban- Metropolitan Region," Anerloan Sociological Review. XVIII (1953), 317-321. Jahn, Julius A. "The Measurenent of Eoologloal Segrega tion," Anerloan Sociological Review. XV (1950), 100-101* . Jones, Victor. "Eoononlo Classification of Cities and Metropolitan Areas," Municipal Year Book. (1953), 1*9-57. Kneedler, Grace M. "Functional Types of Cities," Publlo Managenent. XXVII (1951*), 197-203. Knight, John K. "Certain Changes In Population Distribu tion In the United States," Geographical Review. XXXI (191*1), 1*88-1*90. Kuper, Leo. "Social Solenoe Research and the Planning of Urban neighborhoods," Soolal Foroes. XXIX (1951), 237-21*3. Llebman, Charles S. "Functional Differentiation and Political Characteristics of Suburbs." Anerloan Journal of Sociology. LXVI (1961), l*B5-l*^b. Meyers, Jerome K. "Rote on the Homogeneity of Census Tracts: A Methodological Problem In Urban Eco logical Research," Social Foroes. XXXII (1951*), 361*- 366. Velson, Howard J. "A Serrloe Classification of Amerloan Cities," Economic Geography. XXXI (1955), 189-201. Relas, Albert J., Jr. "Functional Specialisation of Cities," Municipal Year Book. XXIV (1957), 5l*-68. _______. "Some Logical and Methodological Problems in Comsiunlty Research," Social Foroes. XXXIII (1951*), 51-57. 181* Robinson, W. S. "Ecologioal Correlations and the Behavior of Individual*." American Sociological Review. XV (1950), 351-355T Sabagh, (Georges, Van Aradol, Maurice D., Jr., and Zahedl, Hamid. "Urban Population Concentration and Occupa tional Structure In the United States: 1900- 1950," The 1960 Proceedings of the Soolal Statis tics Section. Anerloan Statistical Association. 164-171. Sohettler, Clarence. "Relation of City-Sixe to Eoononlo Services." Anerloan Soclologloal Review. VIII (191*3), 60-62. Schmid, Calvin P. "Generalisations Concerning the Ecology of the Anerloan City." Anerloan Sociological R.Tl.M, Vf (1950), 2iu-5BT;----- --- Sohnitt, R. C. "Demography and City Planning," Social Foroes. XXX (1952), 300-301*. Schnore, Leo F. "Satellites and Suburbs," Soolal Foroes. XXXVI (1957), 121-127. . "The Separation of Hone and Work: A Problen for Hunan Ecology," Soolal Foroes. XXXII (1951*), 336-31*3. _______, and Varley, David W. "Sons Conoonltants of Metropolitan Size," Anerloan Sociological Review. XX (1955), 1*08-1*11*. Smith, Joel. "A Method for the Classification of Areas on the Basis of Denographloally Homogeneous Popula tions," Anerloan Sociological Review. XIX (1951*), 201-207. Stewart, John Q. "Concerning Social Physios." Scientific American, CLXXVIII (191*8), 20-23. _______• "Bmplrloal Mathematical Rules Concerning the Distribution and Equilibrium of Population," Geogranhloal Review. XXXVII (191*7), 1*61-1*85. Swanson, G. B. "The Disturbances of Children In Urban Areas," Anerloan Soolologlcal Review. XIV (191*9) 185 Vlnlng, Rutledge. "A Discussion of the Spatial Aspects of the Structure and Functioning of an Economic System," Econometric a. XX (1952), I j . 78-I4 . 80. TJllman, Edvard. "A Theory of Location for Cities." American Journal of Sociology. XLVI (19ij.l), 853- 064. Wlrth, Louis. "Urbanism as a Way of Life," Amerloan Journal of Soolology. XLIV (1938), Reports Houser, Philip M., Dunoan, Otis Dudley, and Duncan, Beverly Davis. Methods of Urban Analysis: A Summary Report. Air fetoroe Personnel iuid training Research Center, TW-56-1. The Sprlnmfleld Survev: A Study of Social Conditions In ----- ^ r l c y " g h y ."' Sli^etea" fcy Shelby SL BlrrlSon. vol. l-III. Sew York: Russell Sage Foundation, 1918-1920. Unpublished Material Bond, Hudson J. "The Isolation of Urban Subcultures Through Social Dimensions." Unpublished Ph.D. dissertation, University of California, Berkeley, 1950. Dewey, Richard S. "Peripheral Expansion of Population In Metropolitan Milwaukee." Unpublished Ph.D. dissertation. University of Wlsoonsln, 19h7. Friedsam, Hiram J. "Some Demographio Aspects of Old Age in the United States." Unpublished Ph.D. disser tation, University of Texas, 1951* Olbbard, Harold Allen. "Residential Succession." Unpublished Ph.D. dissertation, University of Michigan, 1938. Hatt, Paul K. "A Study of Hatural Areas In the Central Residential District of Seattle." Unpublished Ph.D. dissertation, University of Washington, 19U-5* 186 Keyes, Fenton. "The Correlation of Soolal Phenomena with Community Size." Unpublished Ph.D. dissertation, Yale University, 191+2. Magee, Joseph V. "A Study of the Use of Sociological Indices of the Functional Bntlty of Washington, D.C." Unpublished Ph.D. dissertation, Catholic University, 191+7. Tisdale, Dorothy Hope. "Urbanization: A Study of the Prooess of Population Concentration In the United States and Its Relation to Social Change." Un published Ph.D. dissertation. University of Horth Carolina, 191+2. Van Arsdol, Maurice Donald, Jr. "An Bmplrleal Bvaluation of Social Area Analysis in Human Boology.” Unpublished Ph.D. dissertation, University of Washington, 1957* Walz, Orry C. "The Construction of a Rural Urban Index." Unpublished Ph.D. dissertation, University of Kansas, 1951+* Wendllng, Aubrey. "Suicide in the San Francisco Bay Region 1938-191+2 and 191+8-1952." Unpublished Ph.D. dissertation, University of Washington, 1951+.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
The Construction And Empirical Test Of A Theory Based On Selected Variables In Small-Group Interaction
PDF
Referential Dissociation And Response To Stress
PDF
Social Components Of Housing Cost In The Western Metropolis
PDF
Reference Group Theory, Selection, And The Images Of Professions
PDF
The Career Business Executive As A Definitive Occupational Type
PDF
Ecology Of Negro Communities In Los Angeles County: 1940-1959
PDF
The Effect Of Differential Treatment On Attitudes, Personality Traits, And Behavior Of Adult Parolees
PDF
Middle-Class Marital Roles - Ideal And Perceived In Relation To Adjustment In Marriage
PDF
Differential Fertility Behavior And Values In Rural And Semi-Urban Costa Rica
PDF
Prediction Of Overt Behaviors In Hospitalized Psychiatric Patients
PDF
Flexibility In Thinking And The Social Attitudes Of Catholic And Public High School Seniors
PDF
An Empirical Examination Of The Relationship Of Vertical Occupational Mobility And Horizontal Residential Mobility
PDF
Determinants Of Intercounty Migration: California, 1970-1973
PDF
Perception Of The Power Structure By Social Class In A California Community
PDF
Status Consistency Among The Clergy
PDF
Behavioral Rigidity And College Physical Science Instruction
PDF
Employment Distribution, Income, And City Size: A Statistical Analysis
PDF
Social Stratification In A Selected Community
PDF
Behavioral Seriousness And Impulse-Control Balance In Delinquency
PDF
Some Social Factors Affecting The Power Structure And Status Of A Professional Association In Reference To Social Work
Asset Metadata
Creator
Amemiya, Eiji Cannon
(author)
Core Title
Economic Differentiation And Social Organization Of Standard Metropolitanareas In The United States: 1950
Degree
Doctor of Philosophy
Degree Program
Sociology
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
OAI-PMH Harvest,sociology, general
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Van Arsdol, Maurice D., Jr. (
committee chair
), Lefever, David Welty (
committee member
), Sabagh, Georges (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c18-250643
Unique identifier
UC11358615
Identifier
6203717.pdf (filename),usctheses-c18-250643 (legacy record id)
Legacy Identifier
6203717.pdf
Dmrecord
250643
Document Type
Dissertation
Rights
Amemiya, Eiji Cannon
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
sociology, general