CR5” L155. or L52. The spatial (1‘. usually explained 'L 5cm residence: on Tue social choice :1 consequence of val; 9-1 v» non-economic mane through the intent increasing degree c Tait research feream in Ill: 5 these differences “it“. Life I :y'. based upon an ind; timalized as a txlxlbit reflect t to residential en Data come it Toronto, Ontario ABSTRACT URBAN LIFE STYLES AND ENVIRONMENT: THE EFFECT OF LOCATION AND RESIDENCE ON BEHAVIOR IN A TIME-SPACE FRAMEWORK By Brian Paul Holly The spatial distribution of residential subareas in the city is usually explained by one of two models. The economic competition model sorts residences out on the basis of competition for accessible sites. The social choice model argues for residential differentiation as a consequence of values, social distance preferences, sentiment, and other non-economic motives. Both models are complementary, but social choice, through the intervening construct life style, is seen as offering an increasing degree of explanation. This research tests the hypothesis that there are significant dif- ferences in life style between urban residential environments, and that these differences reflect differential preferences on the part of urb- anites. Life style is conceptualized as a set of predispositions to act based upon an individual's hierarchy of values. The concept is opera- tionalized as a set of activity sequence patterns. The behavior peOple exhibit reflect their desired style of living, and this can be related to residential environment. Data come from a longitudinal survey of residential movers in the Toronto, Ontario metropolitan region. Daily time budgets were recorded for middle income families, both with and without children, living in either single family homes or high rise apartments in downtown Toronto or its suburbs. The sample is relatively homogeneous in terms of income, education, OCCUPi Activities dl four hour day an: Tneactivitiea az| dentiai environs: linking any acti‘; jetted to factor in the activity d Those livin,| fashion as to ref are more family : betaeen housing t; Suburban housediw doses: 1° “TR. pr: imam“ 0!! the c 301“ and leisure l Shel! famil) “Chine: Offered orientation at the Brian Paul Holly education, occupation, and stage in the life cycle. Activities were recorded for every fifteen minute period of a twenty- four hour day and are coded according to a ninety-nine item activity code. The activities are arrayed in transition probability matrices by resi- dential environments and sex to achieve a measure of the probability of linking any activity with any other activity. The matrices were sub- jected to factor analysis to determine the underlying sequence patterns in the activity data. Those living in downtown Toronto sequence activities in such a fashion as to reflect a cosmopolitan style of life, whereas suburbanites are more family oriented in their behavior. Differences in life style between housing types are not significant when location is controlled for. Suburban housewives emerge as spatially isolated individuals who emphasize domestic work, private needs, and child care in their daily routines. Husbands, on the other hand, lead more complex daily lives centered on work and leisure behavior. Single family homeowners close to the city center enjoy the amenities offered by proximity to downtown but maintain a familistic orientation at the same time. Suburbanites in high rise apartments and single family homes pattern their activities in a similar manner. These findings tend to confirm the existence of life style variation by residential location, and support with behavioral data the empirical findings of Social Area Analysis and Factorial Ecology. As a major axis of urban residential differentiation life style deserves more attention by scholars interested in the complex social geography of the city. URBAN LIFE STYLES AND ENVIRONMENT: THE EFFECTS OF LOCATION AND RESIDENCE ON BEHAVIOR IN A TIME-SPACE FRAMEWORK By Brian Paul Holly A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Geography 1974 © Copyright by BRIAN PAUL HOLLY 1974 This docu: Ieen input fro: for their sign- 80(1ch 0f enco-e the Obstacles .i “in“ Here Co the 1d“ from w to 1“ ”mien critical final the manuscript. deserves 'Pec ia‘. Physical Envirm HOUS ins . fl . provided to int all gradute ‘ t for the content ACKNOWLEDGEMENTS This document represents the culmination of an effort which has seen input from many sources. Several persons should be singled out for their significant contributions. My wife, Marilou, was a constant source of encouragement and inspiration, especially at those times when the obstacles were greatest. Many of the graphics and much of the editing were contributed by her. Dr. Ronald Horvath provided me with the idea from which this dissertation emanated, and he guided it through to its completion. Dr. Stanley Brunn made valuable suggestions at the critical final stages and is responsible for helping shape the form of the manuscript. Dr. William Michelson of the University of Toronto deserves special mention for providing the data from his study, "The Physical Environment as Attraction and Determinant: Social Effects in Housing." His generosity is greatly appreciated. Finally, John Stephens provided an intellectual sounding board and the kind of friendship that all graduate students need. I, of course, remain solely responsible for the contents of this work. 11 LZSZOFIABLES. . . . . . LZSIJFHSL'RES . . . . - Canter 1. newton . . 2. iIRSL‘i RESIDENTIAL Residential St Competing nodc Urban Resident 3min Activit Study Design 3‘ mi 5111.3 AND Sock} Eases The Theory c SPltiai A,pi Urban find 8 Life Style: LIFE SWLE M; SYStem CE I THE y‘otiva The Intercl I Activity L ActtVity . TABLE OF CONTENTS LIST OF TABLES O O O O O O O O O O O O O O O O O 0 LIST OF Chapter 1. 2. 3. 4. FIGURES O O O I O O O O O O O O O O O O O 0 INTRODUCTION . . . . . . . . . . . . . . . . URBAN RESIDENTIAL STRUCTURE . . . . . . . . Residential Structure of the City . . . . Competing Models of Residential Structure Urban Residential Models . . . . . . . . . Human Activity Systems . . . . . . . . . . Study Design . . . . . . . . . . . . . . . LIFE STYLE AND LOCATION: A CONCEPTUAL MODEL Social Bases of Life Style . . . . . . . . The Theory of Social Choice . . . . . . . Spatial Aspects of Life Style . . . . . . Urban and Suburban Ways of Life . . . . . Life Style: A Conceptualization . . . . . LIFE STYLE MEASUREMENT: HUMAN ACTIVITY SYSTEMS Systems of Human Activity . . . . . . . . The Motivational Basis of Behavior . . . . The Interdependence of Activities . . . . Activity Linkage Analysis . . . . . . . . Activity Linkage Modelling . . . . . . . . iii ix 14 18 21 27 27 33 37 48 52 62 62 64 66 72 75 A KethOdoloi’y : s. oascamlo“ f I Data 5°urce ° Varild‘m in E Saflple Selecti roronto: 1'“ Residentiil 3 Time sodas“ e. Atmm SYST‘? Methodoloal" ' ”ll“ Life 5' Life Style .w Life Style E Conclusions Life Style Evaluation APPLVDIX A: THE. BL". APPENDIX B: run; Et‘l “’th c: mates . BIBLIOGRAPHY ‘ A Methodology for Analyzing Activity Linkages . . . . 82 55. DESCRIPTION OF THE DATA AND STUDY AREA . . . . . . . . 88 Data Source . . . . . . . . . . . . . . . . . . . . . 88 Variation in Physical Environment . . . . . . . . . . 89 Sample Selection . . . . . . . . . . . . . . . . . . 91 Toronto: The Environmental Context . . . . . . . . . 105 Residential Structure . . . . . . . . . . . . . . . . 109 Time Budgets of Human Behavior . . . . . . . . . . . 113 6. ACTIVITY SYSTEMS AND RESIDENTIAL ENVIRONMENTS . . . . . 120 Methodology . . . . . . . . . . . . . . . . . . . . . 120 Major Life Style Differences . . . . . . . . . . . . 125 Life Style By Location . . . . . . . . . . . . . . . 133 Life Style By Housing Environment . . . . . . . . . . 139 Life Style Differences by Sex . . . . . . . . . . . . 142 Linkage Dimensions . . . . . . . . . . . . . . . . . 161 7. SUMMARY, CONCLUSIONS, AND PROSPECT . . . . . . . . . . 166 Conclusions . . . . . . . . .u. . . . . . . . . . . . 170 Life Style and Residential Structure . . . . . . . . 174 Evaluation . . . . . . . . . . . . . . . . . . . . . 176 APPENDIX A: TIME BUDGET INSTRUMENT . . . . . . . . . . . . . . 181 APPENDIX B: TIME BUDGET CODE . . . . . . . . . . . . . . . . . 184 APPENDIX C: TABLES OF FACTOR LOADINGS . . . . . . . . . . . . 190 BIBLIOGMPW O O O O O O O O O O O O O O O O O O O O O O O O O 205 iv Table 3-1 3-2 5-3 5-5 5-6 M So 6-1 6~3 PRIE‘AR‘i AND SEC: LEA lNDICES C11": Sid. AND ‘ smus . , . Hosea on rat. HOUSING TY} ass or vies. A: ENV 13C 551:2? “55 0? CHILI Disratnuuot; “will? DISH OCCUPATIONAL EDUCATION 1.: 1’53.ng 0P 1 DETACH; MdIROPQL LINKAGE CO; ACTIVII l 1‘3ch C0; ACIIVII: LIN SAG; C Q; ACIIVII LIST OF TABLES Table Page 3-1 PRIMARY AND SECONDARY SPATIAL ARRANGEMENTS OF SOCIAL AREA INDICES . . . . . . . . . . . . . . . . . . . . 44 3-2 CITY SIZE AND 'F' RATIOS FOR SOCIAL RANK AND FAMILY STATUS . . . . . . . . . . . . . . . . . . . . . . . 45 5-1 NUMBER AND PERCENT OF RESPONDENTS BY LOCATION AND HOUSING TYPE, PHASE I . . . . . . . . . . . . . . . 98 5-2 AGE OF WIFB AND YOUNGEST CHILD BY DESTINATION ENVIRONMENT . . . . . . . . . . . . . . . . . . . . 99 5-3 AGES OF CHILDREN BY DESTINATION ENVIRONMENT . . . . . . 100 5-4 DISTRIBUTION OF INCOME FOR HUSBANDS AND WIVBS . . . . . 101 5-5 INCOME DISTRIBUTION BY FAMILY . . . . . . . . . . . . . 102 5-6 OCCUPATIONAL DISTRIBUTION FOR HUSBANDS AND WIVES . . . 103 5-7 EDUCATION LEVELS FOR HUSBANDS AND WIVBS . . . . . . . . 104 5-8 PERCENT OF DWELLING UNITS WHICH WBRB EITHER SINGLE DETACHED OR APARTMENTS FOR BOTH 1951 and 1961 IN METROPOLITAN TORONTO REGION . . . . . . . . . . . . 110 6-1 LINKAGE COEFFICIENTS: HUSBANDS AND WIVBS WEEKDAY ACTIVITIES - DOWNTOWN . . . . . . . . . . . . . . . 127 6-2 LINKAGE COEFFICIENTS: HUSBANDS AND WIVES WEEKDAY ACTIVITIES- SUBURBAN . . . . . . . . . . . . . . . 129 6-3 LINKAGE COEFFICIENTS: HUSBANDS AND WIVES WBBKDAY ACTIVITIES - SINGLE FAMILY HOMES . . . . . . . . . . 131 V 1151: 6-5 6-6 6-5 6-9 6-11 6-12 “ 6-13 6-14 5-15 6-16 LN iAGE COEFFEC‘. ACTIVITIES - LINN’AGE COEFFIC'. ACIIVIIIES - 11.1.1132 COEFFIC ACTIVITIES - LINN-GE COEFFIC ACIIVIIIES - LINKAGE COEFF 1c ACIIVIIIES « LINKAGE OEFFI 0015mm LIN-GE COEFF: 1 0'1 ELIE-0153.1 Table 6-4 6-6 6-7 6-8 6-9 6-10 6-11 6-12 6-13 6-14 6-15 6-16 Page LINKAGE COEFFICIENTS: HUSBANDS AND WIVES WEEKDAY ACTIVITIES - HIGH RISE APARTMENTS . . . . . . . . . . 132 LINKAGE COEFFICIENTS: HUSBANDS AND WIVES WEEKDAY ACTIVITIES - SINGLE FAMILY HOMES, DOWNTOWN . . . . . 134 LINKAGE COEFFICIENTS: HUSBANDS AND WIVES WEEKDAY ACTIVITIES - HIGH RISE APARTMENTS, DOWNTOWN . . . . . 135 LINKAGE COEFFICIENTS: HUSBANDS AND WIVES WEEKDAY ACTIVITIES - SINGLE FAMILY HOMES, SUBURBAN . . . . . 137 LINKAGE COEFFICIENTS: HUSBANDS AND WIVES WEEKDAY ACTIVITIES - HIGH RISE APARTMENTS, SUBURBAN . . . . . 138 LINKAGE COEFFICIENTS: WIVES WEEKDAY ACTIVITIES - DOWNTOWN . . . . . . . . . . . . . . . . . . . . . . 143 LINKAGE COEFFICIENTS: WIVES WEEKDAY ACTIVITIES - SUBURBAN . . . . . . . . . . . . . . . . . . . . . . 144 LINKAGE COEFFICIENTS: WIVES WEEKDAY ACTIVITIES - HIGH RISE APARTMENTS . . . . . . . . . . . . . . . . 147 LINKAGE COEFFICIENTS: WIVES WEEKDAY ACTIVITIES - SINGLE FAMILY HOMES . . . . . . . . . . . . . . . . . 148 LINKAGE COEFFICIENTS: HUSBANDS WEEKDAY ACTIVITIES DOWTOW O O O O O O O O O O O O O O O O O O O O O 0 149 LINKAGE COEFFICIENTS: HUSBANDS WEEKDAY ACTIVITIES SUBURBAN O O O O I I O O O O O O O I O O O O O O O O 1 50 LINKAGE COEFFICIENTS: HUSBANDS WEEKDAY ACTIVITIES HIGH RISE APARTPENTS O C O O O O O O O O O O O O O O 152 LINKAGE COEFFICIENTS: HUSBANDS WEEKDAY ACTIVITIES SINGLE FAMILY HOIIES O O O O O O O O O O O O O O O O O 153 v1 Tale LINKAGE. COEFF‘EL. 6-1‘3 C-1 C-2 C-3 C-5 (2-6 m 03 M MD C-ll C-lz C-13 LIMASE COEFF IC . SINGLE FNIIL‘. HIGH RISE SL’ LINKAGE COEFF IC LIN-113:: 005.9% 81;". RISE L FACTOR STRUCTL' FACICR SIAUCTL FACIJR SIRCCI‘; FACTOR STRUCI; IACTDR SIR‘JCI‘ FACICR SIRE-c: FAC'IOR SIRCCI FACTOR STRUC: FACTOR SIA'JC‘ DWI-02.2" ‘ FACTOR SIRUC SUBURBAN FACIm SIP-‘23:; HIGH &13; FACIO 3 STEAL“: | FACTOR SI“: Table 6-17 6-18 6-19 6-20 C-1 C-2 C-3 C-4 C-5 C-6 C-7 0-8 0-10 C-ll C-12 C-13 LINKAGE COEFFICIENTS: WIVES WEEKDAY ACTIVITIES SINGLE FAMILY SUBURBAN . . . . . . . . . . . LINKAGE COEFFICIENTS: WIVES WEEKDAY ACTIVITIES HIGH RISE SUBURBAN . . . . . . . . . . . . . LINKAGE COEFFICIENTS: WIVES WEEKDAY ACTIVITIES SINGLE FAMILY DOWNTOWN . . . . . . . . . . . LINKAGE COEFFICIENTS: WIVES WEEKDAY ACTIVITIES HIGH RISE DOWNTOWN . . . . . . . . . . . . . FACTOR STRUCTURE: SUBURBAN . . . . . . . . . o FACTOR STRUCTURE: DOWNTOWN . . . . . . . . . . FACTOR STRUCTURE: HIGH RISE APARTMENTS . . . . FACTOR STRUCTURE: SINGLE FAMILY HOMES . . . . FACTOR STRUCTURE: SINGLE FAMILY SUBURBAN . . . FACTOR STRUCTURE: SINGLE FAMILY DOWNTOWN . . . FACTOR STRUCTURE: HIGH RISE SUBURBAN . . . . . FACTOR STRUCTURE: HIGH RISE DOWNTOWN . . . . . FACTOR STRUCTURE: WIVES WEEKDAY ACTIVITIES - DOWNTOWN . . . . . . . . . . . . . . . . . . FACTOR STRUCTURE: WIVES WEEKDAY ACTIVITIES - SUBURBAN . . . . . . . . . . . . . . . . . . FACTOR STRUCTURE: WIVES WEEKDAY ACTIVITIES - HIGH RISE . . . . . . . . . . . . . . . . . FACTOR STRUCTURE: WIVES WEEKDAY ACTIVITIES - SINGLE FAMILY HOMES . . . . . . . . . . . . FACTOR STRUCTURE: HUSBANDS HIGH RISE . . . . . vii Page 155 156 157 158 190 191 192 193 194 195 196 197 198 199 200 201 202 Rule 5.14 FACTOR SIAUC C-lS FACTOR SIRS C Table Page C-IS FACTOR STRUCTURE: HUSBANDS DOWNTOWN . . . . . . . . . 204 viii Figure 2-1 3-1 4-1 4-2 5-1 5-2 6-2 6-3 LIST OF FIGURES RENTS FOR LAND USE CATEGORIES WITH DISTANCE FROM THE CENTRAL BUSINESS DISTRICT . . . . . . . . MAJOR URBAN LIFE STYLES AND RELATED STRATA . . A TRANSITION PROBABILITY MATRIX MODEL . . . . . SAMPLE TRANSITION PROBABILITY MATRIX . . . . . CATEGORIES OF THE PHYSICAL ENVIRONMENT . . . . STUDY ZONES IN METROPOLITAN TORONTO . . . . . . RESPONDENTS' PRE-MOVE LOCATIONS IN METROPOLITAN TORONTO . . . . . . . . . . . . . . . . . . . POST-MOVE LOCATIONS OF RESPONDENTS . . . . . . MAJOR APARTMENT CLUSTERS, 1967 . . . . . . . . ACTIVITY PATHS OF SELECTED COUPLE . . . . . . . ACTIVITIES OF SELECTED FAMILY WITH CHILD . . . TYPES OF COMPARISONS POSSIBLE BETWEEN LOCATION AND HOUS INC 0 O O O O O O O O O O O O O O O 0 ix Page 10 S8 77 78 90 92 93 94 112 121 122 124 Ethan society has decades, particularly been seen in rising, in change, the almost CO: tation form, rapid Su pathological conditicv fully aware. These changes m not an internally ho; seeiety I! I means 0 city has been invest Value: and forms . Merl 'u N either liv'v lutluence area of a‘ Chapter 1 INTRODUCTION Urban society has been undergoing profound changes in recent decades, particularly since Wbrld war II. In North America changes have been seen in rising incomes, higher productivity, rapid technological change, the almost complete dominance of the automobile as a transpor- tation form, rapid suburbanization, inner city renewal, and the many pathological conditions of urban society of which we are all too pain- fully aware. These changes have tended to emphasize the fact that the city is not an internally homogeneous entity which can be contrasted to rural society as a means of understanding social change. Traditionally, the city has been investigated from the vantage point of rural, traditional values and forms. With almost ninety percent of the population of North America either living in urban places, or residing within the effective influence area of an urban place, the traditional distinctions diminish in importance. Society is predominantly urban, and as diverse within urban areas as between urban and rural areas. The city, contrary to earlier conceptions, contains a heterogeneity, both socially and physi- cally, which merits the attention of scholars. This condition has not gone unrecognized by economists, sociologists, geographers, planners, and others. But only relatively recently have significant strides been made in understanding the interaction between spatial structure and social structure in the contemporary city. This I dissertation is an “I prevalent Within soci‘ seggraphic nobility IN More specificallg ficaat findings uhicn the hypothesis that 1 environment in metrop aents which are congr and they seek locatic This process helps tc SPfliil patterniNg o: This research h Wmnent. In the c introduced into urba m“ Previously. I; activity sequence p. sample of Toronto. ences in life 5t“, locations lnd diff: little “sad tiME-s ._ dissertation is an attempt to contribute to the dialogue currently prevalent within social geography regarding residential differentiation, geographic mobility processes, and behavior in urban environments. More specifically this study attempts to add to the already signi- ficant findings which have emerged from Social Area Analysis by testing the hypothesis that life style varies systematically with residential environment in metropolitan areas. People select residential environ- ments which are congruent with the image they possess of themselves, and they seek locations which contain families with similar ways of life. This process helps to account for the observable differences in the spatial patterning of social groups throughout the city. This research has both a conceptual component and an empirical component. In the case of the former, the concept of life style is introduced into urban social geographic research in a more formal manner than previously. The empirical element involves an analysis of the activity sequence patterns (as a surrogate measure of life style) of a sample of Toronto, Ontario residents in an attempt to sort out differ- ences in life styles among urban residents residing in different locations and different housing environments. Within this context the little used time-space budget is employed to secure activity data, and a decision-making framework is presented in order to understand the role of the individual as the fundamental unit of analysis. Within this larger framework, several specific objectives motivated the conduct and guided the methodology of this research effort. First, an attempt is made to present and elaborate upon the construct life style as an organizing framework for understanding residential differen- tiation in the city. In the past residential structure has been explained by reference to cc“— for urban space. 1 emerged as O compel use of life style i defined as an indiv scout I person's 1" (behavior is define Life Style act information about a life cycle, behavio; understanding residc explain the locatior of individual decisj neighborhoods occupi A second objec; Pittems 0f urban re tioned above, life s fro-construe: to m 8 new“! seemed to b 'CIIVlty epiSOde .3 by reference to economic competition among social groups and land uses for urban space. In recent years, the notion of social values has emerged as a competing explanation of urban residential structure. The use of life style is consistent with this latter tradition, and is here defined as an individual's predominant role orientation. Information about a person's life style can be recovered from his or her behavior (behavior is defined as the acting out of roles). Life style acts as a convenient summary construct which incorporates information about a person's social and economic status, stage in the life cycle, behavior, preferences, and personality. As a basis for understanding residential differentiation, life style can be used to explain the locational decisions of individual households. The aggregate of individual decisions results in an urban pattern of residential neighborhoods occupied by groups practicing different styles of living. A second objective of this study centers on the daily activity patterns of urban residents as a measure of their life styles. As men- tioned above, life style is a comprehensive construct. Bridging the gap from construct to measurement is a difficult task in this case. Human behavior seemed to be the best measurement category, with the individual activity episode as the fundamental behavioral unit. But what is im- portant is not so much the activity itself but the manner in which acti- vities connect with other activities over some definable unit of time. In light of this it became necessary to investigate the decision- making process as it relates specifically to activity sequencing to de- termine a summary measure of the likelihood of linking specific activities to one another. A third objective of this study seeks to determine the feasibility of mloyins transiti behavior on the P"lrt allow the identifica: eapnasize in their da the oasis of resident terns of dominant pat This study also haviorally, can be us distribution based or analE’sis has not beer 5 final objecti‘. to a concern for app] frueaork to the ques aMl'jseo ha Ve bec ome studies, criae atudi Petra of urban 80cia a55reéited d“, for a -,, my“: °f Urban so » cannot It eah II of employing transition probabilities as a measure of activity sequencing behavior on the part of urban residents. Transition probabilities allow the identification of sequences which particular groups tend to emphasize in their daily routines. Groups are identified a priori on the basis of residential location and housing type, and are compared in terms of dominant patterns of behavior. This study also seeks to determine if life styles, as defined be- haviorally, can be used to verify the spatial aspects of life style distribution based on other measures (e.g. census tract data). Such an analysis has not been attempted before in geography. A final objective, which incorporates all the previous ones, relates to a concern for applying behavioral data and a behavioral conceptual framework to the question of urban spatial structure. Although behavioral analyses have become more common in recent years (e.g., shopping trip studies, crime studies, intra-urban mobility) the study of spatial as- pects of urban social structure have relied primarily on the use of aggregated data for census tracts. This study follows the premise that analyses of urban society cannot ignore social behavior patterns, and further, cannot treat behavioral categories in isolation. The time- space budget approach incorporates the totality of behavior over some specified time period, and thus focuses on movement and activity in the most dynamic sense. In order to enhance our understanding of the dynamics of urban living and its implications for the future of urban society, social science must pursue all possible avenues of investigation, no matter how novel. In addition, disciplinary boundaries must be bridged in order to facilitate this process. In a recent provocative essay David Harvey noted that, There at sociological work in a 5P4 of a poserfui ness, who fii can have .1 pr the numerous aodern living it is into th approaches to prob contributing to ou process and spa tia. There are plenty of those possessed with a powerful sociological imagination who nevertheless seem to live and work in a spaceless world. There are also those, possessed of a powerful geographical imagination or spatial conscious- ness, who fail to recognize that the way space is fashioned can have a profound effect upon social processes - hence the numerous examples of beautiful but unlivable designs in modern living. (Harvey, 1973: 24) It is into this interface between the sociological and spatial approaches to problems that this dissertation moves, in the hope of contributing to our knowledge of the relationships between social process and spatial form. URL: Residential Struc tut; Vithin the gener and the forces which stream of thought ar fundamental dis tinc ti explaining urban spa The first ; focusing on the on the relative lent. The seCO' oriented, conce give rise to th on the decision The distinction Produce the obs ervec' :10: considered so cc oriented s tandpo in t terms of, " O I a th' Chapter 2 URBAN RESIDENTIAL STRUCTURE Residential Structure f he City Within the general frame of reference of urban spatial structure and the forces which generate the pattern of growth in cities, two main streams of thought are prevalent. Andrei Rogers (1967) identifies a fundamental distinction in the orientation with which the problem of explaining urban spatial structure is approached: The first point of view is macro and system oriented, focusing on the phenomena of urban growth and change and on the relative forces acting to promote orderly develop- ment. The second, on the other hand, is micro and process oriented, concerning itself rather with the factors that give rise to the phenomena by their collective influence on the decision making of societal units. (Rogers, 1967: 108) The distinction reduces to one of which forces are Operating to produce the observed spatial patterns, though the distinction is usually not considered so complete that no overlap exists. From the systems oriented standpoint the spatial structure of a city is interpreted in ". . . the dynamics of natural, social, and economic forces terms of, acting within an urban environment" (Rogers, 1967: 109). Usually, structure (i.e., the sorting out of various land uses) has been explained through the Operation of the land market, specifically vai the medium of rents. The process oriented approach emphasizes the "events" which produce urban growth and change, these events being the consequences of the activities of interacting decision makers. "The active role of human volition replaces societal adaptation to sapce as the principal 6 referent" (Rogers, 1967: 109). Heretofore, the most dominant explanation of urban spatial structure has been a land use theory that specifies a relationship be- tween land use and land value. Land values were responsible for the allocation of land uses to different areas of the city; accessibility of a site to all other locations being a prime determinant of that site's value. Since accessibility has been measured traditionally in terms of transport costs or physical distance, the relationship is postulated as being one of increasing land value with decreasing transport cost to the central business district. Economists still measure accessibility in these terms (Alonso, 1965). However, as Moriarty (1970) points out, while accessibility constitutes an important variable in the selection of land use for a particular site, it is by no means the only consider- ation employed in the selection of residential locations by urban residents. Quinn (1950) has observed that the greater the frequency of interaction between any land use activity and complementary activities, the greater will be that activity's tendency to maximize its accessi- bility to the complementary activities. However, while this postulate may hold true for non-residential land use activities, it cannot be said to describe the locational behavior of residential decision makers. Moriarty (1970) notes that residential decision makers display a geographical bias in both locational and travel behavior which, ". . . conflicts with the prescribed order" (Moriarty, 1970: 19). He further states that even though residentially based actors occupy different household locations from which interaction with different spatially distributed activities is conducted, location theorists still lump them into the same group with non-residential actors in terms of accessibility preferences. He concludes his criticism of the access- ibility model by stating: However, the complementary activities that non- residential decision makers seek to be accessible to (that is, their consumers) are not the same complementary activities to which residential decision makers seek to increase their accessibility. Residential decision makers seek to be accessible to those activities deemed necessary by them to satisfy their day to day needs. (Moriarty, 1970: 19) The contention here is that residential structure cannot be fully explained by models and theories which have been designed to account for the locational patterns of non-residential land uses. The different decision makers do not operate using similar preference structures. In fact, a recent study by Fales and Moses (1972) proposes that current land use theory better describes the structure and growth of nineteenth century cities, and new approaches are needed to explain present day patterns. They suggest that since a different set of processes have been operating in recent decades (particularly dispersal through improved transport technology) new variables must be entered into the postulates which attempt to explain spatial structure. This dissertation will attempt to introduce new concepts to help explain these new phenomena in urban areas. Specifically, it is con- tended that, since urban residents for the most part can ambulate through the city without severe constraints on their mobility, some old notions about the relationship between residential location and access- ibility to central urban functions no longer hold true. Instead, urban dwellers are free to consider locations and housing environments which meet the demands of whatever particular preference structure they embrace. Concepts such as social distance, differential access to leisure activities, sentimentality and life style enter the picture. People's housing choices reflect aspects of their values, needs, and drives more than their desire solely to minimize travel costs. This study will investigate one aspect of that situation. Competing Models of Residential Structure Residential differentiation within urban areas has generally been explained by two competing conceptual devices. These are termed the economic competition hypothesis, a system oriented approach, and the social choice hypothesis, a process oriented approach (Feldman and Tilly, 1960). Presently, land use allocation models incorporate the former postulate to the exclusion of the latter. The economic compe- tition hypothesis views the spatial distribution of all land uses as being governed by the process of competition for locations with fixed values (determined by accessibility to the central business district). All residential land users possess similar preferences but differ in their budget allocation and available resources. Ability to pay explains distance from the central core, ability being associated with quality, size, and location of households. Residential decision makers, according to this model, although they may possess similar housing preferences, compete for locations and are distributed on the basis of financial resources. The difference in ability to pay determines the spatial distribution of socio-economic groups in urban space (Moriarty, 1970). Figure 2-1 displays the juxtaposition of urban land uses as the economic competition model would predict them. Land uses sort themselves out on the basis of bid rent curves. The activity with the highest and steepest rent curve for any parti- cular location would occupy that site. Figure 2-1 shows the rent curves 10 Ceiling rents per acre Offices, banks, stores, other businesses A dependent upon accessibility Manufacturing and warehousing Housing Manufacturing Agricultural uses Distance from market Source: Nourse (1968) p. 115. Figure 2-1 RENTS FOR LAND USE CATEGORIES WITH DISTANCE FROM THE CENTRAL BUSINESS DISTRICT 11 for the four major land use types (commercial, manufacturing, residen- tial, and agriculture) and their optimal spatial distribution. Within this larger context a housing subsystem exists based upon processes of economic competition, wherein the spatial distribution of housing types is dictated by rent bids. "The slope of the rent bid curve depends upon the decreased price per unit of housing as distance from the market increases, the amount of house per unit of land, and on the substitution of land for nonland inputs as rent decreases" (Nourse, 1968: 114). As a consequence of this hypothesis high rise apartments would be located exclusively near the center, and with increasing distance away and declining rents housing types would shift to duplexes, thence to single family homes, and finally single family homes on large acreage (Nourse, 1968). The social choice hypothesis holds that residential locations are chosen on the basis of differential preferences or behavioral considerations. This position views urban space as a reflection or indicator of social values which are governed by sentimental, non- economic, or cultural factors (Feldman and Tilly, 1960). Status significance, ways of life, and personal motives of residential change are the major concerns which determine residential differentiation. Residential decision makers are not so much indifferent to location, but are greatly influenced by the character of established residential areas. The selection of a household involves a conscious residential choice by decision makers who vary in their preferences because of different values, needs, and desires (Moriarty, 1970). It is believed that the latter explanatory system offers a more fruitful framework for condhcting research on urban residential 12 structure. Previous research has demonstrated the inconsistencies which prevail between residential location behavior and the economic competition hypothesis. Both Duncan and Duncan (1955) and Feldman and Tilly (1960) observed that white collar workers and blue collar workers at the same levels of income display different residential patterns. Other preferences, such as social distance may be responsible for the difference instead of a sub-social force such as competition for access- ible sites. Witness also the existence of a cultural artifact such as Boston's Beacon Hill residential section occupying a central city location which should contain high rise office buildings and other commercial establishments according to the economic competition hypothesis (Firey, 1945). Another contradiction to the notion of economic competition producing an orderly arrangement of residential land uses is the 'dispersal of such high density activities as high rise apartment buildings, garden apartments, and townhouses (Fales and Moses, 1972). For the past two decades the character of the housing stock in Metropolitan Toronto has been shifting such that the ratio of one and two family buildings has been declining. Kumove (1966) has noted that: Apartments comprise over 50 percent of all new buildings built in 1958, and over 60 percent of those built since 1961. By the end of 1965, about 70 percent of all new housing were apartments. The vast majority of these apartments were con- structed on vacant land in the suburban municipalities. (Kumove, 1966: 5) Earlier, Hoyt (1939) proposed that the initial location and growth of the high rent areas of most North American cities directly influenced the location of medium and low rent areas. The wealthy or upper class residents located their homes on radial transportation 13 routes emanating from the city center or on, ". . . high ground which is free from the risk of floods and to spread along lake, bay, river, and ocean fronts, where such water fronts are not used for industry" (Hoyt, 1939: 117). The middle range rental areas were situated usually on either side of the high rent sector, with lower rent sec- tors located on either side of those. This would seem to indicate that social distance considerations may predominate over more economic factors in the location decisions of most urbanites. Obviously, considerations other than accessibility to comple- mentary land use activities (although accessibility remains important) must be weighed when constructing models of urban residential struc- ture. In the tack taken above an attempt has been made to elaborate upon the relative power of two competing hypotheses to account for urban residential differentiation. The conclusion to be drawn is that, at the metropolitan level, the forces of economic competition account for the general spatial arrangement of land uses, but that within the residential or housing subsystem other considerations, namely social ones, account for the spatial variation of residential neighborhoods. This dissertation proceeds from this social values framework, but utilizes it more as a guiding principle than as a working hypothesis. The proposition that differences in behavior correspond to differences in residential environment, as defined by location and housing type, will serve as the general question to be addressed. Different behavior patterns are seen as resulting from variation in life style preferences, which in turn are manifestations of the value orientations of urban residents. 1h Urban Residential Models Most students of the urban scene proceed under the assumption that the urban community is neither an undifferentiated mass nor a haphazard collection of buildings and people (Timms, 1971). There seem to be fundamental processes operating which give order and pattern to the physical and social fabric of the city. This is no less true of residential differentiation than it is of the differen- tiation of other forms of urban land use. In fact, since residential land use occupies the largest percentage of space in metropolitan areas, social scientists have devoted considerable effort to identi- fying the processes by which urban residential areas become differen- tiated from each other along income, family status, ethnic, and other lines. Louis Wirth (1938) once said that the residential differen- tiation of the city resembles a mosaic of social worlds. That is, similar groups of people cluster tagether and come to characterize their area. But the means by which this mosaic form has been interpreted in different ways by urban scholars. During the early decades of the twentieth century a group of sociologists at the University of Chicago, under the leadership of Robert E. Park and Ernest W. Burgess, sought to understand the growth and structure of the city from an ecological perspective. They placed emphasis on the subsocial forces which shaped the residential and social structure of urban areas, and they postulated a process of economic competition, or the sorting out of groups on the basis of income. The processes of invasion, succession, and dominance derived their energy from the expansion of the city's population and area through time. The constant turnover of population at the urban 15 core provided the stimulus for growth at the periphery. Those groups with the longest tenure in the city were also the most upwardly mobile. With an increase in economic status they moved to newer homes in the peripheral zones while groups new to the city (usually EurOpean immigrants) and of a lower economic class would replace them. At a very general level Burgess's concentric zone model of urban growth reflected this sorting process on the basis of income and tenure. Hoyt (1939) rejected the concentric zone model in favor of a sectoral arrangement of land uses. He proposed that high and low rent neighborhoods occupied distinct subareas of the city. The spatial pattern of rental areas was determined by those who could afford the highest rents. Higher status areas spread out along major transportation routes and higher ground, while interstitial areas came to be occupied by lower status residences. The sectoral model was also based on an assumption of economic determinism; it merely postulated a different spatial arrangement of residential land use. Also built into it was a social distance element, whereby each socioeconomic group would locate next to areas of the next highest group. In this fashion the poor would be located as far away from the rich as possible. Finally Harris and Ullman (1945) proposed a picture of the city where land uses concentrated around separate nuclei as opposed to a single core area. Until the 1950's one line of research that social scientists pursued involved the evaluating of these three competing models of urban residential structure. Research results conflicted with some expec- tations and it was concluded that they were overly simplistic and suffered from an absence of clearly stated assumptions. It took the advent of Social Area Analysis (Shevky and Bell, 1955) and its subsequent 16 derivative methodology Factorial Ecology to finally unravel the web of conflicts which had arisen. The Social Area Analysis model includes three basic constructs which describe the way in which urban populations are differentiated. They are economic status (social rank), family status (urbanization), and ethnic status (segregation). An examination of these dimensions would lead to a more complete picture of how residential areas were differentiated. As originally conceived, the Social Area method constituted an indexing system and typology. Only later did the authors attempt to establish a theoretical basis for the differentiation that occurred (Shevky and Bell, 1955). Although Shevky and Bell demonstrated the use of the Social Area typology they were interested in a structural rather than a spatial representation of social (residential) space. The task of analyzing spatial differentiation was left to others. The methodology most often employed to test the classical spatial models of urban growth and structure has been summarized by Murdie: A sample of census tracts was selected from a predefined grid of sectors and zones using slightly different criteria in each study. Social area indexes were calculated for each tract and an analysis of variance design was used to deter- mine the relative significance of the spatial models in describing the location of each index. (Murdie, 1969: 25) One weakness in the above approaches is their failure to incor- porate anything other than aggregate census tract data in the analysis. They have provided us, however, with sound empirical generalizations upon which to base other forms of investigation into the structure and dynamics of urban residential differentiation. Murdie (1969) categorizes studies employing Social Area Analysis as focusing on formal characteristics of social space. At least two 17 other characteristics, the functional and the circulatory, are required in any comprehensive social geography of a metropolitan area. Webber (1963) points up the limitations of the purely structural approach in another manner: The kinds of information that can be read from maps showing urbanized areas or land use patterns are therefore likely to be misleading . . . maps of this sort miss the essential meaning of urbanization. Whether the maps rep- resent existing patterns or plans for future patterns, they present static snapshots of locational patterns of people or buildings or activity places and say nothing about the human interaction patterns that are the heart of complex social processes. (Webber, 1963: 49) He further argues that the basis for rational planning should reside in the complex processes of interaction which characterize urban areas over time. Meier (1962) proposes that urbanism and urbanization are both measurable in terms of communication transactions and information flow. In essence, he advocates a theory of the city based upon intensification of interaction and operationally measurable by 'bits' of information passed. By contrast with rural dwellers, who characteristically lead spatially and temporally homogeneous lives, urbanites experience great variability, not only in the selection of activities, but also in the place, time, and duration of their exercise (Meier, 1962). It is the interactional aspects of urban dwellers lives, and not their social, economic, and demographic characteristics, that provide the focus of inquiry in this dissertation. This study will investigate the behavior patterns of various residentially defined groups. It proceeds from a realization that the ecological approach to residential differentiation has already made its most significant contribution. With the concentric zonal, sectoral, and multiple nuclei conflict having been settled by Social Area Analysis 18 and Factorial Ecology we now have some valuable empirical generalizations about the aggregate social, economic, and ethnic structure of the modern western city. The time now seems propitious to seize upon the sugges- tions of Meier (1962), Webber (1964), and others to move to a different level of scale in the search for a better understanding of urban organ- ization. This change entails a shifting from the study of attributes of urban populations to research into their everyday behavior. Human Activity Systems In recent years, scholars and planners have moved from purely structural studies of urban growth and form to investigations which contain a behavioral component to them. That is, many researchers have come to realize the value of studying the behavior of urban dwellers as they interact with their environments. Social scientists are interested in the reasons behind human behavior so that such processes as learning and decision making can be better understood. Social and physical planners, on the other hand, study behavior with more pragmatic objectives in mind. They seek to map and understand behavior so that planning projects can be designed which are responsive to the needs of those individuals who live and interact in the city. The study of interaction has superceded the study of structure because purely structural models of the city are of a low level of explanation. Social scientists and planners in recent years have proposed a human activity systems framework for studying urban spatial structure. This dissertation represents an attempt to fuse the study of human activity systems together with the more traditional inquiry into residential differentiation by investigating the ways in which people 19 structure their daily activities and by relating those differences to variation in residential location and environment. If different regions of the metropolitan area can be characterized by different economic, social, and ethnic structures, then they may also display differences in the interaction patterns of their residents. It is the purpose of this dissertation to make an exploratory incursion into the relation- ship between environment and behavior. Chapin and Hightower view urban activity systems as, ”. . . made up of patterns of human or institutional interaction associated with the functions of the urban center" (Chapin and Hightower, 1966: 4). They see the social system as a basis for studying human interaction, with the household, or family, forming the units of observation. In addition to the social system there are at least two other systems arising from and organizing basic urban functions: (1) The economic system where interaction occurs as a result of production and distribu- tion activities, and (2) An institutional system (or series of sub- systems) resulting from the activities of governmental, religious, political, labor, and other institutions. Chapin and Hightower see the need for research on household activity systems because: This concerns the living patterns of urban residents and the attitudes that influence their use of city space --how they carry on their daily, weekly, and seasonal activities and how well adapted the structure and form of the city is to these living patterns. The urban planner can no longer rely on intuition in these matters. He needs to bring to bear a whole new area of analysis which precedes the use of location and transportation models now being introduced into planning practice. (Chapin and Hightower, 1966: 2) Along these same lines others have suggested that problems of metropolitan growth require theoretical statements of spatial agglomer- ation based upon premises which differ from those of traditional n theory (C911 Locatie an approach founded < assets of tertiary for testing the impa: ace Cullen, God 125: allxation models as beta-nor patterns is figs, 1961) generat relationship between distributed relative In no case usually occupie in as many as t school in yet a near the wife's the action spac Which the resid “I“? | workpl attion space st of the family. In this same re .he study of reside?! concludes that at th C 01 the location of -' 35 . the neic‘snborhood er : In his preside 365132131 Science A. ilodel of daily bk was to explore new '1 20 location theory (Cullen and Nichols, 1971). These researchers offer an approach founded on micro-analysis of the behavior patterns of members of tertiary institutions and the development of models suitable for testing the impact of decentralizing such organizations. In another instance Cullen, Godson, and Major (1971) point to the shortcomings of allocation models as an indication that a more comprehensive view of behavior patterns is needed. Most models (Lowry, 1964; Alonso, 1966; Wingo, 1961) generate residential distributions on the basis of a simple relationship between residence and employment. Services are then distributed relative to the residential pattern. In no case is account taken of the fact that homes are usually occupied by families, the members of which may work in as many as three or four different places and go to school in yet another place. Shopping may have to be done near the wife's workplace or the primary school. In short, the action spaces of the household, in the framework of which the residential location decision is made, is not merely a workplace - home dumbell but a complex joint action space structured by the commitments of all the members of the family. (Cullen, Godson, & Major, 1971: 3) In this same respect, Whitelaw (1972) has attempted to classify the study of residential mobility according to levels of scale. He concludes that at the metrOpolitan wide level of analysis, consideration of the location of workplace looms large, but at a subarea scale such as the neighborhood the gamut of activities and accessibility to services and amenities become important to the location decision. In his presidential address to the European Meetings of the Regional Science Association in 1969 Torsten Hagerstrand articulated a model of daily behavior based upon time-space paths. His purpose was to explore new theoretical approaches to the relationship, . . . between the micro-situation of the individual and the large stale 85:39?“ outc theact'vities of pe ation activities are The purpose of the direction in whi saving. How human a on new importance ir. tion theory in accou dissertation follows hwever, several 33:.- worn on activity 3..” the decision making acquencing. Ac tivi t lit: “5“ “on; Vario 9L. Patten . , 0f acU A fit? 8 equ enc 1' dit; at“. Y\ It 1 s the Cor 21 scale aggregate outcome" (Hagerstrand, 1970: 9). His model focuses on the activities of people and the time and space constraints within which activities are carried out. The purpose of the preceeding statements has been to indicate the direction in which some significant research on urban structure is moving. How human activities are organized in time and space has taken on new importance in light of the inadequacies of traditional loca- tion theory in accounting for urban residential structure. This dissertation follows in this human activities tradition. There are, however, several ways in which this study differs from much recent work on activity systems. This investigation will not attempt to model the decision making process or the rules governing activity choice and sequencing. Activities will be analyzed from the point of view of linkages among various types of activity categories. It will be assumed that a logical connection exists between empirically identifiable patterns of activity sequences and the conceptual term, life style. The intent here is to relate the relatively unexplored phenomenon of activity sequencing to the more established concept of life style, and to attempt to relate differences in life style to questions of residential differentiation. Others have suggested this activities- life style link (Chapin and Hammer, 1972) but as yet no empirical research has been attempted along these lines. The unique contribution of this dissertation will be that of relating life style to residential differences via the medium of activity systems. Study Design It is the contention here that the particular types of resi- dential environments found in the city will have associated with them 22 certain patterns of behavior. People select residential neighborhoods and individual dwellings in which to live for a variety of reasons, such as social status, life cycle considerations, space considerations, location (accessibility to downtown or to open spaces, etc.), income changes, and others (Rossi, 1955). Running through this list of reasons is the desire on the part of individuals and families to find a resi- dential environment which is congruent with their preferred way of living (Michelson, 1970). This dissertation will attempt to investigate one aspect of this life style-residential environment relationship. Namely, are there patterns of behavior which seem to be characteristic of certain environmental situations, and can these differences in behavior be explained by those environmental situations? In order to achieve this objective, residential environments will be defined in terms of location within the urban area and in terms of dwelling type lived in. The concept of human behavior will be approached from a relatively new perspective, that of the human time budget. Traditionally, the behavior of variously defined groups living in the urban area has been identified in terms of membership in organizations, political activity, neighboring, employment, recreation, child care, and the like. The time budget, by contrast, with its recording of 511 the activities undertaken by an individual over some specified period of time, represents a more comprehensive approach to the identification of behavior patterns. It focuses on the micro- behavioral units of individuals and groups rather than on rates of participation and other aggregate forms of behavior. The time budget allows the researcher to make finer distinctions between and among groups regarding the activities they emphasize in 23 their daily lives, the allocation of those activities to temporal units, the cyclical nature of activities, and their spatial manifesta- tions. Within groups defined by such characteristics as social class, family status (or life cycle state), education, income, etc., the analysis of time budgets may lead to the discovery of preferences for ways of life on the part of subgroups. In order to accomplish the objectives of this study the time space budgets of a group of families residing in the Toronto, Ontario Metropolitan area will be studied for regularities in their behavior. In addition to investigation of time budgets while holding the residential environment constant, controls are placed upon the social class and life cycle stage of the subjects. By limiting the study group to one broad social class as defined by income (middle to upper middle class) and two life cycle categories (married couples in the childbearing years both with and without children) it is hoped that the contaminating influence of such variables can be minimized. In this way, differences in patterns of behavior which may emerge can be attributed to the life style preferences of the individuals involved. A further control will be placed upon sex, since this has been considered to be an influencing factor in the assessment of differences in behavior. Women and men take on different roles quite often, and since the role differences are quite pronounced from city to suburb and from housing environment to housing environment, the ensuing analysis will take sex into consideration. More importantly, in a later chapter, life style will be conceptualized as a dominant role orientation. Therefore, differences in sex (which orient roles based upon social organization) must be accounted for. 2h Behavior, in this instance the activities people pursue, will be analyzed and described by answering the questions: how do individuals link their activities over the course of a day? Are certain patterns of activity linkages characteristic of groups living in particular housing types and/or locations within the metropolis? Which activity types tend to be more closely linked to each other for which groups of urban dwellers? Can these linkage patterns be connected to particular aspects of the various environmental conditions in the city? For example, it is believed that individuals select downtown high rise apartments over suburban single family homes in order to pursue life styles which emphasize cultural activities, entertainment, and work related leisure. Therefore, the manner in which such individuals link their activities in time and space should differ in a fundamental way from that of their suburban family oriented counterparts. The means by which activity linkages will be summarized and modeled in this dissertation was first suggested by Hemmens (1966) and later refined by Brail (1969). The transition probabilities matrix of Markov Chain Analysis will be the initial grouping procedure used, whereby the probability of movement from one activity to another will be calculated from matrices of frequencies for various subgroups of the sample. This dissertation takes a methodological approach which differs from earlier attempts at modeling activity linkages. First, earlier analyses have been confined to trip data gathered in transportation surveys. Such data are highly restrictive in that they consist only of out-of-home trips and are coded by land use at origin and desti- nation and not by activity type. Secondly, with the exception of 25 Horton and Wagner (1968), no other social scientists known to this author have examined activity linkages for different groups residing in the city. The focus Of this dissertation, then, is on the activities which peOple living in different environments select to follow over the course Of a day's time. It is not the intention Of this research to Offer activity linkages as an Operational definition Of life style. It is Obvious that sO general a concept could not be adequately defined by so specific a measurement as a transition probability. Nevertheless, a knowledge of how people link their daily activities can provide a descriptive measure Of one aspect of life style, namely behavior patterns. This dissertation will not only attempt to utilize this approach, but will provide an assessment Of the activity linkage methodology as a tool for relating differences in life style to residential differentiation in urban areas. To this end, this dissertation has been divided into the following seven chapters and several appendices. Chapter Three provides a review of the construct life style. It proceeds from an historical recapitu- lation Of the concept as it has been defined in Sociology to an analysis Of life style as it has been incorporated into urban differentiation. Finally, a conceptualization Of life style is Offered which makes it useful as a framework within which tO cast this study. Chapter Four establishes the behavioral basis of activity systems Of human beings. It provides an exploration of the most significant and recent thought on activity research and attempts to define the conditions under which activity decisions are made. Further, the chapter outlines several conceptual schemes for considering decision mains in 8“ ‘CL In the gift? and study area» Selected: a lens smile data to t the demosraphic’ Also included is Ioronto 3‘ a“ 8.2 Finally, ‘ histfi adiition, the 1“ described. The Chapter Six hypotheses which includes a descr analyzed and sum? An evaluatic vided in Chapter suitable for rese. of sctivity links, I conclusion, dire and the prospect 0‘ 0f study will be 0‘ 26 making in an activities framework. In the fifth chapter the data are described along with the sample and study area. Because of the unique manner in which the sample was selected, a lengthy section is devoted to the applicability of the sample data to this particular study. Information is provided as to the demographic, social, and economic conditions of the respondents. Also included is a brief justification for the selection of Metropolitan Toronto as an appropriate urban area within which to carry out research. Finally, a history and description of time budgets are given. In addition, the instrument by which the activity data were secured is described. The instrument itself is included as an appendix. Chapter Six presents an analysis of the data and offers several hypotheses which are evaluated in light of the analysis. The chapter includes a description of the methodology by which the data were analyzed and summarized. An evaluation of the method and a summary of the study are pro- vided in Chapter Seven. The utility of life style as a concept suitable for research on residential differentiation, and the value of activity linkage as an analytical device will both be assessed. As a conclusion, directions for additional research in this area of study, and the prospect of time budget analysis as a viable and useful method of study will be outlined. Chapter 3 LIFE STYLE AND LOCATION: A CONCEPTUAL MODEL Social Bases 2£_Li£g,§gylg The previous chapter introduced the notion of social choice as possible explanation of the process of residential mobility of urban dwellers. Although this study concerns itself with the general subject of urban residential structure, it is not specifically investigating the residential mobility decision. Rather the objective is to postulate a relationship between residential environment and behavior patterns, to test for the existence of such a relationship using empirical data, and, by inference, to advance statements about the determinants of the observed relationship. In order to accomplish this, the study will be cast in a theoretical framework which will provide direction, both in terms of the specification of definitions and also in the formulation of hypotheses and elaboration of a methodology. It is suggested that the conceptual device which provides both a literature and frame of refer- ence for the study of residential differentiation is that which is referred to both scientifically and pOpularly as life style. Embodied in the term life style is the notion of differentiation. A style of life is a way of living, and by implication, ways of living vary across society. Current trends in the structure of urban populations, in particular the growth of the middle class, the rise in real incomes, suburbanization, mass education, and the elimination of illiteracy, indicate that the old criteria of differentiation are losing their 27 28 applicability. Groups are becoming more sharply defined according to differences in styles of living (Greer, 1962). Urban pOpulations, then, can be ordered along a style of life dimension. This does not imply that the objective of this study is to present a typology or classification of life styles. The concept will be used instead to present a general frame of reference to which patterns of behavior will be related. The purpose of this chapter will be to review the theoretical status of life style in the literature of social science and to conceptualize life style in such a manner that it will be useful for the examination of urban activity patterns. Social scientists at least as far back as Weber have incorporated notions of living styles in their theorizing about stratification systems. According to Weber, similarities in both education and occu- pation among groups of persons provide the antecedents of given common life styles. Classes are stratified according to their relations to material production and acquisition, but status groups are stratified according to consumptive patterns as represented by special styles of life. Social status stratification becomes more pronounced during periods of economic stability, and replaces economic class as the basis of stratification. As a result, class status tends toward homogeneity, resulting in an emphasis on stylization of consumption and social activity. Therefore, life styles proliferate and, ". . . class dis- tinctions are repressed by differentiation by life style rather than by stratification by (class) status" (Gerth and Mills, 1958: 193-94). In other words, Weber elaborated a theory stating that, as a majority of a society's members gain entry into the affluent middle class, status groups emerge which stratify themselves according to consumptive patterns 29 representing special styles of life. Certainly, modern society can be said to be moving toward styli- zation by consumption patterns. While classes are broadly categorized as upper, middle, and lower, most persons in the deve10ped world are favorably situated with respect to the production functions of society. How the rewards of production are distributed and utilized becomes more an important differentiator as the modern affluence society approaches. Two individuals may receive equal rewards for their contri- bution to the productive sector of the economy, but they may opt to utilize these rewards in divergent ways. It is at this point in societal develOpment that variation in life style becomes more pro- nounced. Life style as a theoretical construct has been examined most thoroughly within the confines of sociolOgy and social psychology. The literature, although voluminous, remains somewhat vague in terms of formal theory, propositions, hypotheses, definitions, and typologies. Conceptual designs are diverse, ranging from personality and psychological definitions (Ansbacher, 1967) to global statements about entire societies (T3hnies, 1967). The conceptual framework which seems to have the most elaborate formulation of a theory of life style is Social Area Analysis (Shevky and Bell, 1955), which rests on a theoretical foundation of increasing societal scale. It will be demonstrated how this theory of societal scale has been employed to elaborate the notion of increasing diversity of life style through the mechanism of social choice. Godfrey and Monica Wilson (1945) were among the earliest to propose increasing scale as a perspective from which to view social and 30 economic differentiation in societies. According to the Wilson's, ". . . by the scale of a society (is meant) the number of people in relation and the intensity of these relations" (Wilson and Wilson, 1945: 25). The fundamental distinction between traditional primitive societies and more modern ones is that of scale. As society increases from small to large scale there occurs a series of changes in the patterns of functional differentiation, the complexity of organization, and in the range and intensity of relations (Timms, 1971: 125). Drawing heavily upon the works of Colin Clark, particularly on the concomitants of economic growth, and on Louis Wirth's concept of urbanism as a way of life, Shevky and Bell established an ex post facto theoretical basis for Social Area Analysis, that of increasing societal scale. In the Shevky-Bell theory increasing scale is synonymous with the emergence of urban-industrial society. Society can be differentiated more easily as a result of ever increasing scale (McElrath, 1968). Three conditions are postulated as a result of increases in scale: (1) changes in the range and intensity of relations, (2) a greater differentiation of function, and (3) a growing complexity of organization (Robson, 1969). Three broad structural trends result from such changes, and account for the distribution of attributes upon which social differentiation is based. These trends are reflected in the changing distribution of skills, changes in the organization of productive activity, and changes in the composition of pOpulation with greater movement and increasing diversity (McElrath, 1968). The manner in which these changes can be observed and measured gives rise to three constructs and their associated indicants (or indexes). Changes in the distribution of skills leads to changes in 31 the arrangements of occupations. This can be measured through the construct, social rank. The second construct, labeled urbanization, reflects changes in the distribution of function. Such changes occur in the structure of productive activity, the growing importance of cities, and in the diminishing importance of the household as an economic unit. Concurrently, there occurs a movement of women into urban occu- pations and a proliferation of alternative family patterns. Finally, the growing complexity of organization produces a redistribution of population in space and the isolation and segregation of groups. At the structural level these changes are reflected in the construct termed segregation (Robson, 1969). In his revision of the social area model, McElrath (1968) added a fourth major trend occurring as a result of increasing societal scale, that of the aggregation of population. Such aggregation, particularly the concentration of population in urban areas, mirrors increasing social differentiation by migrant status. In addition, McElrath posits these four structural changes in society as being derivable from two processes: industrialization and urbanization. Urbanization and industrialization, then, yield four basic dimensions of social differentiation along which the rewards and resources of urban communities are distributed. In the (modern) city the range of opportunities available to an individual or family is subject to the multiple con- straints of economic status (based on skills); family status (based on life style option); migration status (based on migration experiences); and ethnic status (based on social visibility). (MeElrath, 1965: 104). As indicated by the preceeding quote, a relationship exists between family status (urbanization) and style of life; the fact that an option prevails introduces further the concept of free social choice in large scale urban-industrial societies. In order to understand the 32 relationship between the family status construct of Social Area Analysis and life style, some elaboration is necessary. In its most general terms, the relationship can be expressed in a proposition of the form: as a society increases in scale the greater is the social differentiation according to family status, and the greater the variety of life styles. No causal link is suggested; only a strong association. Shevky, Bell, and others, however, assume a causal connection, as elucidated below. Three aspects of social life are purported to be influenced by the changing nature of production: (1) the relationship between popu- lation and economy, (2) the structure and function of kinship units, and (3) the range of social relations concentrated in the city. Each of these in turn reflects the construct Shevky termed urbanization. With the population freed from Malthusian constraints and with the widespread adoption of contraception, family size becomes a matter of individual decision and presents the choice between family and career mobility. People may breed to the limits of the available food supply or invest in property, career, etc. Similarly, the family ceases to act as a production unit, causing a realignment of kinship functions. Differences in family structure, rather than reflections of social status differentials, mirror the choice between alternative forms of life. The individual is faced with the Opportunity of selecting from alternative life styles (Timms, 1971: 130). The family status construct, then, represents a continuum along which life styles are arrayed. Increasing scale causes a differentiation of society in the form of alternative styles of life. 33 The Theory 2; Social Choice Wendell Bell (1958, 1968) has provided an articulation of the societal scale theory by proposing that social choice, through the construct life style, explains the social and residential differentiation extant in urban society. This approach is more in line with McElrath's (1968) revision of the social area model, where he asserts that residen- tial differentiation is a function of social differentiation, which in turn is coordinated with certain systematic changes accompanying social deveIOpment. These changes are assumed under the general process labeled modernization. Although one of the originators of Social Area Analysis, Bell rejected the theoretical stance maintained by Shevky, and formulated different labels for the operational constructs derivable from the theory. 0f the three, urbanization, or family status as Bell suggested, proved to be the center of considerable disagreement. Those measures related to urbanization: age and sex characteristics, type of tenancy, and house structure, and more importantly, the indicants used in the calculation of the index, according to Bell, refer to family related phenomena. Family status does not measure degree of urbanization, but rather the indicants suggest a direct measurement of a set of value orientations of familism, careers, or consumption (Bell, 1958). No assumption should or could be made that populations high in family status are necessarily low in urbanization. Such an hypothesis is inconsistent not only with the theory of which it is a part, but also with reality. Familism and non-familism (or urbanism) are concomitant developments of urbanization through the intervening mechanism of social choice. 3b In outlining his theory, Bell advanced several propositions re- lating to the degree of modernization a society achieves and the character of its social system. With modern methods of contraception available in urban-industrial society, family size and birth rate rep- resent manifestations of the sum of many individual decisions by people concerning numbers of children. Such decisions are neither randomly made nor are they a function of individual differences in personality. They represent systematic preference patterns or "social choices" which are fashioned by the character of the social system and shaped by value systems (Bell, 1968). In effect, the greater the degree of modernization a society achieves, the greater the freedom from the constraints of food supply, disease, and overpopulation, the greater the range of available social choices. As a direct result of this social choice hypothesis he postulated a range of preference patterns which includes familism, upward vertical mobility (careerism), and consumership. Bell (1968) observes that greater per capita wealth, leisure, and energy in advanced societies is being spent on children and other aspects of family living, as well as other alternatives. Family living as a distinct and desired activity is a contemporary phenomenon, and places a high valuation on family living, marriage at young ages, a short childless time span after marriage, child-centeredness, and re- lated characteristics. The familistic preference pattern can also be equated with certain norms of consumption, an interwoven set of activities and possessions. The upward mobility pattern is characterized by spending time, money, and energy on one's career. The upwardly mobile person engages 35 in career relevent activities at the expense of alternative activities. As a result, sociOIOgists posit an inverse relationship between familism and upward mobility, and spending time and money on career may limit family life by delaying marriage and postponing children (Bell, 1968). Research indicates that the upwardly mobile make little distinction between work related and leisure activities. These career oriented peOple increasingly view their leisure activities as instrumental to advancement. Social life is used to promote contacts with colleagues and professional clients. Family oriented persons tend to separate these two lives; social contacts and leisure have no connection with employment (Willmott, 1969). The consumership life style eschews both familism and careerism for as high a level of living as possible in the present. Money, time, and energy are expended on "having a good time," "living it up," or "enjoying life as much as possible" (Bell, 1968). These individuals develop "consumption as a way of life" (Greer, 1962). This is exper- ienced in a manner unconnected with family or career goals. These life styles are by no means class bound, nor does one lock himself into a way of life throughout his life time. Although some do this, others are able to combine in their life style elements of each preference pattern at various periods during their lives. For the majority some choice is possible, and most place differential emphasis on the different choice patterns as various stages in their life cycle. Although the foregoing possesses some of the elements of a logical coherent system at the higher levels, it begins to break down at the level of the postulated preference patterns. It does not necessarily follow that the life styles associated with familism, careerism, and 36 consumerism are logical derivatives of social choice. These constructs do not possess the independence necessary to relate them to a higher order statement such as social choice. In addition, they are defined in the vaguest of terms, and in fact, tend to be defined in terms of attributes rather than as a set of dispositions to act. The preferences inherent in their definitions are inferred rather than logically de- duced. In addition, the three dimensions are too highly intercorrelated to be of much value as differentiators of life styles. At a gross, metropolitan wide level of analysis employing census data, the familism construct reveals basic differences in family status and therefore describes a generalized social structure. However, no comparable data exist for examination of other preference patterns. Therefore, no uniformity could be achieved in empirical analysis at aggregate levels, and no comparable data collected from survey instru- ments are available for review. Greer (1962) and McElrath (1968) also subscribe to the social choice approach to social differentiation. Greer offers some partic- ularly interesting propositions relating to life style as an axis of social differentiation. He sees the entire social system of the metro- politan complex travelling along three separate dimensions of change. By examining these dimensions the researcher can order and compare different neighborhoods, different cities, or the same city at different points in time. These dimensions are: social rank, life style, and ethnicity. The formerly high correlation among familism, ethnicity and social rank is breaking down. Declining segregation by social rank and ethnic identity accompanies increasing concentration by life style. "As life style becomes freer from the limits of occupation and ethnic 37 origin, it becomes a more significant differentiator of the population" (Greer, 1962: 76). As a result of these processes says Greer, urban populations can be located along a style of life continuum which represents a commit- ment to familism. At one extreme lies the urban style of life, where families are small, and husband and wife generally both work (many are also unmarried). At the other extreme we find the devotees of familism. These people live in the vast middle range of social rank, raise larger families, and the wife usually remains at home neighboring and providing intensive child care (McElrath, 1968). Spatial Aspects gf Life Style A common thread running through most research into styles of living at the macro level, especially among urban sociologists, has been the postulation of a relationship between preferred life style and residential location. As a consquence, research questions are being asked, such as: Do individuals decide where to move on the basis of life style aspirations? Do changes in ways of life (perhaps precipitated by life cycle changes) induce residential mobility, and vice-versa? What relationships exist between housing environment and life style, neighborhood environment and life style? According to the social choice hypothesis the manner in which groups sort themselves out in urban regions results from the individual decisions of residential location seekers. This sorting out process, formerly on the basis of class, social status, or ethnic differentials, now occurs as a result of more subtle differences in preferred life styles. Although the range of life styles is great, variation in life 38 style is not correlated strongly with variation in social rank. At each level of economic class the range in life style varies considerably (Greer, 1962). Therefore, status differentiation and status enhancement, as manifested through life style, are likely to lead to emphasis in small differences between one residential area and another (Pahl, 1970). The recent work of urban sociologists and urban geographers represents an attempt to link life style with residential location. Bell (1958, 1968) postulated the existence of a direct relationship between an emphasis on familism and the growth of suburbs. The move to the suburbs expresses an attempt on the part of the movers to find a location in which to conduct family life that is more suitable than that offered by central cities, i.e. that persons moving to the suburbs are principally those who have chosen familism as an important element of their life styles as over against career or consumership (Bell, 1968: 151). Bell (1968) tested this hypothesis for a sample of recent movers in the Chicago area and found that upward vertical mobility is not associated with choosing to live in the suburbs, whereas familism is. According to Greer, "The utility of different parts of the metro- polis for different styles of life results in a concentration of similar persons with similar needs in given neighborhoods" (Greer, 1962: 95). The career oriented style of living is more pronounced in the apartment areas of central cities, whereas the familistic style is better suited to the suburbs and outer city (McElrath, 1968). The growth rate of a population will be affected by the proportion of persons selecting these styles of life. Where familism dominates birth rates will generally be higher and in an urban society this will place pressure on peripheral areas of metropolitan regions. The recent growth of the United States population, and the growth of suburbs, derives from an increase in the 39 emphasis on familism. A recent research report on moving behavior and residential choice included in the conceptualization of the residential mobility process such variables as familism, consumerism, and urban-suburban orientation (Butler, g£_gl;, 1969). Moore (1972), in a recent monograph, incor- porates an examination of life style aspiration as a motivational element in movement decisions. To the three types originally pr0posed by Bell he adds a fourth, community oriented aspirations. Practitioners of this life style stress interaction with others possessing similar group oriented values. Michelson (1970) suggests that life style is an important dif- ferentiator of urban populations and is intimately related to resi- dential locations. Pahl (1970a, 1970b) feels that the spatial structure of the city cannot be fully understood without some knowledge of the social structure. He views life style as an influencing factor in the locational structure of residential patterns. The primary concern of sociologists has been the existence of a social structure based on life style preferences along with other economic and social conditions. Their interest in the spatial mani- festation of these conditions has been confined largely to the work of Bell and Greer. Even Social Area Analysis was limited to the aspatial study of census tract data. Location was not considered important. In recent years urban social geography has come to the fore in the study of urban social patterns. Working from the Social Area model geographers have extended the analysis to the mapping of various indices of social organization. With this development it became feasible to ho test the validity of the competing models of urban structure regarding the spatial distribution of social phenomena. The application of factor analysis to social, economic, demo- graphic, and housing variables enhanced the geographer's ability to make statements about the spatial structure of the city. This multivariate technique allowed a greater range of variables to be analyzed, and it generalized about social structure by allowing variables that are very highly correlated to cluster together. Each census tract (or other enumeration area) can be located on all factors via its factor scores. The mapping of factor scores permits the generalized social structure to be viewed spatially. The primary theoretical benefit derived from the multivariate analysis of census tract data has been resolution of the debate surround- ing the social patterning of urban residential areas. The concentric zone model (Burgess, 1925) viewed growth as central, wherein economic status of families increased outward from the urban core. Expansion resulted from invasion and succession by lower status groups living closer to the city center. Hoyt's sector model (Hoyt, 1939) supported axial growth and argued that social rank varied sectorally about the city center. Higher status groups followed scenic and higher ground along radial transportation lines while lower and middle status groups pro- ceeded along other wedge-like vectors. A third area of inquiry concerns the segregation of ethnic groups in particular neighborhoods without conforming either to the concentric or axial explanations. The introduction of Social Area Analysis and its theoretical foundation viewed urban social structure from the three axes of social rank, family status, and ethnic status. A major research question L1 generated from these three axes of urban residential organization centered on which model of urban structure best incorporated the three dimensions. From separate analyses of residential distributions by social rank, familism, and ethnicity, it is thought that the zonal, sectoral, and nucleated models of urban structure may actually reflect different sets of properties. Brian Berry has suggested that the zonal and sectoral models may be considered "independent, additive contri- butors to the total socio-economic structuring of city neighborhoods" (Berry, 1965: 115). According to Berry, the city's basic residential organization may be seen in terms of the sectoral variation of neighbor- hoods by socio-economic rank and the concentric zonal variation of neighborhoods according to family structure. Berry concludes: Thus, at the edge of the city are newer, owned, single- family homes, in which reside larger families with younger children than nearer the city center, and where the wife stays at home. Conversely, the apartment complexes nearer the city center have smaller, older families, fewer children, and are more likely to be rentals; in addition, larger pro- portions of the women will be found to work. (Berry, 1965: 115-16) and suggests that: If the concentric and axial schemes are overlaid on any city, the resulting cells will contain neighborhoods remark- ably uniform in their social and economic characteristics. Around any concentric band communities will vary in their income and other characteristics, but will have much the same density, ownership, and family patterns. Along each axis communities will have relatively uniform economic character- istics, and each axis will vary outwards in the same way according to family structure. Thus, a system of polar co- ordinates originating at the central business district is adequate to describe most of the socio-economic character- istics of city neighborhoods. (Berry, 1965: 116) Some empirical tests exist which have attempted to support the contribution made by Berry. Analysis of variance techniques have proven to be the most popular. Timms notes that: 142 The major source of relevant material consists of a series of studies which have used analysis of variance techniques in an attempt to unravel the separate effects of zones and sectors in the distribution of social rank and family status scores. Less attention has been paid to the distribution of ethnicity and virtually none to that of mobility or migration status. (Timms, 1971: 230-31) Among the early attempts to sort out the spatial patterning of residential differentiation was Anderson and Egeland's study of four U.S. cities (Anderson & Egeland, 1961). They determined that family status is distributed primarily by zones while social rank (prestige value) followed a sectoral pattern. From his study of the factorial ecology of metrOpolitan Toronto Murdie concluded that, ". . . economic status and family status tend to be distributed in sectoral and con- centric patterns respectively" (Murdie, 1969). In a study of the social areas of Rome, Italy, McElrath (1962) tested the sectoral and zonal hypotheses in a cross-cultural situation. His analysis revealed that social rank and family status vary both by zones and sectors. Social rank varies inversely with distance from the center of the city, and family status varies directly with distance. In Rees' study of Chicago's ecological structure the joint influence of zonal and sectoral effects on the distribution of social rank and family status also prevails (Rees, l968). In Chicago, the relative strength of the two effects varied with the boundaries of the study area. When the entire Chicago metropolitan area was considered both indices varied predominantly be concentric zones. When the industrial satellites and rural periphery were excluded from the analysis the zonal effect became secondary to the sectoral in the distribution of family status. (Rees, 1968) h3 From the results of a Factorial Ecology of Brisbane Timms (1971) tested the spatial patterning of both social rank and family status by analysis of variance. His results reveal significant zonal and sectoral effects for both indices and an additional interaction effect for familism. The sectoral effect predominated for social rank while the familism profile exhibited a stronger zonal distribution. A summary of the information presented above appears in Table 3-1. Table 3-2 reveals the existence of a pattern formerly noted by Rees (1968), that of a relationship between the relative strengths of the zonal and sectoral effects and city size. For the larger the city the greater the importance of zonal variation of socio-economic status as compared with sectoral variation, although sectoral variation remains the more important in all the cities. Similarly, the larger the city the greater the importance of sectoral variation of family status, as compared with zonal variation, though zonal variation remains, by far the more important in all the cities (Rees in Berry & Horton, 1970: 373). The major exception to this pattern is the high sector to zone ratio for social rank in Toronto. Rees cites the particular combination of variables which load on the Household and Employment Characteristics factor, a factor which exhibits, " . . . an extremely strong zonal pattern of spatial variation" (Rees in Berry & Horton, 1970: 373). Timms subsequently reduced this situation to the generalization that, "The larger the city the more complicated its spatial structure” (Timms, 1971: 234). He credits this to the greater heterogeneity of the large city population, whereby low status minority populations confined to certain inner city locations distort the simple nature of zonal and sectoral effects. The axial movement of minorities may be expected to increase the sectoral component in distribution of family hh Table 3-1 OF SOCIAL AREA INDICES PRIMARY AND SECONDARY SPATIAL ARRANGEMENTS Effect City Index Primary Secondary Chicago* Socio-Economic Status Zonal Sectoral Family Status Zonal Sectoral Chicago** Socio-Economic Status Sectoral Zonal Family Status Zonal Sectoral Toronto Economic Status Sectoral - Family Status Zonal - Rome Social Rank Sectoral Zonal Family Status Zonal Sectoral Brisbane Social Rank Sectoral Zonal Familism Zonal Sectoral Akron*** Prestige Value Sectoral - Urbanization Zonal - Dayton Prestige Value Sectoral - Urbanization Zonal - Indianapolis Prestige Value Sectoral Zonal Urbanization Zonal - Syracuse Prestige Value Sectoral - Urbanization Zonal - * Includes the inner city, outer city, inner suburbs, outer suburbs, industrial satellites, and rural periphery. ** Includes all of the above sections except industrial satellites and rural periphery. *** Although the four cities were tested together the results are presented separately. Source: Compiled by author. 1:5 Table 3-2 CITY SIZE AND 'F' RATIOS FOR SOCIAL RANK AND FAMILY STATUS Social Rank Family Status City Population Sectors/Zone Ratio Zone/Sectors Ratio Chicago 5,959,000 1.15 2.60 Toronto 1,824,000 30.00 10.70 Rome 1,530,000 1.00 1.56 Indianapolis* 639,000 1.44 - Brisbane 594,000 1.30 4.17 Dayton 502,000 5.89 Akron 458,000 6.03 29.90 Syracuse 333,000 19.89 * Indianapolis included in Dayton, Akron, Syracuse set. Source: Timms, 1971, p. 234. no status. The infilling of inner-city areas may be ex- pected to increase the zonal component in the distri- bution of social rank. It may be that the more homo- geneous a population in terms of its degree of modernity and differentiation the more its spatial patterning will approximate to the simple, additive model suggested by Berry. (Timms, 1971: 234) Such findings set the stage for the testing of several hypotheses relating the social and spatial structure of urban areas and the patterns of behavior exhibited by urbanites. The index of social differentiation of primary concern here is the family status dimension, which has often been referred to as style of life. Most often, however, ways of living have been extrapolated from a knowledge of attributes of families at certain stages in the life cycle. A lack of followup investigations into urban life styles has led to inferences about the behavior of certain groups based upon a knowledge of certain social attributes. This dissertation seeks to shed some light on the relationship between behavior and location, and therefore provide an empirical test of the life style-location hypothesis of urban ecological theory. Janet Abu Lughod (1969) has specified the necessary conditions for the emergence of the two leading dimensions of social status and family status in Factorial Ecology. They include: (1) a ranking system by socio-economic status, (2) a clear linking of family types to specific stages in the life cycle, (3) a housing market structured to cater for each possible combination of these characteristics in distinctive sub- areas, and (4) a population consisting of independent households mobile enough to use the possibilities. These conditions have been met in the case of Toronto. Murdie's (1969) study confirms the existence of both social rank and family status dimensions in the study area. Such findings set the stage for a test of the hypothesis which states that 1:? within a particular socio-economic stratum style of life will vary by location in the metropolitan region. More specifically, this thesis will test for the existence of behavior patterns which support the generally accepted proposition that urban society, at least in North America, can be differentiated along a style of life continuum referred to as familism. Polar posi- tions on this continuum can be described by the following conditions: (1) A family oriented style of life with the wife home caring for children and engaged in neighboring and domestic activities. In this situation the husband leaves home daily for work, and engages in family oriented leisure activities when home. (2) A cosmopolitan way of life which is confined largely to the apartment building areas of the central city. The family is most likely childless or with older children, and both husband and wife work. It is expected that fundamental differences prevail between these two general life style categories in terms of daily activity systems. Within each group individual activity patterns quite likely vary consid- erably. No two individuals are exactly alike in their styles of living, but when many individuals are aggregated on the basis of shared values, stage in the life cycle, residential location, and housing environment, there tends to be a convergence of life styles into a limited number of broad categories. Therefore, in addition to the spatial clustering of households on the basis of family status, a similar tendency exists wherein urban areas can be characterized as consisting of subareas with particular living patterns. At this point certain qualifying statements must be made regarding the appropriateness of the data for carrying out such a test and the 1:8 measurement of the variables involved. Location, a critical variable, of necessity has been defined nominally. This circumstance results from the lack of families interviewed at intermediate locations from the Toronto CBD. Further limitations resulted from the contractual agree- tnents made when the data were secured. Thus, the most desirable con- ditions for testing hypotheses relating residential location to life styles must be lodged in further efforts. Nevertheless, the behavioral information provided in the time budgets allow for adequate empirical inquiry into the geography of urban life styles. The sampling design provided for the exclusion of a large propor- tion of the Metropolitan Toronto area at an intermediate distance from the CBD. This condition precluded the testing of any hypotheses incorporating gradients (e.g., life style could not be related to any set of concentric circles). Hence the necessity of treating location as a dichotomous variable. If Cans (1962) insistence that the outer city and suburbs are essentially the same in terms of life style, then the nominal categories of location presented here are adequate. In fact, the differences between inner city and suburban life styles have long held the attention of urban scholars. The following section ex- plores this subject further. ‘ggbég and Suburban ways 2£.£££E One of the most obvious results of urban growth, particularly in advanced economies with a sophisticated division of labor, has been the growth and spread of suburban areas. In the post "Orld War II era this growth has been of such significant proportions that the suburbs have generated considerable research activity on the part of social scientists, planners, and politicians. So significant an impact has suburban growth _ b9 had on current social thinking that controversy inevitably ensued. The controversy centered around the seemingly different ways of life practiced in the suburbs and central cities. While some saw the suburbs as ideal areas to raise children, recreate in Open spaces, and conserve privacy, others condemned the demoralizing and conforming effect they exerted on their residents. This dialogue reached its emotional climax with David Reisman's "The Suburban Sadness," (Reisman, 1958), but the debate has subsided in recent years. This may be the result of the growing feeling among social scientists that suburban growth constitutes merely the latest phase of an ongoing urban growth process (Berger, 1966; Gans, 1968). Contrary to the notion of a melting pot, the American culture is characterized by some as being more pluralistic than not (Glazer and Moynihan, 1963). The suburbs, then, do not differ radically from the cities in terms of the social makeup of their residents. Great cultural variation can exist across suburbs and can be accounted for by the variety of personal and collective motivations on the part of urbanites for moving into suburban areas. The conclusion most often drawn is that physical and cultural differences between cities and suburbs do not contribute significantly to the living patterns of their inhabitants. Berger contends that although statistically significant differences do exist they are 'sociolOgically spurious,’ since the most meaningful comparisons should be made not between the suburban areas and cities as entities, but between particular suburbs and urban residential neighborhoods (Berger, 1966). Both Cans (1968) and Berger (1966) argue that suburbs were seen as highly ordered and homogeneous because only highly ordered suburbs , " 24W 50 came under the scrutiny of social scientists. Their main contention is that changes in ways of life cannot be attributed to the move to the suburbs from the city since the life styles constituted the impetus for relocating in the first place. Not only did the knowledge of suburban areas precipitate a shift in preferred life style, but the suburbs allowed for the realization of that preference, and therefore are responsible for causing changes in living styles for significant numbers of people. In 1938 Louis Wirth's famous article, "Urbanism As A Way Of Life," was published. In it Wirth argued that the city was ". . . a rela- tively large, dense, and permanent settlement of socially heterogeneous individuals". He saw the factors of number, density, and heterogeneity creating a social structure dominated by secondary contacts in place of primary group relationships. These contacts were both fleeting and superficial in addition to being segmented and impersonal. Such con- ditions transformed the city dweller into an anonymous, isolated, secular, and sophisticated person who, in order to function pr0perly in society, organized corporations and unions, participated in voluntary organizations, established representative forms of government, and communicated through mass media. Wirth deduced two social consequences from number, density, and heterogeneity which he said explained the major features of urban life. These consequences were: (1) the segregation of homogeneous populations into well defined neighborhoods, and (2) increased social contact between groups due to close physical proximity, which encouraged acculturation and assimilation - the melting pot effect. He argued that, the latter being a more powerful tendency, primary group relationships 51 characteristic of segregated neighborhoods would disappear under pressures generated by dominant social, economic, and political institutions. A situation characteristic of Tgnnies' "Gessellschaft" would emerge. Gans (1962) challenges Wirth's postulates on the grounds that urban society, as we perceive it today, does not conform to the urban- rural dichotomy as Wirth saw it. Instead of a situation where urban living is characterized solely by secondary group relationships and reliance on formal organizations, Gans sees a distinct difference be- tween conditions in the inner city and those in the outer city and suburbs. A reanalysis of Wirth's conclusions from this perspective suggests that his characterization of the urban way of life applies only - and not too accurately - to the residents of the inner city. The remaining city dwellers, as well as most suburbanites, pursue a different way of life, which I shall call 'quasi-primary.' (Cans, 1970: 71) During the nineteen twenties and thirties urban sociologists at the University of Chicago (Wirth included) organized their research efforts around the problems of the inner city (Park, Burgess, and McKenzie, 1925), and were concerned about the consequences wrought by social disorganization caused by urban living (Burgess and Bogue, 1964). The human ecologists defined urbanism in terms of the differences between city life and rural life. According to Gans, Wirth ". . . examined the city from the distant vantage point of the folk society - from the wrong end of the telescope, so to Speak" (Cans, 1970: 72). Therefore, the social consequences of number, density, and heterogeneity, if there are any, will most likely be found in the inner city. Even here, most groups are relatively homogeneous and are protected fairly effectively from the consequences of number, density, and heterogeneity by certain 52 social and cultural institutions. Persons residing in the outer city and suburbs, who constitute the majority of the urban population, enjoy relative immunity from the consequences suggested by Wirth. Thus, Cans is saying that the pattern of living in the suburban and 'outer city' zones, although they may vary somewhat, are not so much different from each other as they are from those of the inner city residents, a group comprised of: cosmopolites, unmarried or childless, ethnic villagers, the deprived, and the trapped. Gans' thesis seems to be that with a few inner city exceptions a strong degree of homogeneity permeates metrOpolitan living. Life styles do not seem to vary to any great extent, except when one gets into the inner city area. He is, of course, talking about all the residents of the city, in all income categories and all racial groups. When examining ways of living from this perspective it becomes quite clear that there are fundamental differences between the inner city life styles and the life styles of the majority of the urban pOpulation living in the outer city and suburbs. But what differences might be present between groups of urban dwellers who seemingly practice the middle class style of life. Do owners of single family homes in residential neighborhoods in the city lead the same style of life as their counterparts in the suburbs, and ‘ is the same true for city and suburban apartment dwellers? Finally, might there be life style differences between apartment dwellers and people who reside in single detached homes? These are some of the questions being explored here. Life Style: A_Conceptualization Up to this point the focus of attention has been directed at an ecological view of life styles in urban settings. That is, the discussion 53 has centered on an extremely generalized continuum ranging from extreme familism at one end to extreme urbanism at the other. Within this frame- work large aggregates of urban dwellers can be classified, and inferences can be drawn concerning their socio-economic attributes, generalized preferences, and possible behavior. However, this conceptualization, although valuable from the standpoint of a low level theory, is limited by its very broad distinctions concerning life styles, and in its failure to differentiate the urban milieu into anything finer than central city and suburb. In view of these limitations it is the purpose of this section to articulate a conceptualization of the construct life style which will facilitate the study of spatial behavior and relative location. The approach, then, will focus on human behavior, with the concept of life style acting as a tool for organizing individual behavior into recogni- zable and empirically valid patterns. This conceptualization of life style will necessitate the examination of certain aspects of social roles, attitudes, and decision making. It is to be the literature on these topics that this chapter will now direct its focus. No comprehensive typology of life styles exists which can be employed in urban social research. As a result the concept has been subject to varying interpretations by social scientists. The position taken by previous researchers has been that of empirical generalization. Typologies, such as they have been, emerged largely from examinations of empirical data gathered from census records or social surveys. While it is recognized that empirical generalization presents certain limitations as a means of theory formulation, empirical research can lead to the formulation of prOpositions and testable hypotheses. These in turn can Sh be incorporated into deductive theoretical systems. This research will constitute an empirical, and therefore inductive, study of urban behavior, and most certainly is exploratory in nature. Along with others, the author is searching for a methodology which will be most appropriate for the study of human time budget behavior. The methodology outlined in this dissertation is considered a first tentative step in that direction. Practically speaking, life style is useful as a research concept in the examination of the social and spatial characteristics of urban populations; social because life style reflects the social values and social interaction patterns of individuals and groups, and spatial insofar as similar individuals seek to position themselves in spatial proximity, and display spatial regularity in their day to day behavior. Increasingly, neighborhoods are coming to be identified by the styles of living of their inhabitants. The social area map which peOple of any sizeable town carry in their heads is more clearly defined and socially specific . . . each (area) representing important expres- sive aspects not only of the income but of the occupations, social proclivities, emotional background, and social pre- tentions of the people who live in them - or rather, of the kind of people who are supposed to live in them. . (Burns, 1968: 56) Life style, as an empirical construct, finds utility in the study of spatial behavior; particular forms of spatial behavior may be characteristic of certain life style groups. How people move about in the city and avail themselves of the facilities of the metrOpolitan area is postulated as being a direct reflection of their preferred style of living. Spatial interaction can also be used as an Operational method of identifying living styles. 55 If the needs of social planning require a more micro analytical frame of reference (Chapin, 1968), then a conceptual framework which incorporates behavioral propositions in the study of urban structure is more desirable. The level of aspiration involved in this approach, as Olsson (1969) would see it, aims at higher order statements about life style than that of Bell (1968) or Greer (1962). Emphasis should be placed more on the behavioral processes shaping interaction rather than on the structure of social areas. To this end, life style is being conceptualized as role emphasis and value orientation, which can be translated into motivations to act and thence to behavior. Others, notably Michelson (1970), Michelson and Reed (1970), Boskoff (1970), Tallman and Morgner (1970), and Havighurst and Fiegenbaum (1959), have incorporated the notion of social roles in their efforts toward the understanding of life style. Role emphasis here refers to the predominant set of social roles an individual selects to play at various stages in his life cycle. These roles, both ascribed and achieved, are manifestations of the self image and ideals individuals possess. The objects, affiliations, and activities associated with given ". . . directed and given roles are, in the words of Michelson and Reed, consistency by the preferences and orientations of the individual - in short, by his hierarchy of values" (Michelson and Reed, 1970: 17). The formulation and shaping of values are, of course, life long processes. Some values in the hierarchy, the most fundamental, remain ingrained and rarely change. The Protestant work ethic and certain moral values will serve as examples of these. Many values are constantly undergoing change, however, with new ones being formed and old ones being discarded as the individual encounters new experiences and environments. 56 The notion of feedback in the form of reinforcement or change assists in the evolution of value orientation. Such characteristics as age, sex, education, occupation, and ethnic and religious affiliation are the primary interacting variables which determine the hierarchy of values held by individuals (Michelson and Reed, 1970). These, in turn, influence the selection of roles to be emphasized. Ogburn and Nimkoff defined a role as, ". . . a set of socially expected and approved behavioral patterns, consisting of both duties and privileges, associated with a particular position in a group" (Ogburn and Nimkoff, 1958: 153). To a significant extent roles are voluntarily assumed by most individuals (achieved roles) but many are not (ascribed roles). Parsons and Shills (1951) view the role as the most significant unit of social structure: The role is that organized sector of an actor's orientation which constitutes and defines his participation in an inter- active process. It involves a set of complementary expectations concerning his own actions and those of others with whom he interacts. Both the actor and those with whom he interacts possess these expectations. Roles are institutionalized when they are organized around expectations of conformity with morally sanctioned patterns of value-orientation shared by members of the collectivity in which the role functions." (Parsons and Shills, 1951: 23) Individuals take on social roles which fit into their predominant value orientation. Roles are selected on the basis of how the individual perceives the set of social expectations which is attached to each role. To Boskoff, style of life reduces to a selective ranking of available roles. "Each life style signifies to the typical adherent some range of role choices and a distinctive hierarchy of such roles" (Boskoff, 1970: 199). Michelson postulates the minimum elements included in life style. One consists of a behavioral set necessary to satisfy a role. Behavior, then, in the form of an activity or set of activities which are related, 57 can be said to constitute the fulfillment of a role. The typical behavior of a middle class family man should differ markedly from that of a child- less, career conscious executive. Their respective consumption of time, space, and activities, as well as material goods, should vary according to their predominant value orientation. The second element concerns which sphere of life is emphasized. Generally, five spheres are cited as central elements of any society: political control, economic supply, propagation, socialization of the young, and explanation of the super- natural. Depending upon the roles people select to play in life, certain of these spheres would be emphasized at the expense of others. The direct result of the determination to select specific roles to play is the set of behaviors necessary to fulfill each role. It must be noted that the individual does not consciously decide which activities are congruent with a particular role. The conventions and norms society attaches to specific roles usually define the context of each role. In this sense most behavior is habitual or conforms to patterns of regularity. The individual is constrained, to a degree, in which activities he may employ to fulfill the content of a role. The patterns of behavior which typify the role choices of urban residents can be thought of as defining their style of living. A central assumption of this conceptual schema holds that the patterns of activities which characterize each individual are a direct reflection of that person's hierarchy of values. Boskoff (1970) has outlined a schema such that if a class or stratum can be identified by opportunity variables (education, income), then a related style of life can be identified by the manner in which opportunities are used, evaluated, and rewarded by a relevant reference group. Boskoff interprets style of life as a product of experience in 58 trying to provide adequate means of adjusting to available facilities, resources, limitations, and frustrations. Adjusting would entail some forms of behavior. In addition to the methods outlined above, he suggests three indicators by which styles of life in urban regions may be distinguished. They include: (1) training, which refers to acquired skills that concern ability to perceive opportunities and those that enable use of opportunities; (2) patterns of associational affiliation; and (3) dominant role conception (Boskoff, 1970). From these indicators, style of life categories can be derived which correspond to those in Figure 3-1 s Consumatory Style The Social Elite Striving Style Nouveaux Riches Exurbanites Organization Men Derivative-Vicarious Old Middle Class Style New Middle Class Stable Poor Persistent-Rebellious Bohemian Style Fugitives Resignation-Alienation Unstable Poor Style Social Outcasts Indolent-Apathetic Source: Boskoff (1970) p. 202. Figure 3-1 MAJOR URBAN LIFE STYLES AND RELATED STRATA 59 What remains unclear with respect to Boskoff's typology is the relative degree Of influence each Of his three indicators contributes to each life style category. The exact position of training, associational affiliation, and role conception in this arrangement remains unclear. Unfortunately, he fails to satisfactorily relate the indicators to suCh terms as Old Middle Class and Unstable Poor. The value Of role emphasis in the investigation and subsequent delineation Of life styles is manifest in two aspects. First, since the tendency exists in almost all human beings to conform to some outside set of influences and pressures (mediated by a system Of punishments and rewards) there is likely to be a significant convergence of life styles into a finite number of general categories. Those categories will be based on common patterns Of role emphasis shared by individuals. Second, at an Operational level, dominant role patterns can be inferred from an examination Of human behavior. This can be accomplished through a whole range of pattern recognition devices. There is a paucity Of research treating the concept Of life style as a role oriented set Of behaviors. None proceed from even a skeletal conceptual framework. However, several deserve mention. Havighurst and Feigenbaum (1959) defined life style as a description of a person's characteristic way Of filling and combining the various social roles he is called on to play. Operationally, a life style was defined as a pattern of role performance scores for leisure activities shared by a group of peOple. Using data gathered from a sample Of adults living in Kansas City they arrived at four general styles of life: (1) com- munity centered, emphasizing outside the home activities, (2) home 60 centered high, (3) home centered medium, and (4) low level. Based upon much Of the same data is a life style categorization devised by Williams and Wirths (1965). They defined six styles of life among adults, and attempted to relate the process of aging to each style. Their categories were: (1) world of work, (2) familism, (3) living alone, (4) couple- hood, (5) easing through life with minimal involvement, and (6) living fully. Tallman and Morgner (1970), in an attempt to assess differences in life style between urban and suburban blue collar families, assumed a definition of life style which included a number of behavioral activities and orientations. Each behavioral activity required a distinct investment Of resources of time, energy, affect, or money. A series Of questions was designed, the answers to which indicated degree Of local intimacy, social participation in voluntary associations, church participation, family organization, subjective class identi- fication, mobility orientation, and political orientation. Although the data provided a role profile for each respondent, no attempt was made to classify individuals according to dominant life style. One of the more significant conclusions concerned the fact that residence is associated with fundamental differences in life styles for working class families. Class and class related cultural factors interact with residence to influence some of the life style patterns (Tallman and Morgner, 1970). Within the confines of this conceptual framework the relationship between life style and location will be investigated. The location variable will be defined as being either suburban or central city, 61 while life style will be operationalized in a more complex form. The following chapter explores the behavioral basis of life style together with the literature on time-space budgets and activity systems. The second part of the chapter outlines the methodology employed in this study. Chapter 4 LIFE STYLE MEASUREMENT: HUMAN ACTIVITY SYSTEMS Systems of Human Activity In order to more fully understand the analytical aspects of this dissertation it is deemed important to investigate the general concept of activity systems and the process of activity choice. Recent re- search on time budgeting has been devoted to the specification of the rules under which particular activity decisions are made. Thought has progressed to the point where an activity decision is no longer con- sidered to be made independently of other activities, but rather the decision is viewed as being interdependent with other activity choices which may precede or follow it. An additional develOpment concerns the movement from the assumption of a rationally economic man to that of a more realistic position Of intended rationality in decision making. This chapter explores these concepts and relates them to the methodology which is applied to the analysis Of activity patterns. Although time budget research has been carried out for quite some time, it was Chapin (1968) who attempted to give some conceptual and even theoretical structure to this type of research. He and several of his colleagues have built a conceptual framework around household activity systems. They argue that, ". . . as an area of theoretical investigation it offers promise Of supplying some conceptual guidelines for relating resident behavior patterns to the spatial organization of the city" (Chapin and Logan, 1969: 306-7). 62 63 Activity systems, as conceptualized by Chapin (1965,1968) concerns the nature of human satisfactions and dissatisfactions and how activity choices are affected by them. All activities are carried out within the confines of a social system, here defined, ". . . as a milieu made up of a universe of behaviors and environments (physical, economic, social)" (Chapin, 1968: 12). Consequently, the manner in which the environment influences daily activity routines, and is in return shaped by them, is a function of personal values of individuals and their motivations for satisfying their value prescribed needs. This is in keeping with the value orientation definition of life style elaborated on earlier, where- in the hierarchy of values held by an individual or group of individuals can be deduced from their daily activity patterns. According to Chapin and Hightower (1966), the activity choices of individuals can be inter- preted as an independent indication of life style. The extent to which activities can be considered as a measurement of life style is open to question, but certainly activities must be considered in any definition of the concept. It seems appropriate to state here that the choice of any activity is the outcome of many factors which are operating on the individual decision maker. These factors include the goals the individual seeks to fulfill, the actual and perceived constraints which channel activities in space and time, the individual's values, his personal characteristics, and other activities already completed or being contemplated. Of course, these general factors can be further subdivided into more specific com- ponents such as the actor's location, the time of day, activity just completed, the degree to which any activity must be completed, age, sex, and so on. It is the purpose of the next section to explore the 611 relative importance of these factors, and to demonstrate how they result in patterns of activity sequences. The Motivational Basis 2: Behavior To some extent, all behavior can be said to be motivated or di- rected toward some goal. The purposefulness of behavior is quite variable, however, both between individuals and among activities. It is quite clear that individuals do not operate as rational economic men, and that even intended rationality in decision making may be an un- realistic assumption (WOlpert, 1964; Fred, 1967). Some see rationality as an unsolvable issue in that all behavior has some purposefulness to it, and that any assessment of rationality on the part of the observer is a dubious process at best, which is subject to a variety of cultural, ideological, and personal biases. To construe a pattern of behavior as being rational or not depends upon an understanding of the values, goals, and purposes toward which the behavior is directed (Cullen, Godson & Major, 1971). Individuals make activity decisions based upon some motivational component and the amount of information that they possess regarding dis- tance to the site, anticipated duration, time of day, and other factors. In most situations this information is incomplete, creating conditions of uncertainty. Golledge (1969) carries the rationality argument to the point of distinguishing among problem solving behavior (involving rational decision making), weakly motivated behavior (random behavior), and habitual response making behavior. He mentions that routine or habitual response making behavior may predominate more frequently than the other two forms. 65 Routine behavior generally follows problem solving- behavior and is of greater duration through time . . . Routine behavior . . . often follows a path of minimum effort, it serves to reduce uncertainty in the decision process, and reduces consideration of alternative sources of action. In other words it is the behavior most used to cope with contingencies of everyday living. (Golledge, 1969: 103-4) One might surmise at this point that the distinction between problem solving behavior and habitual response behavior is synonymous with that between discretionary behavior and obligatory behavior. To a certain extent this may be the case, since obligatory behavior (work- ing, eating, sleeping, etc.) usually forms an habitual pattern. But to classify activities as either discretionary or obligatory entails a failure to account for individual differences in the way people perceive their acts. What is discretionary for one person (the decision by a businessman to attend a concert) may be considered obligatory by another (a music professor considering it part of his professional duty or committment). We may be dealing with a continuous variable here (degree of obligation or discretion) which itself may be the result of a series of factors. It is important to remember, then, that the decision to engage in a particular activity at a particular location and time may not really be a decision at all, but a patterned response to some set of stimuli which recurs each time the stimuli are present. Rather than saying work is obligatory, it might be more precise to classify it as habi tual s In summary, it can be said that some behavior appears to be almost instinctive while other behavior is obviously highly calculated. To classify behavior as consistently rational seems unjustified. 0n the other hand, it is neither completely habitual. A more realistic 66 approach is to view behavior as highly organized activity episodes which impart structure and pattern to blocks or sequences of activities. Thus, it is advanced here that most activity sequences are not random but are the result of a certain amount of organization on the part of individual ac tors a The Interdependence p£_Activities Most research in human behavior fails to take into account the fact that any activity in a person's day constitutes part of a sequence of interdependent events. It follows, then, ". . . that the decisions governing these events must therefore also be interdependent owing to the necessity of continuity of action in both time and space" (Cullen, Godson & Major, 1971: 2). Although some rather general suggestions have been made by Meier (1962) and Weber (1968) concerning the inter- dependencies of flows and networks in urban areas, . . . there has been very little development of either ideas or data relating explicity to the question of the interdependence between activity choice decisions over time. The task of pattern recognition has not yet taken the form of identifying sequences of activ- ities which can be viewed as activity modules or of trying to understand the processes which govern the integration of the individual's time-space behavior. (Cullen, Godson & Major, 1971: 3) Capin and Hightower (1966) have proposed that the activity de- cision is in reality a two stage one. First, the individual selects the activity he wishes to pursue from among the alternatives he has available to him, and then he settles on where he will do it (presumably the time component is lodged in the activity choice decision). This framework seems unrealistic since a knowledge of alternative possibili- ties includes knowing some aspect of the location of those opportunities. A more acceptable conceptualization would be that the choice of activity 67 and location entail a single linked decision. An individual is typically confronted with a set of choices between different activities at differ- ent locations (walldéh, 1968). An elaboration of this concept is neces- sary in light of its importance to the subject of activity linkages. The activity choice process involves a linked or interdependent decision in two different but related ways. First, the choice of any particular activity involves consideration of the range of activities available, time and space. When, where, and how long an activity will take place will exert an influence on its selection. Second, activity choices are linked to previous and subsequent decisions because the choice of a particular activity at one point in time and at a certain location will influence the likelihood of other activities being under- taken. Hemmens (1970) sees the individual as acting out two distinct sequences of activities over the course of a day. One is the set of out-of-home activities comprising the full daily cycle, and the other includes the activities which occur on each foray outside the home, those activities linked together on each individual journey. In the latter category can be considered single activity and multiple activity movement (a sequence of activities performed at different locations out- side the home). This framework fails to consider in-home activities which must be taken into account in any analysis of activity patterns. In-home activities contain a spatial component in that the decision not to leave home places the actor in a unique location. For instance, given a certain amount of leisure time, one may Opt for bowling at a local establishment or for playing cards at home. Once that decision has been made, different sets of activity sequences become more likely 68 because of one's location. Employing a two dimensional time-space framework, Hagerstrand (1970) has suggested a model of behavior, whereby a person's daily time- space path is structured by different sets of constraints. These he breaks down into: (1) capability constraints, primarily biological and technological, (2) coupling constraints, or the necessity for interaction with objects and other human beings, and (3) authority constraints, or those imposed by outside agencies or other individuals. Simply stated, the individual, at certain times, must be at particular locations. In order to meet these committments, the individual's area of movement dur- ing intervening periods is constrained by time and the availability of transport. The activities which are fixed in time and space limit the individual's choice of other activities to those which fit into a time- space prism which defines a potential space of activity. Since certain activities are fixed in both time and space, the day can be divided into allocated and unallocated, or fixed and unfixed periods. The allocated periods dictate the potential (time-space bud- gets), not only for the entire day, but particularly for the unallo- cated activities. Hagerstrand's model thus demonstrates the interde- pendent nature of activity decisions. It appears fruitless, then, to consider the decision to do something at a particular place and time as being related solely to the individual's place of residence. The de- cision is related to a very specific time and space position which con- nects the individual's present location and that of his next committed activity (Cullen, Godson & Major, 1971). Within this framework behavior may be highly purposeful, random, or habitual. As an extension of Hagerstrand's model, Cullen and others (1971) 69 have elaborated on the questions of priorities and constraints and their influence on the flexibility and scheduling of activities. By priorities they mean the order of importance individual's assign to activities ac- cording to considerations of finances, presence of other persons, order of planning, personal likes and dislikes, and so on. In conjunction with priorities, activity choices are constrained to the extent that physical, economic, institutional and accessibility deterrents are operating at any time in any place. One may be limited by the amount of money he has, business hours of certain institutions, and availability of transport, in addition to a whole range of perceived constraints which may in- fluence an activity decision at any particular time (Cullen, Godson & Major, 1971). The flexibility with which activities can be scheduled in sequences depends upon the degree of committment the person affixes to them, and on the real or perceived deterrents to their being carried out. The degree of committment controls the flexibility of the types of activities which are selected, while the constraints Operate on the location of activities in time and space. In this manner, activities can be related to both the individual decision maker and to the organization of his environment. By contrast, Hemmens (1970) mentions the three activity choice dimensions of (l) the activities themselves, (2) time, and (3) space, but fails to organize them into a coherent framework and relate them to the organization of the city. Kofoed (1970), on the other hand, has organized activities in such a manner that they can be related to the social, temporal and spatial organization of an urban area, thus making it possible to locate any activity in a three dimensional activity space. 70 Due to urban social organization the activities imposed upon people can be related to: (1) (2) (3) (4) (S) (6) PhysiolOgical needs (eating, sleeping, inactivity caused by illness) Social duties (school attendence, jury service) Agreements (working, attendence at certain meetings) Sudden urgent needs (repairs, visiting the dentist) Services (shopping, using banks) Leisure (social visits, recreation, walking) Temporally, societal organization imposes schedules which restrict (or enhance) the Opportunity for engaging in different activities at various hours of the day. As a result, activities may take place: (1) (2) (3) (4) (5) Over a number of consecutive and predetermined hours (walking, attending school) At any time between two hours (shOpping) At a fixed hour (an appointment) At any time but related to other persons (visiting) At any time but unrelated to other persons (walking) Finally, in the spatial dimension, activities tend to be more or less fixed at single points in space. These may be: (1) (2) (3) (4) At fixed locations (most work, family visits) Within administrative borders (social services, visits to or by the doctor) In a purpose related facility (shopping, visits to the cinema) At any place (walking, enjoying the company of other peOple) According to Kofoed (1970), by means of these three dimensions, activities can be related to the prime factors influencing the activity sequence. Thus, activities are predetermined to the extent of their 71 Ilocation in the activity space. For example, groceries can be purchased atLa self-service store without prior arrangement at any time at almost ‘any store, while an office visit to a doctor requires an appointment, ‘and may follow consultation with a referral agent. Different activities relate in different ways to other activities, and will occur therefore as a result of and precipitate different activity sequences. The individual's day becomes organized around certain key activ- ities which rank high on some scale of committment. Such activities Inight include work, shopping, business or physician's appointments, care to children, housework, school, or meetings of voluntary organizations. Degree of committment can be related to a whole series of characteristics such as age, sex, occupation, family status, income, education, etc. That is, certain key activities may be characteristic of particular groups of people living in the city. Their life styles may revolve around the most highly committed activities of their daily routines. Built around and between the more stable activities are various combinations of other activities which may be highly purposeful, only weakly motivated, or even close to habitual. The choice of some of these more flexible activities may be facilitated by being in a partic- ular place at a particular time. Travelling to the local shopping center for the weekly groceries may cause one to remember to pick up some previously left clothes at a nearby dry cleaning establishment. Immediately following the trip home might be a period of unorganized relaxation in anticipation of another key activity later on in the day, preparing dinner for example. The manner in which various activities are related to each other by linkages results in the individual daily routines which, when 72 aggregated, form the interaction patterns of an urban area. Groups of ,people:have significantly different patterns of behavior, and one way of ldefining these patterns is by the sequences in which activities occur. ‘The following section outlines means by which patterns can be extracted from time budget data, and presents some past attempts to do so. .Activity Linkage Analysis The fundamental question being asked in this dissertation concerns the differences which exist among individuals who inhabit different resi- dential environments. Differentials in location and housing type act as either attractions or deterrents to persons with different life style preferences. Once settled in a particular housing environment the indi- vidual begins to act out his preferences for living. It is assumed here that information about life style can be recovered by an analysis of the behavior exhibited by urban dwellers. One aspect of behavior through which differences in life style can be studied is that of activity linkages. As individuals and groups ambulate throughout the city over the span of a twenty-four hour day they engage in a variety of different activities. As a result, different in- dividuals perform different activities which they link sequentially into patterns of association (Hemmens, 1966). If we are to assume that per- sons travel about the urban area with a set of goals they expect to ac- complish, then the manner in which they link activities together, both temporally and spatially, can be considered a kind of strategy for real- izing those goals. It has been suggested further that the dimensions of household activities and the linkages between them could be isolated and the range of patterns organized into a typology of life styles 73 (Horton and Hultquist, 1972). The study of activity linkages can provide a clearer understanding of the manner in which groups of people associate different types of activities in temporal sequences. Secondarily, it provides insight into the complexity of urban travel patterns. In a more general sense: Knowledge of the extent to which different activities and different land users are knit together in the daily travel of persons in an urban area is crucial to our ability to plan for the future. (Hemmens, 1966: 1) From the standpoint of spatial planning and social engineering a comprehension of the complexity of processes Operating in the city, such as activity linkages, can provide a better understanding of the spatial structure of urban areas. The process which is under consideration here concerns the interaction between households and other households and be- tween households and the various spatially separated economic and social functions in the city. Thus activity linkages possess a spatial as well as temporal dimension. One direct result of an activity linkage analysis will be a determination of how various facilities in the city are linked functionally and spatially. The intensity of multi-purpose shopping trips and the clustering of related shOpping facilities is one example of the relationship between travel linkages and the spatial juxtaposition of certain urban functions (Nystuen, 1967). The primary concern here is the differences in linkage patterns for the activities of various residentially defined populations. As stated before, the selection of a residential environment in which to live is based upon, among other things, life style aspirations. One measure of life style can be achieved through identification of dominant behavior sets or activity patterns. The aspect of behavior under scrutiny here is the linkage between activities. 71: ‘Most studies which have treated the subject of activity linkages have directed their attention to out-of-home activities, which in turn means the study of travel patterns or spatial behavior. Much research dwells on the relationship between residential site selection and de- sired movement patterns. The two are intertwined since the choice of a residential site is influenced in part by the set of intricate travel patterns required to satisfy day to day demands for goods, services, and leisure (Marble, 1959, 1967). As a result, much effort has been expended on the manner in which individuals array functions, or activities, on trips through the city. People do, however, spend a considerable amount of their time at home, and aSpects of the immediate home and neighbor- hood loom large in the selection decision, especially for women (Michel- son, Belgue & Stewart, 1972). To the extent that most activity linkage research has been con- ducted by planners and geographers, the emphasis has naturally been placed on spatial or movement aspects of activities. Therefore, research methodologies, including definitions, measurements, statistical tests and mathematical models have been employed with a definite spatial bias to them. As an example, consider the fact that most activity data have come from traditional transportation surveys where activities are de- fined by land use or trip purpose at origin and destination. Also, travel time to an activity site has been considered to be an important cost an individual weighs when he decides upon an activity. Duration of an activity has also been another important factor in the activity decision. Closely related to duration are the time of day and sequence of activities. The duration of an activity may be directly related to the ,._ ‘2 u“, 75 time of day when it occurs. The sequencing of activities may influence their duration, and time of day may influence both sequence and selection. Using data gathered in a transportation study in Buffalo, Hemmens (1970) found that the total number of activities undertaken in the morning, afternoon and evening were evenly distributed. However, through Chi- square tests and Analysis of Variance, he demonstrated that the proba- bilities of certain types of activities occurring were significantly different by time of day. For example, more social recreation activities were undertaken in the evening than in the morning, and personal business activity prevailed in the afternoon hours. Activiterinkggs Modelling One method of describing the activity linkage patterns of any population, or sample thereof, is to aggregate their daily activities into a matrix of associations or linkages. Such matrices can be con- structed for various subgroups of people for purposes of comparison. The frequency of movement from one activity to another can be calculated and compared with the frequency links for another group of individuals. The main weakness of this type of approach lies in the fact that fre- quency counts are totally dependent on sample size, and therefore offer no basis for analysis. What is required is a means for assessing the relative strength of any frequency across a row of linkages. In other words, a measure of the probability of moving from one activity to another in a sequence is needed. Since these probabilities will vary for each activity sequence, the linking of activities can be considered a stochastic process. Therefore, a commonly used model for describing the activity choice process for aggregated data has been the Finite Markov Chain. 76 The Finite Markov Chain belongs to a class of probability models called Markov Process Models. The central element in these models ia a probability pij which refers to the likelihood of transition or movement from one state (i) to another state (j) within a given interval of time. The pij is a transition probability which, together with all other pos- sible outcomes, sums to one. Given data (aij) reflecting movement be- tween pairs of locations or change for pairs of classification over some time interval, the transition probabilities (pi ) can be readily esti- J mated as: an = 311 4:313 Each aij can be considered the number of times a particular activity was selected following any other activity. The Pij would be this number expressed as a probability. Hemmens (1966) refers to these as linkage coefficients. A matrix of linkage coefficients is a stochastic matrix containing the probabilities (pij) that a particular activity would fol- low any other activity. As a result, the typical matrix of activity linkages may take the forms of Figure 4-1. As a simple example of a transition probabilities matrix, consid- er the following. Let us assume that the probability of following an in-home activity with another in-home activity is .80; the probability of selecting an out-of-home activity becomes only .20. Once out of the home, however, the probability of returning home reduces to .60, and of remaining outside, .40. With this information we can construct a transition matrix (Figure 4-2). 77 * A1 9 A2 , A3 , sssossssssssssssassoosssss , An Al P11 P12 P13 asssssssassssssssssaoososs Pln A2 P21 P22 P23 assessssssosssossaaaasses. Pl“ A3 P31 P32 P33 aoososoooooaosooooossoaooo P3n An Pnl Pnz Pn3 0Otooooooososssssssssssos Pun *(Al’ A2, A3 .........................., An) is the set of all activities and the pij are the probabilities of linking any activity with any other activity. Figure 4-1 A TRANSITION PROBABILITY MATRIX MODEL 78 In Out-of Home Home In Home 080 .20 . 100 Out-of Home 060 040 = 1.0 Figure 4-2 SAMPLE TRANSITION PROBABILITY MATRIX We now have the estimates of the likelihood of moving from one state (the vertical axis) to another (the horizontal axis). The probab- ility of being in state j at time tel is conditional only upon being in state i at time t and nothing more. In the transition matrix, which serves as input to the Finite Markov Chain, the pij remain stationary. That is, they do not change through the steps of the chain. With the preceeding information it becomes a simple matter to con- struct a matrix of transition probabilities from time budget data to estimate the manner in which people sequence activities. Activities, or groups of activities, can be considered the states. The raw data matrix can be constructed by counting and summing the number of times each activity is followed by another activity sequence. The transition matrix, P, is then easily derived. As previously noted, much activity data in the past has been de- rived from traditional transportation surveys. One of the first to employ the transition probabilities matrix in a study of activities with the use of such data was Hemmens (1966). His activities were defined by the trip purposes, to and from home, work, shopping, school, social- recreation, and personal business. He compared the trip data from 79 transportation surveys conducted in Buffalo, Chicago and Pittsburgh, and concluded that the linkage structures for both Chicago and Pitts- burgh were very similar but different from that for Buffalo. However, Hemmens cautioned that, "It is quite likely that there are substantial variations in the data from the three cities due to field, coding and processing procedures. Since the amount of variation in the data is not known the results observed here can only be considered a suggestion for further research" (Hemmens, 1966: 20). Hemmens realized the inadequacies of the data he was working with, but was interested in presenting a method whereby activity data could be analyzed in a productive way. One difference between the research re- ported on in this dissertation and Hemmens' activity linkage study is the use here of time budget data from a sample of urban dwellers. Another difference is that comparisons will be made between groups in the same sample, and not between independently collected samples from different cities. Sampling, coding, and interviewing errors have been minimized in this study. Other studies utilizing trip data have employed the Markov Chain Model as a means of summarizing and predicting activity choice. Marble (1964), for example, was among the first. Employing data from the Chicago Area Transportation Study he constructed a transition probabili- ties matrix for the same trip purpose categories listed by Hemmens. His results reveal little that is of analytical value in the study of activity linkages because he makes no comparisons with other similar data. His categories are too gross and deal with only vehicular trips. Like Hemmens (1966) Marble was demonstrating a method and advancing sugges- tions for its future application to movement data. 80 Horton and Shuldiner (1967) employed an analysis similar to that (sf lflarble and Hemmens. They subjected both links between land uses and tnflip purposes to Markov Chain Analysis, but did not make any comparisons with other times or places. Conducted by transportation geographers or planners, all these studies have demonstrated a concern for the pre- dix:tion of the average number of stOps on any trip and the percentage of persons found in any particular state (trip purposes at any random time during the day). In a later article, Horton and Wagner (1968) attempted to analyze trip»1inkage structures for different groups in the city of waco, Texas. IPirst, they performed a modified Social Area Analysis of the city, finding four different types of residential neighborhoOds based upon household, ethnic, occupational and economic characteristics. A principal compo- ‘nents analysis identified neighborhoods displaying: (1) upper socio- economic, (2) lower socioeconomic, (3) middle class, and (4) ethnic characteristics. By categorizing each sample respondent according to one of ten socioeconomic occupational groups and one of the four neighbor- hood types, 40 combinations of socioeconomic class groups based upon upon neighborhood and occupation were identified. Next, in order to define groups with similar travel patterns, a matrix of trip linkages based on trip purposes for each of the groups was subjected to a group- ing routine and yielded the three classes of high occupational status, low occupational status, and housewives and students. A trip purpose matrix of transition probabilities was derived for each group, and dif- ferences in specific linkages were compared. The most significant findings were that non-workers such as women and students substituted strong linkages between home, school, and 81 social-recreation for the traditional home-work linkage, a not too sur- prising fact. Clearly, for the other groups, home-work linkages pre- dominated. When the transition matrices were subjected to multipli- cation in a Markov Model it was discerned that the middle to high socioeconomic groups had a more diverse trip pattern than the lower group, and showed more multi-purpose trips. This study, although of more analytical value than previous works, suffered from several weaknesses. First, the use of trip purpose categor- ies (almost always including the "home" category) proved too cumbersome to adequately assess trip linkages because of the dominance of home and work. Second, the multitude of grouping procedures performed upon the data causes this writer to wonder at least if any meaning can be imparted to the end result. Grouping has a function, but it can be carried too far. Finally, and this criticism can be directed at the previously mentioned works, it appears that the activity data (by trip purpose) analyzed in these research papers do not meet one of the fundamental as- sumptions of Markov Chain Analysis, that of independence of activity choices proposed here, it seems unrealistic to employ a model whose central assumption is that location in one state is only dependent on location in the immediately preceeding state and nothing more. As a final example of modelling activity linkages, consider the work by Hanson and Marble (1971). To a matrix of trip linkages consist- ing of twenty-six different land use categories they applied the Savage- Deutsch-Goodman transaction flow model. This model presents a systematic method of determining the important linkages once the effect of different absolute levels between categories have been eliminated. This technique revealed significant linkages between shopping goods locations and 82 another set between work and certain service uses. Once again, no com- parisons were made between groups; the description of travel linkages in the data was the objective. A Methodology For Analyzing Activiterinkages When this work was in its formative stages Markov Chain Analysis was considered as a possible modeling device for summarizing activity linkage data. However, the assumption of independence in the Markov Model and the conceptualization of interdependence of activity choices presented earlier proved to be antithetical to each other. As a result, the Markov Model was dropped from consideration. For purposes of model- ing, Markovian models are of great value but since the objective of this dissertation is not so much to model as it is to compare and describe, other more simplified techniques became more desirable. Specifically, this research attempts to analyze the differences in activity linkages or sequences between groups defined by residence, location and sex. To this end, linkage matrices have been constructed from the time budget data described in Chapter 5. The matrices take the form of both fre- quencies and transition probabilities and reflect all combinations of sex, location, and residence. The Objective will be to assess the variation between linkage matrices while controlling for the factors just mentioned. In this way extraneous variance will be minimized so that differences in life style can be considered. The analysis will be constructed on three levels. At the first level, transition probabilities matrices reflecting only differences in location and then housing type will be compared for variation in activity sequences. At the second level a similar Operation will be performed on matrices which reflect differences in residence type while controlling 83 for location, and for differences in location while controlling for residence type. Finally, at the third level of analysis, comparisons will be made between residence types while controlling for sex and lo- cation, between locations while controlling for residence types and sexual differences, and between locations by sex with residence con- trolled. At each level Of analysis the Observed differences will be evaluated in light of certain hypotheses which will be advanced. Of course, it is expected that significantly large variations will manifest themselves only for certain linkages and not for others. For different groups different activity sequences should predominate. Those which differ by some significant amount should give an indication of the activity links around which certain groups anchor their day. For most activities the variously defined groups should display quite similar linkage patterns, and these will consequently be of little concern, even though they may be of a high magnitude. The anchor points, or pegs, may then be considered indicative of life style differences, such differences then being related to variation in residential environment. The reader may be prompted to question why any differences should be considered significant on the basis of visual comparison of transition probabilities. He or she would be expressing a legitimate concern, one which is not easily countered. To put the question another way, why could not the differences be attributable to chance alone? Such a question cannot be answered definitely because no technique exists for statistically testing the difference between any two cells of two matrices. Chi-square tests exist for comparing any observed matrix with a theoretically expected one. Such a test cannot be employed in the comparison of two different observed matrices. 81: Although the transition probabilities matrices provide a summari- zation of activity linkage data for purposes of analysis, they remain rather large and cumbersome (100 cells to each matrix) for purposes of comparing different pOpulations on the underlying dimensions of activity sequences. Whereas the individual linkage coefficients in each transi- tion matrix can be compared with linkages in other tables, it is dif- ficult to comprehend any patterns of activity flow in any of the matrices. They are simply too large. As a means of reducing the transition matrices of activity link- ages to a more manageable state for analytical purposes each matrix was subjected to a principal axis factor analysis whereby the 100 linkages in each table were reduced to their most fundamental dimensions. As a result, the basic activity patterns for each level of analysis have been identified and can be compared and contrasted across the variously de- fined groups in the sample. Among geographers, factor analysis has enjoyed immense popularity in recent years, particularly in the study of urban ecology (Berry, 1971). However, by comparison, factor analysis has found fewer applications to the analysis of flow data. Nevertheless, the technique has proven to be beneficial in the analysis of flows between spatial units (census tracts, cities, counties, states, countries, etc.) and a methodology of flow analysis has been prOposed (Britton, 1971). Studies treating the aspatial case of linkages between activities as Opposed to spatial units are rare. The factor analysis of an activity matrix, such as those under consideration in this study, results in a series of factors which identify clusters of activities that are alike in terms of the activities 85 which precede them in sequence. When considered in a flow matrix, the activities arrayed along the vertical axis are regarded as origins and those along the horizontal axis act as destinations. In an R-mode factor analysis, correlation coefficients are computed among the columns, or destinations. The factors which emerge isolate groups of destination activities which are similar in terms of their origin activities. In the Q-mode analysis correlation coefficients are computed among the rows and factor analyzed to reveal groups of origin activities which are alike in terms of the activities that follow them. Factors can be identified by the strength or magnitude of the factor loadings for each activity type on each factor. Loadings of .50, or greater, are generally considered to be large enough for purposes of factor identification. For each observation (origin activity) a factor score is generated which can be interpreted much like a standard de- viation. In both Q-mode and R-mode analysis the strength of the factor scores indicates which activities are strongly associated with each factor. Usually, factor scores of greater than 1.0 are considered significant. In the R-mode, the factor scores identify those activities which usually precede the clusters of activities loading highly on each factor. In the Q-mode, the pattern is reversed. Wheeler (1972) performed both a Q-mode and an R-mode factor analysis on a ten by ten activity matrix of trip purposes. His results revealed a factor structure which was remarkably similar for both modes. Three factors in each analysis were identified and labeled as: (1) a social trip component, (2) a business trip dimension, and (3) a personal business and shopping trip component. Both sets of factors were ro- tated to achieve "simple structure,‘ wherein each variable loads highly 86 on, and therefore is identified with,on1y one factor. Those factors to be rotated were determined by the standard criterion of possession of an eigenvalue of at least 1.0. The R-mode resulted in five factors, the Q-mode in three. Although both forms revealed similar activity linkage structures, the R-mode accounted for a higher degree of variation in the data and resulted in more factors. In another study, Hanson and Marble (1971) subjected a 26 by 26 matrix of travel linkages to a Q-mode factor analysis which yielded three identifiable components. Factor One identified a series of func- tions which were linked strongly with the home place, while a second factor linked the place of work with such functions as restaurants, banks, parking, and taverns. The third factor emerged as a shopping goods component which demonstrated functional linkages among retail activities. Both of the preceeding studies dealt with linkages among out-of- home activities which were defined by trip purposes, and therefore they differ from the activity types under investigation in this dissertation. Nevertheless, they do point up the utility of factor analysis as a means of identifying the underlying structure of activity linkages. A series of R-Mode factor analyses were performed on each of the ten matrices of transition probabilities for the Toronto sample in an attempt to isolate significant activity sequence patterns for the var- iously defined subgroups. This was done for economy of analysis and also in the hope that the factors would allow easy identification of the anchor points, or pegs, around which people build their daily routines. The factors which emerged from the analysis were not subjected to 87 rotation. This decision was made for two reasons. First, the forcing of each activity to load on only one factor was not thought to be realis- tic in view Of the belief that the activity categories were themselves broad summarizations of different but related activities. Different types of activities may form links in more than one sequence pattern. For example, going to sleep at night (private needs) may be closely related to evening activities such as entertainment, civic participation, and passive leisure. But other private needs, such as meals and personal hygene may form part of another unrelated sequence. Therefore, rotation was ruled out in order to allow each activity type to load on as many factors as necessary. A second reason for not rotating factors lies in the difficulty of interpretation which results when one is dealing with rotated factors. By the use of transition matrices and their subsequent factor analysis it is hoped that a pattern of activity sequences will arise for each subgroup in the sample. Major differences in activity sequencing which arise from the analysis will be compared across groups with the purpose of identifying significant variations in life style. These variations will be related to differences in location and residential environment in particular and to urban spatial structure in general. Chapter 5 DESCRIPTION OF THE DATA AND STUDY AREA Datg Source The data which serve the empirical objectives of this study con- stitute the output of a longitudinal survey currently underway in Toronto, Ontario. The research project for which the data are being collected in- volves an examination of the relationships between physical environment (here defined by housing type and location within the metropolitan area) and social attitudes and behavior. At a very general level two questions are being asked: (1) What kinds of individuals come to occupy a given type of housing or location, and (2) How can their subsequent behavior be related to their surroundings? This dissertation will concern itself ‘more with the latter question. Entitled, "The Physical Environment as Attraction and Determinant: Social Effects in Housing, the study is attempting to assess the impact of a change in residential environment on the pattern of life of the new resident. How does he adapt to his new surroundings? Do aspects of the environment influence behavior, associations, attitudes, and preferences, and to what degree (Michelson, 1969)? A central objective of the study revolves around the specification of broad life style groups from an examination of: (1) Interaction pattern within and without the local neighborhood (2) Descriptions Of self and neighbors (3) Voluntary association participation (4) Patterns of comercial behavior 88 89 (5) Time budget patterns (6) Participation in hobbies and pasttimes (7) Intentions of life in new residence (8) Perceived differences in personal life style vis-a-vis other residential settings (9) Why other housing was rejected (10) Influence of housing environment on expected life The scope of the housing study ranges far and beyond the research outlined in this report. Although the present research is designed to be integrated into the housing environment project no further mention *will be made of those study objectives except as they relate to specific aspects of this research. Variation l2 Physical Environment Since the overall objective of the parent study on the physical environment as attraction and determinant is to assess the impact of the physical environment on ways of living, the environment was systemati- cally varied for different groups in the sample. Specifically, the groups were defined by the particular environment they were moving into, while as many other variables as possible were held constant. The major variables which defined the physical environment are: (1) housing type lived in, and (2) access to the city center (Toronto). These combine into the four major environmental categories as described in Figure 5-1. In order to investigate the influence of both housing type and location, the study was designed such that both varied independently of each other and were clearly differentiated. Attention focused, therefore, on single family homes and high rise apartments as extremes in housing types, and on central locations and far suburban locations, thereby eliminating an intermediate zone (the outer city). In terms of Figure 5-1, cell I describes single family homes both a short distance 90 from and easily accessible by public transit to the city center, while cell II contains single family homes distant from the center (on the fringes of Metropolitan Toronto). Cell III is indicative of high rise apartments centrally located, and cell IV describes high rise apartment situations in the far suburbs. Access To City Center Close Distant Single Family I II Housing Type High III IV Rise Figure 5-1 CATEGORIES OF THE PHYSICAL ENVIRONMENT Through the cooperation of various segments of the housing in- dustry in the Toronto area an inventory of new and used, but appro- priate, housing units was compiled. Single family homes were considered appropriate if they were fully detatched and carried a selling prive of between $35,000 and $100,000. Appropriate apartments were those located in modern high rise buildings of at least five stories and with eleva- tors. Minimum acceptable rent was set at about $200 for two bedroom apartments, although location influenced this factor to a degree. At the time these decisions were made these figures were thought to 91 ‘represent the 'upper middle' portion of the housing market. Location found its definition partly through access to the city «:enter and partly through distance. A downtown location was Operationally defined as being within a ten minute walk of a subway station within the city limits (The city of Toronto is relatively compact, occupying only a small percentage of the total area of Metropolitan Toronto), or on a ‘bus or trolley line leading directly to the CED without transfer, or directly within walking distance to the CBD. Minimum travel time to the city center served to define a suburban location. The size of the suburban zone includes the major commuting areas serving downtown Toronto. A broad band of space between the downtown area and far suburbs lacked eligibility for inclusion in the study. Figure 5-2 displays the relevant zones of analysis in the metrOpolitan region. Figure 5-3 represents the distribution of locations of the sample prior to their move. Approximately one out of every six families had their initial location outside the Metropolitan Toronto area and are not displayed on this map. The locations of reapondents following their move are shown on Figure 5-4. Any families that relocated subsequent to the first move were retained in the sample only if their new destina- tion lay within the study area. Sampl e Sel ec tion A quite elaborate and exhaustive procedure was implemented which resulted in the sample of families participating in the above mentioned study. Although the sampling procedure was designed to meet the speci- fic requirements of that study, the selection process and the charac- teristics of the sample will be reported here as aids to understanding the empirical content of this research. _4 _ mmxtmz g zemmomam I 2392158 thomOh z<._._..0n_om._.w§ z_ mMZON 53.5 NA mason. .. . 533mm: 025on I 24.53; . 245qu a i ii 020.255 ‘ $3.25 .10 Ems; \\ skin” V 9288 z<50aoEms z. _ . .- .1. . .. . . a see a. see. mzo_I—vos_..1.mOn1 1 . . 1., TF.IOI .I (AM. o0 o o o Jim. o o Sn $181.1 .1. . 1.. .. .. . . 1...? 1.5.1.. / .12.. mu..- «.1: (.t... .1» . o O s _ .111... lowcw O.- .l.r|..l_.lln 11111 A. _ . _ 1 . 4.1 . ,r... . 1 . . . . 1!. 1 . . n. .r. . 1 ._ . o O _ 1. If”... _ X354 1 . O o . .w 1 -_ 1 . 1 . .. . 1|.1 .1 _ . . _ . O .1 o. . . . . . .J. 1. u .1 . . . x . 1 . . _ . . . . . .s... 1 L . . . f. Tuiiiiii- -!»..-......1...--. ............. 6 021.555 _ . . . ..1. 1 ._-I-I-I1i.1 _ .1 .1 _ .1 .1 m .1 .31: 02012on .1 1 .1 .1 ._ .1 .1 f x a Ill'llt (034mm1mm12 95 The needs of the physical environment study required as close to a one hundred percent sample of persons who were naturally moving through the relevant housing environments and who met certain criteria of family structure and age. The study is being conducted as a natural experiment, in which certain parameters are being measured before and after the natural application of an experimental stimulus, i.e., a change of residence (Michelson, 1970). The first consideration involved finding peOple who were moving to the environments of interest prior to their making the move. The objective was to interview the subjects while they presently resided in their pre-move environments. In order to isolate potential respon- dents, arrangements were made with real estate agents, builders, high rise rental agents, etc., to inform the investigators when appropriate families signed purchase agreements or rental leases. These families were then approached, and if agreeable, were interviewed for their in- clusion in the study (Phase 1). Each family was programmed to be inter- viewed two months following their change in residence (Phase II), one year after (Phase III), and four years after the move (Phase IV). The later three phases were to be conducted during the same time of year to control for any seasonal variation in activity patterns. At the time of this writing, data from the first three phases have been collected, coded, and made suitable for machine storage and processing. No restriction was placed on the residential location of the respondents prior to their move into the metropolitan area. However, in the interest of controlling certain key variables several restric- tive criteria were established which governed the selection of subjects to be included in the study. Respondents were selected on the basis of 96 their relative affluence vis-a-vis other segments of the population. This insured that housing choice clearly represented choice since, on the basis of income alone, they would have been able to choose any of the environmental situations under consideration. Given the systematic control over environment and a relatively homogeneous group of respondents in terms of degree of affluence, the decision followed to maintain as much homogeneity as possible by focusing on a single large cohort of people. They are married couples still in their child-bearing years with children (usually under 18). For com- parative purposes couples in the same age brackets, but without child- ren, were included. However, newlyweds settling in their first home were not sampled. As a final restricting criterion, areas of the city region with pronounced ethnic characteristics were excluded to control for any variation which may be attributable to that non-environmental condition. The inventory of potential respondents was compiled from diverse sources, both from the apartment rental market and the home buying mar- ket. A list of families moving into resale homes (all homes in the downtown area were resales) was generated, as well as an inventory of persons moving into new homes (The majority of moves into suburban homes were into new structures). Some suburban resales were included in order to assess differences between newly settled and established areas in the suburbs. A subsample of rented homes (originally desired) failed to materialize due to the absence of an available source of in- formation and the fact that in the Toronto area rental homes are not too numerous. Awareness of the possible incompleteness of the inventory of 97 respondents exists, but it is believed that as close to one hundred percent sample of all relevant movers was identified. Because of the tailing off of a housing boom as the inventory was commencing in June, 1969, the span of time required to fulfill the sampling quota extended to Christmas of 1970. During this period the number of rentals was low as families renewed leases rather than changed buildings. In addition, too few families with children sought downtown apartments during this period, thus delaying filling of the cells in the sampling design. As a result of this sampling procedure 989 families meeting the criteria for inclusion in the study were approached and of these, an acceptable rate of 77 percent consented to interviews. The final sample size for Phase I totaled 761 families. Phases II and III numbered respectively 687 and 593 completed family interviews. This completion rate is deemed acceptable and rests well within the maximum acceptable loss. The set of interviews conducted with each family consisted of the basic interview with the wife and supplementary ones with the bus- band and one child aged 10 to 17 years (if applicable). For any unco- Operative husband the wife completed basic financial information from his protocol. As a result, Phase I consisted of 527 interviews and 218 partial protocols for husbands. Of 202 eligible children, 166, or 82.percent, completed their interviews. The breakdown of the sample according to housing type and location is presented in Table 5-1. Tables 5-2 through 5-7 present data relating to the family structure, income, occupational, and educational characteristics of the 761 families interviewed in Phase I. The distribution of families Within the categories of housing type and location according to their 98 Table 5-1 NUMBER AND PERCENT OF RESPONDENTS BY LOCATION AND HOUSING TYPE, PHASE I Location Downtown Suburbs Total Single 94 272* 366 Family (12.4%) (35.7%) (48.1%) Housing Type High 109 286 395 Rise (14.3%) (37.62) (51.9%) Total 203 558 761 (26.7%) (73.3%) (100.0%) *Of this total, 209 were new homes and 63 were resales. stage in the life cycle is an uneven one (Table 5-2). Nevertheless, the frequencies within each major category are considered high enough for purposes of comparison. Information relating to age of children (Table 5-3), income levels (Tables 5-4 and 5-5), and occupational distribution (Table 5-6) indicates that the sampling procedure produced a group which met the criteria of family structure and affluence. Given the general level of affluence of the sample as a whole, the variation in the educational attainments of the respondents (Table 5-7) proved to be a mild surprise. In summary, in spite of the constraints which frus- trated the realization of the original sampling plan, an acceptable sample emerged which met the intended social criteria. 99 How AdHOH me o m.a H.mm ¢.¢¢ a.~ antenna as mono: cannon mom o.H o.oa m.m¢ m.mn o.¢ «banana a« mason owm ~.m ~.m ~.~m o.nm «.mm «nuance cw uaoauuuau em m.m m.m m.wm m.Hm H.mH csouczov um90£ 2: c6 m.m m.n m.m 12 success ucoauucmc oammua convawnu on um>o no mfl vaasu nmwvn magma n have: vaano couvawno o: HzmzzomH>zm Hugo“ .nm uo>o .nm “was: .mm have: .mm have: .mn have: onacm sumo now HmuOu mo N naHmo HmflUZDO>VQZ< mmHB ho m0< HzmzzbmH>zm ZOHH¢ZHHme Mm GAHmo HmMQZDow Qz< mMHz mo MU¢ Nam mamdfi lOO AwaHzzm Aao some now Houou no N zmmnqamo mo mm0< HzmzzomH>zm onadzHHmmn wm zmmaquo mo mmo< Mun mqm<fi 101 mom 0.0 o.a e.a m.“ o.m~ n.oo mm>H2 amo ¢.oH ¢.~H ~.oH o.o~ o.n~ n.o mazemmee assoc ooo.o~w ooo.o~w- mmm.qs- mam.HH- meow- coco» ua>o coo.maw ooo.~H» coca» coco» Hove: xom some now done» no N mzovzH mm>HB 92¢ mQdemDm mom MZOUZH mo onHDmHMHmHQ can mammH 102 TABLE 5-5 INCOME DISTRIBUTION BY FAMILY INCOME PER FAMILY ' DISTRIBUTION IN 1 not answered 7.0 under $6,000 2.2 $6,000 - 8,999 9.5 $9,000 - 11,999 17.0 $12,000 - 14,999 25.1 $15,000 - 20,000 23.4 over $20,000 15.9 TOTAL 761 103 TABLE 5-6 OCCUPATIONAL DISTRIBUTION FOR HUSBANDS AND WIVES OCCUPATION DISTRIBUTION IN I husbands wives manager 27.1 2.2 professional & technical 32.5 15.5 clerical 6.0 18.5 sales 9.6 1.8 service recreation 3.8 1.8 transportation & communication 2.4 0.4 craftsmen 13.5 2.0 other labourers 0.5 0.3 other 0.5 0.5 not in labour force 4.1 56.9 TOTAL . 761 761 1011 TABLE 5-7 EDUCATION LEVELS FOR HUSBANDS AND WIVES EDUCATION LEVEL DISTRIBUTION IN Z wives husbands elementary or less 5.7 4.6 some high school 22.2 18.3 grade 12 23.0 13.9 grade 13 12.0 12.1 vocational 4.9 5.9 some university or teachers college 17.2 12.9 undergraduate degree 9.5 14.8 post graduate training 7 5.4 15.8 other 0.3 1.6 TOTAL 761 761 105 Toronto: The Environmental Context The selection of the metropolitan region of Toronto, Ontario as the geographical unit within which to conduct the sampling program is inti- mately tied to the design and objectives of the physical environment study from which the data come. This fact in no way hinders or reduces the viability of the research reported on in this dissertation. To the contrary, the parallel objectives of both studies can only result in positive fallout for this study as a result of the sampling being con- ducted in Toronto. In other words, this researcher has not had to settle for a study area which is less than desirable in terms of the objectives of the design. The reasons which made Toronto an ideal location for the purposes of the physical environment study are the very same which make its selection preferable for this one. This section, then, will outline the rationale for the selection of Metropolitan Toronto as the context of the study. In addition, certain aspects of the growth and structure of the Toronto region will be outlined particularly those which bear relevance to the research objectives of this study. The existence of an ongoing study, such as that of the phy- sical environment as attraction and determinant, provided an available data file. Comparable data could not possible have been collected in such de- tail and for so large a metropolitan region by the author alone. However, other considerations were present which made the data particularly ap- propriate, besides their availability. Canadian cities have experienced processes of growth and change in a parallel fashion to urban areas in the United States. On the other hand, the Canadian situation exposes certain contrasting trends which present the social scientist with a unique laboratory situation for the 106 observation and study of alternative living patterns. While Toronto displays much of the promise and many of the problems of most metro- politan areas its size, the social and spatial structure of the city result from a continuous process of expansion which differs slightly from large cities in the United States in terms of the nature of inputs, outputs, and pace. In recent years several valuable works have appeared documenting the growth and structure of Toronto which make an extensive recapitu- lation of the city's geography and history both redundant and unnecessary. Of particular importance are Goheen's (1970) analysis of the social geo- graphy of the city during the latter half of the nineteenth century; Murdie's (1969) comparable social ecology of more recent decades, the investigation of Toronto's suburbanization (Clark, 1961, 1966; Seeley, Sim, and Loosley, 1956), the study of structure and change in Toronto's retail sector (Simmons, 1966), the process of redeve10pment and structural change in the central city (Bourne, 1967), and a comprehensive urban geography which emphasizes site, situation, and historical development in understanding the Toronto of today (Kerr and Spelt, 1965). In ad- dition to these works, the University of Toronto's Centre for Urban and Community Studies maintains an ongoing program of publications relating to the current spatial and social conditions in Toronto and its en- virons. In terms of age and size of the central city, Metropolitan Toronto more closely resembles the city regions of the northeastern United States than other parts of the country. In some ways, similar forces were operating in Toronto as in many U.S. cities, which account for the pattern of land uses and social structure which typify the modern 107 commercial-industrial city. Growth proceeded largely unencumbered by municipal planning authorities, resulting in a distribution of land uses which can be attributed to competition for accessible sites. Residential development seemed to have followed along the same lines as cities in the United States. Functional specialization was accelerated in the latter half of the nineteenth century when the intro- duction of streetcars allowed the wealthy to live in newly opened suburban areas adjacent to the transit lines and commute to work in the central city. Dif- ferentiation of residential areas became particularly apparent as the older housing stock close to the center of the city was inherited by newly arriving immigrants and the less well-to-do. (Murdie, 1969: 44) Innovations in intra-city transportation greatly accelerated the movement of higher status groups to suburban residential locations. Residential expansion proceeded unchecked (Murdie, 1969). Significantly, residential growth seemed to conform to Hoyt's sectoral model insofar as the high income sector expanded to the northwest, pulling total residen- tial development in that direction. Despite these similarities in residential development between Toronto and most urban areas of the northeastern United States, certain recent events provide some interesting contrasts between the two situ- ations. It should also be noted that urban and industrial growth in Canada is more recent than in the United States. Canadian cities such as Toronto, did not experience to the same extent . . . a paleo-technic period of industrialization. As a result they usually do not contain extensive areas of tenements and standardized industrial housing. (Bourne, 1967: 60). Despite the fact that in-migrants accounted for a large portion of the city's population increase between 1911 and 1941, by the latter year persons of British background still accounted for 78 percent of the total 108 population (Murdie, 1969). Toronto has experienced an influx of large numbers of European immigrants only in the past two decades. This con- trasts markedly with the experience of the United States in the late nineteenth and early twentieth centuries. Additionally, Toronto lacks a sizeable non-white minority and its attendant influence on residential change (Bourne, 1967). In terms of pOpulation densities, which are a reflection, in part, of differences in residential patterns, Toronto displays a slightly singular situation. For example, in 1941 the city of Toronto held 73 percent of the inhabitants of the metropolitan area. For 1951 and 1961 the figures were 61 percent and 42 percent respectively, indicating a marked growth of suburban regions not unlike that in the United States. However, in contrast to cities in the 0.8., where emptying out of the core areas is common, the absolute population of Toronto city varied by less than two percent during this period (Hill, 1970). These figures indicate that the central city retains its power to hold residents within its boundaries. As will be demonstrated, this condition is partly attributable to redevelopment in the central city. As a final contrasting condition in the Toronto experience it must be mentioned that a metrOpolitan form of government was introduced in 1953 which integrated the city itself with the twelve outlying munici- palities. One obvious positive effect of this occurance has been the facilitation of orderly growth and a reduction in the cost of essential public services (Bourne, 1967). By assuring better control over sprawl, by assuming responsibility for regional services, by equalizing assess- ment and taxation, and by providing for an integrated subway and express- way system, Metropolitan Toronto government has instigated rapid growth 109 of the outer suburbs (Murdie, 1969), while at the same time has halted the deterioration of the central city, and in fact assisted in its re- development (Bourne, 1967). As a general conclusion, despite these recent developments in Toronto's changing structure, the urban area suffers from many of the same pressures of growth which typify cities in the United States. How- ever, the particular combination of conditions which differentiate this city from others reveal a spatial distribution of housing types which allows for the systematic study of variation in behavior and its effect on choice of residential environment. A discussion of this situation follows. Residential Structure In Metropolitan Toronto no pure relationship exists between housing type and location. That is, the stereotypical situation where the sub- urbs are dominated by families in detached single unit dwellings and where the central city remains the domain of multi-family buildings and high rise apartments does not characterize the Toronto region. While it is true that the largest prOportion of single family detached dwellings existed in the suburban fringes in both 1951 and 1961 (Murdie, 1969), this percentage has been steadily declining in recent years, particularly during the decade of the sixties. A concentrated surge of apartment develOpment, particularly high rise apartments, has significantly con- tributed to this decline. Information relating to the percentage of dwelling units which were either single detached or apartments in both 1951 and 1961 for different sections of the metropolitan area serves to highlight the changing conditions in housing type. Table 5-8 presents the appropriate figures. 110 TABLE 5-8 PERCENT OF DWELLING UNITS WHICH WERE EITHER SINGLE DETACHED OR APARTMENTS FOR BOTH 1951 AND 1961 IN METROPOLITAN TORONTO REGION Metro City of Inner Outer Toronto Toronto___ Suburbs Suburbs 1951 1961 1951 1961 1951 1961 1951 1961 Single Detached 52.1 52.1 31.8 28.5 71.7 54.1 89.5 75.4 Apart- ments 22.0 28.9 29.7 36.5 14.9 23.1 7.8 24.4 Source: Adapted from Murdie (1969), pp. 56-57. The number of single detached dwellings as a percent of the total number of dwelling units declined in the city, inner suburbs, and outer suburbs. No change was recorded for MetrOpolitan Toronto as a unit. The greatest percentage loss occurred in the inner suburbs. Apartments, on the other hand, as a percent of the total dwelling units, increased in all areas, with a dramatic increase occurring in the outer suburbs. This movement to a less lopsided distribution of dwelling units from one dominated by single family houses commenced in the late 1950's. Prior to this, in the late 1940's and into the 1950's, when rates of family formation were rising, couples were marrying at younger ages, and birth rates were increasing, single family housing spread over a wide area. There were miles of one-story bungalow-type buildings closely hugging the ground. Changes occurred in the late 1950's. Lot sizes decreased, more semi-detached houses and row houses appeared and there was an increase in apartment building. (Kumove, 1966: l) 111 The increases in apartment construction have been so marked that today this type of structure constitutes the dominant form of new dwelling unit construction. In 1958, apartments comprised over 50 percent of all new units built. By 1961, the figure exceeded 60 percent, and by the end of 1965, 70 percent. The majority of new apartments have been built on vacant land in suburban Toronto (Kumove, 1966). Lest the reader be mis- lead, in the central city area of Toronto apartment construction has also been on the increase (See Table 5-8). The metrOpolitan wide development of high rise apartment construction is graphically displayed in Figure 5-5. Apartment clusters appear to be uniformly distributed throughout both the city of Toronto and its suburbs. This trend in new dwelling unit construction away from single de- tached units to multi-story apartment buildings cannot be attributed solely to the rising cost of land and increased building costs. Changes in locational preferences and in living conditions have played a signi- ficant role in altering the demand for different housing designs. Demand elasticities for new housing have changed, but to what extent is uncertain. Nonetheless, the greater mobility provided by rental accomodation, as well as, generally lower costs, built-in conveniences, and higher accessibility either because of central location or proximity to mass transit, have accelerated the shift to apartments. (Bourne, 1971: 325) Increases in apartment construction in the suburbs cannot be explained by accessibility to the central city. Other preferences must be finding fulfillment in suburban high rise structures which are unrelated to the location of amenities in the urban core. An analysis of activity pat- terns of the families who chose to reside in these dwellings should assist in isolating these preferences. In conclusion, it can be said that Metropolitan Toronto, with its mixture of single family houses and high rise apartments at all distances can :5 .258 £858 mote—Enos euaoeom «3.qu 225.23 n -m 5.8:. '0. ........ ,. Iwbomommcdm \C. 00000 ........ ........ ........ ........ ........ s9 successes ...... secs 0 0s f0 . I ”uOJuhs O ....... seem-successes: oe- os H so .0 safiss D’s". m......... .11? zomoe e. u ..... O I 0000000 ...... ..... 00000000 ..... I o. sssss 0000000 ..... O. OOOOOOOOO cos.“ a... . seed .flu ...... 0000000000 es 000 00000 sssss ssssss sssssss sssssssssss ssssss sosnu ...... ...... ...... ...... ..... sssssss sssssss sssssss so I I Q 0 so s sssss es ss- ssssssssss 113 from the CBD, offers the residential decision maker a variety of housing choices at different locations. Thus, he is not constrained by the lack of any one housing type at any location. He can more freely fulfill his life style aspirations and can act out his preferences at will. These conditions should produce different patterns of activities for each residential environment under study. As a result, activity patterns can be more easily analyzed as an aspect of preferred life style. 213; Budgets 9; Human Behavior The particular focus of this dissertation requires the collection of data in a rather unusual form, that of the time budget. Although, hardly a unique research design, the time budget remains a little known and even less used mode of data collection. It is by no means a mys- terious research instrument, and in fact is a straight forward means of collecting some of the most intimate information a respondent can relay. Simply, the time budget is a survey instrument on which respondents can provide information on how they spend their time over some specified period. This can be accomplished by means of either a diary format or an interview where recall is required. Typically, individuals are queried on what they did, how long each activity took, with whom it was performed, and where it transpired. With such data for a sample of peOple a variety of research questions can be pursued which are of a behavioral nature. Several such questions will be addressed in sub- sequent sections of this chapter and succeeding chapters. In a recent monograph on human time allocation, the authors la- mented the fact that, "For a nation nearly three quarters committed to an urban way of life (by defacto place of residence), we know surpris- ingly little about how American live this way of life . . ." (Hammer 111; and Chapin, 1972: v). Indeed, despite considerable knowledge of dif- ferences among various status groups, aggregate movement dynamics (e.g. social and spatial mobility), and local and national indicators, we maintain no accounting system of the patterns of life being acted out in our cities, nor are we as scientists able to, " . . . probe the transition of social phenomena from one state to another, to trace their meaning, and to gain insights into the dynamics of evolutionary shifts in life style systems as they become reconstituted into new systems under dif- ferent environmental influences" (Hammer and Chapin, 1972: v). Sur- prisingly, in spite of much research into behavioral processes, we suffer from an ignorance of human time allocation and the systems of life ways present in modern societies. Richard Meier first called for the maintenance of ". . . an index representing the variety of life, particularly the ways people choose to spend their time . . ." (Meier, 1959: 27). He saw the need for a set of social accounts which would reflect process and change in society in a measurable way. In this way the social efficiency of various urban and regional structures could be assessed. Such a set of accounts would parallel the existing national income accounting system. However, a social accounting scheme would consist of indices of human values and be supplementary to economic accounts as the basis for policy making and planning. Measures of the expenditure of time would assist planners in assessment of potential costs and benefits associated with particular schemes, the projection of social trends, and measurement of goal achievement (Meier, 1959). As the philosophy of environmental planning has shifted from a more architecturally dominated one, where emphasis was placed on determinis- tic postulates relating form to social behavior, to a value oriented 115 approach where planning action responds to values and desires of urban residents (Cans, 1968) research has been conducted into dynamics of time allocation and spatial behavior. Most notable among those involved in such efforts have been a group of planners at the University of North Carolina. Perhaps taking their cues from Meier's work (Meier, 1959, 1962) they have been conceptualizing human behavior as systems of activ- ity which suggest profound implications for the present and future form of urban complexes. Lest it be construed that the time budget methodology has been an innovation of recent years mention must be made of the fact that time and motion studies have been a concern of industrial management for a half century. Indeed, living conditions and expenditure patterns were the subject of social investigation as far back as eighteenth century England. The reader is referred to Hammer and Chapin (1972) for an extensive discussion of the evolution of time budget studies, and to Brail (1969). The objective here is to avoid redundancy and instead to concentrate on the methodology of time budget analysis. Since this study is not primarily concerned with time allocation but with the sequential and spatial nature of activities, a review of the literature on time allocation reduces in importance. As with all useful innovations in scientific research, time budget methodology has been gaining in popularity with scientists who embrace the social values or behavioral mode of explanation. In spite of some early works on time allocation (Lundberg, et al., 1934; Sorokin and Berger, 1939) only in recent years has a noticeable increase in time budget studies taken place (Hammer and Chapin, 1972). Certainly the most ambitious project attempted to date must be the International Time 116 Budget Research Project (Szalai, 1966) in which cross-national time budget surveys were simultaneously conducted in a dozen countries. Publication of the summaries of this work are still pending. Recent research employing time budget methodology include studies by the North Carolina group (Chapin and Hightower, 1966; Chapin and Logan, 1969; Chapin and Brail, 1969; Brail, 1969; Hitchcock, 1970), Michelson (1972), Pappas (1968), and Hungarian Statistical Office (1964). As a conse- quence of the increasing use of time budget methodology, a permanent secretariat has been established and charged with the duty of dissemi- nating current research information pertaining to time budgets. The primary device employed in the reporting and collection of data relevant to the objectives of this study is the human time budget. While the term time budget can be thought of as a description of individ- ual or group behavior (i.e. how persons allocate their time) it is also a survey instrument which can be used to plot human activities through time and space. Most researchers employ a form of the latter to gen- eralize from the data collected about the former. Data gleaned from such survey instruments provide information, not only about what activities individuals engage in, but also about who else is involved, when and for how long the activities take place, the typicality of the activities, and where the activities were acted out. In this manner a time budget describes the flow of activities through the course of a day (or longer) and over space. It provides a picture of the contact patterns of the individual as well as information about how activities are linked sequentially. In addition, information is provided with respect to the spatial distribution of individual behavior patterns. A further advantage of the time budget over, say, the 11? traditional transportation survey lies in the range of behavior which can be recorded. Whereas the latter records only out of home trips, and usually only vehicular trips to particular land uses, the former also provides data on in-home activities and all out-of-home activities, in- cluding walking trips. Time budget data, then, allow concentration on the set of activities which characterize a person, as well as on the temporal and spatial dimensions of that person's behavior. Time budget instruments can be of two general types. The first is organized as an open ended question on an interview schedule, where the respondent is asked when he or she awoke on a particular day (usually the day immediately prior to the interview), what activity commenced his or her day, and the nature of each subsequent activity. Length of time is introduced by asking the respondent the approximate duration of each activity. Both Hammer and Chapin (1972) and Butler et a1. (1969) em- ployed a form of this approach. An alternative form of time budget asks each respondent to recall the time of day that the initial activity and each succeeding activity began. The rationale for using this approach derives from the notion that individuals can be more accurate about starting times than they can be when giving estimates of elapsed times. The question is one of reliability in answers. Both this form of the time budget and the preceeding one may require that the respondent re- late only those daily activities which consumed a minimum amount of time (15 minutes, 30 minutes, etc.). A third form of time budget which has been little used in North America is the diary, where the respondent keeps a diary of his activities for some specified period, say, a week. Cooperation of subjects seems to be the major deterrent to the wide- spread use of the diary form. 118 Some research has been undertaken to assess the reliability and validity of various methods of collecting activity data. wa11déh (1968) employed home interviews, self-administered diaries, and telephone interviews to compare the relative ability of each form to secure re- liable results. The quality of information gathered did not differ sig- nificantly for the three methods when the time lapse between the per- formance of the activities and their recording was not great. When information was requested for extended periods of time, however, the diary method emerged as the most fruitful technique. In another ex- perimental situation the validity of survey methods for gathering activity data was assessed when compared with actual unobtrusive obser- vation of the respondents' movements and contacts. With the discrepency between observed time and stated time averaging only three minutes, the reliability of the time budget survey requiring recall on the part of the respondents was strengthened (Carlstein, Lenntorp, and Martensson, 1968). When recall of immediate past activities is desired (one or two days prior to the interview), interview techniques provide reliable information on the nature and times of daily activities. The longer the interval between the recording of activities and their actual occurrence the less accurate the responses. This situation, of course, increases the desirability of the diary method when information of several days duration is required or when the interview takes place much beyond the day(s) in question. The time budget format which has been used to collect the appro- priate data for this study was subdivided into segments of fifteen minutes in length. The respondents were asked by an interviewer to 119 relay all the activities they engaged in on the last weekday and previous Sunday that required a minimum of fifteen minutes. Questions included in the time budget elicited information about what activities were completed during the day, when they began and terminated, who else was involved, where the activity took place, and whether the activity was considered typical by the respondent. For analytical purposes, trips to and from activity sites were coded independently of each activity. That is, they constitute a separate activity, and can be analyzed as such. As a result the relationships between travel time, trip length, and time spent in an activity can be explored. A specimen of the time budget schedule is included as Appendix A. As a means of categorizing activities for purposes of analysis a slightly modified version of the 99 item code of the International Time Budget Research Project was employed. The code is appended for reference purposes. Chapter 6 ACTIVITY SYSTEMS AND RESIDENTIAL ENVIRONMENTS Methodology Chapter Six is devoted to an analysis of the activity linkages of the sample population. The methodology employed in this analysis em- bodies a multi-level approach. At the first level, interaction matrices will be described and compared on the basis of location and housing type. That is, all downtown residents, both male and female, in high rise and single family homes, will be compared with those living in the suburbs for differences in life style. First, transition matrices, and then interaction factors will be scrutinized in order to identify the major sequence patterns of each group. At subsequent levels of analysis sex, location, or housing type will be allowed to vary systematically while variation across groups in linkage and factor structure will be analyzed. In this manner differences in styles of living can be related to varia- tion in physical environment. In dealing with such a massive amount of data, and with numerous comparisons to be made between and among linkage coefficients and factor loadings, only apparent differences can be noted and analyzed. One of the difficult aspects of engaging in time budget analysis is manipu- lation of the volume of data. Figures 6-1 and 6-2 represent the daily time budgets of two sample families, both living in high rise apartments in downtown Toronto. These graphics display individual activity paths throughout the day, where activities have been collapsed down into one 120 121 'Q‘O'O' bs-O—s—g—o— e —. - a ._J. e —.—0 .- o ._.—0-0.. ._.-o—s—s—0 Wife Husband 24 16 MZHH cassava e>auesm unsound e>auo< seesaw» Iihuuam .uuwm u Ufl>HU some unsonvu 09'? nwum moan . -eoem 608 .uvawnu owu .ueoaoa £33 Figure 6-1 ACTIVITY PATHS OF SELECTED COUPLE 122 Husband ousowog u>maeem _ assume; _ o>Huo¢ _ passage» uuoucm _ nunmm . ow>wu _ _ . j j "._..- Wife '——"' Child .eowu umosvm .ouc> _41..|l...”..h.l. ”I”. .Il.||.||.4“.l|l.., naum . ”can _ _ -aoem _ ._. GUM _ _ _ -vaaeo . . . as rul— .3809 r Figure 6-2 ACTIVITIES OF SELECTED FAMILY WITH CHILD —.i-—————————— are: I I 10 p ' I 24 . 22 « 20 18 . 16 14 12 . 8 6 T 4 2 0 wEaH 123 of the ten major categories. Although such displays possess no analytical values, they demonstrate the complicated nature of behaviors patterns of individuals in an urban environment. For any average in- dividual a typical vector of activities may consist of twenty-five or thirty episodes. When these vectors are combined for several hundred individuals in a sample the data matrix becomes quite large, hence the need for reduction to more manageable proportions. When aggregating time budget data for various subgroups within a given culture the analytical problem becomes one of assessing significant differences in time usage and activity linkages. As Michelson (1972) has noted, the differences which emerge are not so much of kind as of degree. That is, for a homogeneous group, such as the one under study here, patterns of activity linkage may be similar for all subgroups, that is, their life styles may be quite similar. The main differences may lie in the extent to which certain subgroups emphasize particular sequence patterns over others. Such differences in emphasis, it is thought, emanate from the particular locational, housing, and life style characteristics of each group. Therefore, variations in life style by environment should be revealed by subtle but regular differences in linkage coefficients, factor loadings, and combinations of the two. The differences should be more pronounced at the lower levels of aggregation due to the control exer- cised over variations in location housing type, and sex. At the first level of analysis dichotomies in location and housing type constitute the categories of comparison. At this level the types of analyses which can be performed correspond to the first two stages in the analytical design outlined in Figure 6-3. In the first stage 124 TYPE I Location D S Housing HR T T Type 1 1 SF TYPE II Location D S D - HR <———-) s - Housing Type HR - SF (------3 SF - TYPE III Location D S HR Housing Type SF Figure 6-3 TYPES OF COMPARISONS POSSIBLE BETWEEN LOCATION AND HOUSING Downtown Suburban High Rise Apartment Single Family Home 125 (Type I) activity matrices can be compared for different housing sit- uations while controlling for location. In a Type II analysis loca- tional differences come under scrutiny while housing type acts as the controlled variable. The Type III mode of analysis becomes feasible only in the second and third levels of comparison. Here the variables of housing type and location can be varied or systematically controlled. Comparisons then become possible between polar situations in environment. Differences and similarities between high rise apartments downtown and single family homes in the suburbs can then be interpreted in terms of life styles of the residents. 531 E Life Style Differences The literature in both urban sociology and urban geography is replete with references to the differences in life style between those who live in the downtown sections of metropolitan areas and persons residing in the suburbs. As noted in a previous chapter, suburban living is thought to be built around a set of fundamentally different activities and attitudes from that of the city. As a result, with social class and economic status controlled, any significant differences in activity linkages between downtown and suburban households should re- flect differences in the influence of the physical environment. In light of current knowledge of suburban living and the role of distance in travel activity it is expected that, for the suburbanite, out of home activities such as shopping, education and training, civic participation, and active leisure would display strong interconnections. On trips outside the home, with distances being greater in the suburbs, families tend to link activities together on one trip more often than 126 (10 downtowners. Persons residing close to the city center rely upon jproximity to facilities (educational, cultural, etc.) to allow the freedom to move back and forth quite easily between in-home and out-of- home activities (Cans, 1968). Since differences between matrices cannot be determined to be significant in a statistical sense, some arbitrary level of difference is required such that some confidence can be achieved in interpretation. The figure deemed appropriate to consider as a significant difference between a single linkage category for any two matrices is .10. Any linkage that differed by at least that amount contributed to variation in life style. Any difference which ranged between .05 and .10 was noted and considered a moderate difference. For families living downtown, regardless of housing type, passive leisure seems to be a significant activity in terms of following other activities. (Table 6-1) Transition probabilities range from .08 for the link with working time to .20 for the link with active leisure. working time, domestic work, care to children, and private needs connect strongly with other activities, primarily following other activities. Care to children and private needs show strong linkages with domestic work and passive leisure. Domestic work, care to children, entertain- ment, and active and passive leisure are all strongly linked to private needs, a common pattern throughout most of the transition matrices. The connection can be explained by the fact that these activity types are primarily home-based, while work, shOpping, civic participation and education usually occur outside the home. Therefore, given the assump- tion of orderliness in human behavior, it seems logical that people would move back and forth between and among such home based activities. 127 soo V a co HN ON ~H ma ma ca NH sa ca mo Oa so HN mo oo Ho do No no so No no no on mN no no no no so so oo oo oo on oo oo oo ao oo oo No oo mo oo as no go No do #0 ss 5N ma oo 0H ow mo ma mm oo No No 00 oo no mo ms no No so so mo No Ma mo oa mo sN ma do sH mo no oo ma ma mN No no no oH Na Ma No ma mo no mo an «panama o>uemem saunas; o>wuo< ucoacamuuoucm coaueaqofluusm ow>qo mcficqena use cowumosvm moooz oue>wum mswmaonm convawno ou sumo xuos owumoaoo saga mcwxuoz ozouczoo u mowua>fiuo¢ xmvxooz mo>a3 was evenness mHzmHonhmou mumquA duo mammH .o~ .m 128 The linkage pattern for families living the suburbs shows some of the same strong connections between activities that characterize the city dwellers. (Table 6-2) To repeat the similarities seems unneces- sary; therefore, only the differences will be highlighted. Surprisingly, differences in sequencing between work, domestic work, care to children, and shopping do not appear to be significant; nor are those associated with private needs and active and passive leisure. Significant differ- ences do emerge for activities related to education and training and civic participation, the latter including meetings of voluntary assoc- iations. Suburbanites display stronger links between education and private needs and civic participation and private needs (.29 to .16 and .11 to .00). This gives rise to speculation that educational activities and those connected with organizations occur mainly in the evening; with the return home, suburbanites retire for the evening. For downtowners, civic participation is strongly linked with work, care to children, and entertainment. The differences between suburbanites and downtowners suggests that the former are more restricted in the time that they can devote to organizations and civic activities. The latter, on the other hand, can mix up these activities with others, indicating fewer re- strictions as to the time of day or sequence of such activities. On the whole, the differences in activity linkages between those living in the suburbs and those residing in central locations are not too pronounced. Their activity patterns display a remarkable degree of similarity. The few notable exceptions include educational and participatory activities, both considered discretionary forms of behavior. An analysis of linkage differences by housing type yields little 129 soo W n oo NH sN mo ma ON sH oH mg mg oo oH no NH No oo No No No no No Ho no so mm Ha oo no am no No 00 oo oo oo ss oo ao Ho oo oo oo Ho oo oo oo sm oo oo #0 oo oo ms om sa Ha mm mm oo ma mm mo No oo no no oo no as so so No ad mo mo oo no ma mo Na ma No ma ma Ha oo no mN sH ow ow no mo oo ma No No sH so No am ousmwoq o>wmmmm ousmaoq o>auo< ucoacfimuuoucm cowueoaowuumm ou>wo wcwcwmua one moausosom moooz oue>wnm wmwmoonm conoamno Om sumo xuoz owumoaon mafia wcwxmoz csnuonam u momuu>wuo< meoxooz eo>43 one mocsomsx mHzmHUHhhmoo MOassso H~ om no oo oo oo no as No no «sesame o>suos om mo no oo so No oo no so no unmeasuuuuuao om oo no on oo oo oo No om so coassosusuuso uo>so as No oo oo os AN No oo so oo unmeasue oes eoossuooo so Ho no so oo on so om mm no souoz uos>soo so No om oo oo No ms eo om so wasooooo om so so oo so so so No o~ so cauomsoo o» memo no No no oo oo on so me ow No soo: assomSoo so so No oo oo so no ~o No mo some oasxnoz oo o o a o m s m N a .o .n .s .n .N .a moaom hawamm oLMCwm moquu>wuo< hmoxomz mo>w3 one monsoon: MHZMHUHhmmou MUsssso no as No oo oo so go so so so ouossma s>suos om No no oo so No oo oo oo om somaesssroseo so oo so an oo so oo oo oo oo nosssosusuuso os>so as so oo oo os ow No Ho oo mo assesses sea cosusoaoo no No no oo No so No oo No om soouz ous>spo NH No on oo oo as ~s so om oo oesooooo om no No oo oo oo no sN no so assesseo ou sumo no no no so so No no No no so goes ususoaoo oo No so oo oo No no so no se mass mousse: om o o a o m s m N H .oH .m .o .m .s .m .N mucoauuoas seam swam moguw>muo< mmoxou: no>m3 ocs monsoon: mHzmHUHmmmoo mU¢MZHA sue MdmmH 133 «one broad socio-economic class may be significant. For these reasons, differences in life style by housing type, for the Toronto metropolitan area, do not appear to be significant. % S_ty_l_e _By Location At the second level of analysis control of one variable is intro- duced, and at the outset, this variable is location. A comparison of transition matrices for single family homes and high rise apartments located in central Toronto reveals a number of significant differences in linkages. The high rise group (Table 6-6) show stronger linkages between both private needs and work and active leisure and work, whereas for single family homes (Table 6-5) the link between civic participation and work is quite pronounced. This latter group shows much stronger linkages to care to children for private needs, education and training, and civic participation. They also link education and training, civic participation, and entertainment strongly with entertainment. The strength of these links suggests that care to children and entertain- ment form important points of activity during the day. They may define the rhythm of the day for single family homeowners. Among single family homeowners domestic work, entertainment, and active leisure are more strongly linked with themselves than for per- sons in high rise situations. One explanation for this situation might be that high rise residents spend smaller amounts of time in these activities and move on to activities of a different nature. Inhabitants of single family homes, perhaps more tied down by small children, must find large blocks of time for entertainment and leisure activities. Instead of linking domestic work strongly with itself and with care to 13h soo w u oo oa ad Na ma ma ma mo ma o« oo oH mo mm no oo oo oo #0 no no no no no ss mm oH so so mo no no oo oo oo on oo Ho oo «o oo #0 No oo so 00 as Ho do No 00 oo as NN no oo NA om ma no mo oo oo oo so mo os so no No muo m4mwuu< muoxoos mo>w3 onn monsnmnm snowmen o>aeonm ouneaog o>muo¢ unmanwnuuounm nouumomouunnm ow>wo mounaeua one nouunosom noooz oun>aum mnwmmonm nonoaqno on «one xuos oaumoaon saga onaxuoz .oa .o 13S soo W n oo NN no oo oo No es No No on no monsoon s>nmsso oN an oo oo oo on so so oo on snosnoa usages ~n No on oo so an no no so on sooacssunouom oo oo oo oo oo oo oo oo oo oo canusosonsnsn us>no on no oo oo on on no oo so no masseuse one eosusuaoo an no oo oo no on no so on nN soooz ass>suo on No oo oo oo on ss no on oo monsooom on no oo oo no No no o~ on no canonnoo on menu no so so oo no on no so an so 1003 onsssaoo so No no oo no oo mo oo no as many mangoes on o o a o n s o N n .o~ .o .m .s .m .N naounsoo u munoauuemn 03m swam 3.3.33»; knoxooz ooh“: one monsoon: mHZMHUHhmMOU mug ZHA one mqmfla 136 cfiaildren those in downtown high rise apartments link domestic activities astrongly with passive leisure. Since passive leisure activities lead 'to night-time retirement almost fifty percent of the time for this ,group, it can be deduced that housework and other domestic chores are left until evening before retiring to bed. Such a situation can be attributed to a high percentage of working wives with no children. On the suburban side of the coin it was expected that activity sequencing would not vary so much between housing types. This hypo- thesis was advanced because it is thought that suburbanites, regardless of housing type, are more like each other than they are like city dwellers of a similar housing environment. A comparison of the linkage matrices bears this hypothesis out. (Tables 6-7 and 6-8) Only two sets of linkages appear to be significantly different, civic partici- pation to domestic work and to itself. High rise dwellers link civic participation more strongly to itself (.57) than do single family home inhabitants (.38), although the reverse is true of the link to domestic work (.00 to .12). The relative similarities of these two linkages structures suggests that in the suburbs family life styles do not differ very much, even when housing type is considered. In the downtown situation the dif- ferences are much more pronounced. Stage in the family life cycle and the presence of a working wife would seem to explain this fact. 0f further note is the fact that the suburban single family home linkage matrix closely approximates that for all single family homes. The same is true of the high rise apartment situation. This would result from the greater percentage of the sample residing in the sub- urbs. 137 soo w .. oo NH no mo Ho oo Ns mo HH oH No onasHoH u>Hssso sN NH no oo oo NN no oH NH no «nomHoH o>Huos oo no on oo oo NH oo no NH NH uaoeaHsuuouem oN oo oH on oo oH oo oo NH No ooHssoHuHunso oH>Ho oH so oo oo on Ho oo so oo oo ooHoHsna oes eoHususoo NH No No Ho oo NN so mH oN sH souoz oum>Huo oH No oH Ho oo No Ns No oH so oaHaooom oH so so oo Ho NH so NN on so eonoHHoo on sumo oH No oo oo oo No so mH oN No one: UHHsoaoo oo Ho No oo oo HH no No No oo uaHe NeHxnoz oH o o N o o s o N H .o .m .s nonunonm a mason haaanm oawnwm nomuq>auo< anoxooz mo>w3 one monsoon: mHzmHonhmoo MUauo< mnoxooz 00>“: one monsoon: ouomfloq o>wmmnm unnamed o>wuo< unuanwnuuounm nouunmuowuunm oapwo wnwnwnua onn nowunonom moooz oun>wum wnwomonm nouoawno ou sumo xnoz owumoaon oEHH mnfixnoz .oH .o .o .s .n .s .m .N 139 In the third type of analytical comparison extreme combinations in the housing and location scheme are compared. (Tables 6-6 and 6-7) Since social conditions, housing type and location interact to in- fluence behavior to a certain extent, expectations exist that the most extreme differences in behavior patterns should manifest themselves in an analysis of linkage characteristics of the most unlike environments. Since most comparisons between urban and suburban living are character- ized by persons who live in high rise apartments and single family homes respectively, life styles should crystalize around these two situations. An examination of differences in linkage coefficients reveals that this is not the case. Differences in linkage coefficients occur in only four instances; the two matrices are remarkably alike otherwise. Families living in suburban single family homes have significantly greater linkages between care to children and domestic work and between private needs and domestic work than do downtown residents in high rise apartment buildings. They likewise link private needs more strongly to care to children. On the other hand, those in apartments link active leisure more strongly to work related activities than their suburban counterparts. This work linkage emphasis on the part of apartment dwellers strengthens the notion that many activities occur close to the workplace, in this case the central business district. Life Style By Housing Environment Another mode of analysis calls for control of housing type while differences in life style by location can be assessed. At first, single family homes in both the suburbs and central city provide the focus. 1110 (Tables 6-5 and 6-7) By investigating the effects of location on life style it is hoped that some inferences can be drawn concerning the effects of distance on activity linkages. Accessibility to the city center is hypothesized as having an inverse relationship to the strength of linkages for out-of-home activities. In the case of residents of single family homes this hypothesis is not supported. Of the eight linkage coefficients which are significantly different between the two matrices the downtown inhabitants show a pattern of stronger linkages for discretionary, out-of-home activities. An examination of the individual linkages reveals that downtowners link education more strongly to care to children and to itself, while the suburbanites link education and training more strongly to private needs. For the downtowners, strong linkages exist between civic par- ticipation and work, civic participation and entertainment, entertain- ment and itself, and active leisure and itself. For the suburbanites, a strong link shows itself between active leisure and passive leisure. All this suggests that suburbanites in single family homes are more homebound than downtowners and tend to link fewer activities on trips outside the home. They may substitute spending larger amounts of time on fewer activities because of the travel investment involved in getting to an activity sight, a finding documented by Hammer and Chapin (1973). Elsewhere, Michelson (1973) has found that homeowners living in the central city do so to avail themselves of the cultural amenities downtown, and to be near the husband's place of employment. Residents of downtown single family homes patronize places of public entertainment more often than other groups in the sample. In addition, these people feel that access to facilities is more difficult in suburbia and 1111 disagree with most in that they feel that suburban space is not superior for the purpose of raising children. A final observation concerns the difference between suburban and downtown residents in discretionary out-of-home activities. In most of the comparisons by location noted so far, both in home activities and activities considered to be obligatory show linkages of a smaller magni- tude across groups. This finding substantiates the claim that life style differences are reflected in patterns of behavior over which the individual possesses the greatest degree of control in terms of time spent, time of day, and sequencing. Residents of high rise apartments display greater similarities in life style between the suburbs and the central city (Tables 6-6 and 6-8). Here, housing type may be the greater influencing factor. This suggests that families who chose to live in high rise apartments probably select their location in response to location of the workplace and not on the basis of accessibility to cultural and shopping amenities. For both downtowners and suburbanites in high rise apartments, working time is linked with itself 70 to 75 percent of the time, whereas for single family residents the occurrence is usually 65 to 70 percent. Michelson (1972) found that for this sample in general persons moving to suburban locations, to both high rise apartments and single family homes, in- creased their travel time for the journey to work, although for hus- bands, moves to high rise apartments in both the suburbs and downtown resulted in a net decrease in travel time. Therefore, although resi- dential location is influenced by workplace location, housing type is most likely influenced by life style preferences, which seem to be realized in high rise situations. 1112 In terms of differences in activity linkages, suburbanites link domestic work more strongly to itself, shopping more strongly to care to children, and active leisure more strongly to domestic work, while downtowners link education and training much more strongly with itself (.51 to .33). Otherwise, the sequencing of activities for the two groups remains quite similar. The few differences reflect the slightly more family oriented life style of the suburban dweller, where the wife is less likely to be employed, and therefore more involved with children and housework. This accounts for the strong movements between care to children, shopping and housework. L$§£H§£lls Differences §1_§§§ To this point the analytical procedure has ignored differences in life style that may result from variation in sex roles. Previously, men and women were grouped together as though there were no differences in the roles they play within the family and in the urban community. By separating out the influence of sex differences a more accurate picture of the life styles of urban residents may be constructed. Life style, as conceptualized previously, reflected a dominant role orientation for each individual. Perhaps the most fundamental division that society makes in terms of role differentiation is that of sex. The manner in which roles are played out by individuals in various locations within the urban area should provide information on the spatial distribution of life style. Differences in activity patterns between suburban and central city wives can be determined along two primary axes (Tables 6-9 and 6-10). They relate to domestic activities in the home and activities involving 1113 soo w .. oo ma mm oH mN mm ma AH ma me no oH so mg no oo 00 do No no so No no no mm mm oo no no no so no oo oo oo mN oo oo oo ao oo 00 No oo no oo ws No #0 No Ho oo mm mm ma oo ma mm no ma so no no No no oo so no es no No mo 00 oH od mN oo sa mo nN sH Ho am mo mo oo 00 mm oH om sm mo No 50 oo oo oo NH no No N0 N5 nsounzoo n mouuu>auo< xnoxooa mo>H3 mHzmHDHhhmoo MOwmmnm ouanoA o>wuo<, unmananunounm nowusaaoaunnm oa>ao wanwenH one noaunonom moooz oue>qnm wnwmmonm nonoawno ou onno goo: owumoaon mafia wnwxuoz 1M; soo W n oo sH HN mo 0H NM NH mo oH sa so oH no ma No oo oo No No no No Ho no no an ad oo no He no No no No oo do on oo Ho do oo oo oo oo oo do oo mm oo oo oo oo oo mm mw 0H 0H no ma co ma NM no neonnnnm n eoHuN>Nuo< meoxoeB mu>a3 no mo Ad wo oo so ms so no mHzmHUHMhmOU MUfimeem unnamed o>auo< unmanaeuueunm nowuemaoaunem ow>fio wnqnwena one noaueonom eoeez eue>unm wnwmnonm nouoaano ou oneo xmo3 caueeEoo oENH wnaxnoz .oa .m 1115 education and training. In the instance of the former the activity linkage coefficients reveal that suburban wives firmly anchor much of their daily routine around housework and related domestic activities. All other activities, with the exception of work, link strongly with domestic work. Domestic work, care to children, private needs, and leisure behavior form the strongest sequences. This indicates that most suburban wives are bound closely to the home. With respect to these women Michelson has concluded that suburban women, especially those in single family homes, spend significant proportions of the average week- day alone, watching television, engaged in casual conversation, visiting, and at the homes of neighbors. This pattern results from conditions of isolation which plague suburban housewives. Downtown wives do not peg many of their activities around domestic work, in contrast to the suburbanites. Although downtowners and sub- urbanites do share some similar linkages with domestic activities, the downtown wives demonstrate lower linkages in general and significantly lower linkages for private needs, education and training, and active leisure. This situation is a reflection of the emphasis on out-of-home social behavior by downtowners. Along the second axis, differences emerge with respect to the emphasis on activities which follow education and training in a sequence. Downtown wives tend to follow with private needs and education and training more often, whereas the suburbanites emphasize domestic work and passive leisure, both predominantly in home activities. Once again we see suburban wives influenced in their sequencing patterns by home based activities, while the downtowners have a more variable pattern. lhé With respect to housing type, the differences between wives in high rise apartments and single family homes, with location allowed to vary, are negligible (Tables 6-11 and 6-12). This indicates that dif- ferences in life style are more subject to the influence of location than to housing environment. Wives in high rise apartments link civic participation more strongly with itself. In general, women in single family homes show stronger links between most activities and care to children, and a significantly stronger link for education and training. The pattern suggests the importance of child raising in the single family home, and substantiates the notion of the family oriented life style of single family home owners, whether in the suburbs or central city. Husbands appear to display a more pronounced variation in life style than wives, both locationally and by housing environment. In terms of differences by location, the most immediately apparent are the linkages for care to children, education and training, and entertainment with passive leisure (Tables 6-13 and 6-14). Suburban men link care to children and active leisure with passive leisure more often than husbands living downtown. The latter, on the other hand, follow education and training more often with passive leisure. This group also combines care to children more often with private needs, education and training more strongly with itself and active leisure more often with itself. Suburban husbands show stronger connections between entertain- ment and working time, education and private needs, and active leisure with passive leisure. Patterns which emerge from these differences in activity sequencing are difficult to interpret. Almost all husbands in the sample work, and 1117 soo W n oo ma sN mo mm 0H na na na ca no oH no mo do oo oo No Ho no no Ho so oo NM ma 00 no AH no no mo oo oo Ho os oo oo oo oo Ho oo Ho oo do oo ms ao oo oo oo oo os 5N oN ma ma mm mo oN mm no No mo so oo no no on so no so mo so mo oo oo Ha mo mm ma do sN AN ma oo Ha Nm sa ow RN 00 No wo mm oo no HA no so mm eunoaunees seam swam s eoquu>auo< meoxoo: mHZMHonmmoo modszA date MAmaeeem unnamed e>wuo< unoenweunounm noHuenaoHunem ow>wo wnmnqena one nowueonom eo>a3 eoeez oue>uum wnannonm nonoamno on oneo xnoz owueosoo mafia wnaxnoz .od .m 1h8 soo w u oo ma HN oH sH MA Na mo ca sH Mo oH Mo NH Mo oo oo No Mo so No #0 Mo Mo MM Ha oo Mo oA so No so Ho oo oo NM oo do No oo oo oo oo oo No oo Ns oo oo do 00 oo oM MN MH sH AH ma oo sH MM oH Mo mo oH Ha Mo mo Ms so so Mo Ma Na mo so ad ON mo NN ca No MN sa NH s~ co sM Ma MM NN co Ho oo Mo oo oo No No NH AN eoEo: Nawaem eHmnHm u eoquw>uuo< Neoxeoz mezmHlohmoo mosquA Name mqmdw anneaea e>aeeem managed e>wuo< unoenmeuuounm noNuemaoNunem 0N>No wnwnaens one nowueonom eo>H3 moooz oue>anm wnfimmonm nouoaqno ou oueo xmoa owumoaon QSNH mnwxuoz .o~ .o lh9 soo w u oo sN MH NH oo MN NN MH MN MN 00 OH Mo MN Mo oo No Ho No mo No No Mo MM MN oo Mo No oo so so oo oo oo on oo oo oo oo oo oo No oo No oo HM Mo oo oo oo Ho N... N oo oo 3 2 2 3. Ns No Ho Mo mo oo No Ho Ms Mo N0 M0 Ho no No oo oo no No ON Mo Ho no mo Mo oo No oo N0 M0 OH HO nsounsoo n oeHuA>Huo< Neoxeoz mHzmHUHmmmoo musszH MHno MHmsH mo oH AN Mw Mo sN HA HA so oN eonenoam eunerA e>Aeeem enneHeH o>Huo< uneanHeunounm noHueaHoHunem oH>Ho wanAenH one noHueonom eoeoz uue>Hnm mnHmnonm nonoHHno ou eueo xnoz oHueeaoo oaAH monnoz .M .N .H 150 soo w u oo MN AM no AN mo MA MA MM oM No oH Mo mo Mo 00 Mo Ho No Mo Ho Ho No Mo QM MH oo Mo NH Mo No No oo oo oo sM oo Ho No Ho Ho oo Ho oo oo oo NM Ho oo Ho Ho oo MM NN HH so 0M oM NH sH MM AH No N0 M0 oo oo Mo MM Mo so No No Mo Ho 00 Mo 00 so sN oH Ho so No Ho so oo oo 00 Mo Mo Ho nennnnnm u eoHuH>Huo< Neoxooz MHzmHonmmoo MUAemem onneHoH o>Huo< uneEnHeuneunm nOAuenAoHunem oH>Ao manHenH one noHueonom eoeez eue>Hum wnHom05M noHoHHno ou oneo xnoz oHueoaon oEAH wnAxnoz .M .s .M .N 151 so their weekdays are occupied by considerable amounts of obligatory time. Therefore, it seems logical to assume that differences in activity would be most pronounced in the categories of leisure behavior. Passive leisure forms an important juncture in the suburban husband's day, with strong connections to child related and active leisure be- havior. After a full day of work the suburban husband retreats to his domicile to enjoy the fruits of suburban living, namely family related leisure activities. The primary beneficiary of suburban living (especially for single family dwellings) is the husband, not the wife. Wives moving to suburban homes express the least amount of satisfaction with their daily schedules, with husbands registering an Opposite view (Michelson, 1973). That suburban husbands link education and training less strongly with itself and passive leisure but more strongly to private needs than do downtown husbands suggests that, for the latter group, this activity may be less time and space fixed. It is more normally part of their day, whereas the suburbanites link these activities with private needs fifty percent of the time. For suburbanites, educational activities constitute a significant set of behaviors, one which is usually the last important activity of the day prior to retiring. Ignoring location and concentrating on activity sequencing for residents of high rise apartments and single family homes, two differ- ences become apparent (Tables 6-15 and 6-16). Husbands in single family homes are more prone to link care to children with work while the high rise dwellers very strongly link child care to passive leisure. The former also follow shopping more often with care to children and education more often with private needs. Husbands in high rise apartments 152 soo w u oo MN MN MH oo ON MN MA Ms MN No oH so NH No oo No Ho M0 M0 so No No so MM MA oo Mo Mo oo so Mo oo oo oo MN oo Ho oo oo oo oo No oo Ho oo Ms Mo 00 oo No Mo Ho 3 HH NH 8 No 3 S on Mo No so Mo oo No No Ns No so Mo Mo so oo oo N0 M0 so oN HH Ho M0 N0 Ho oo N0 M0 Mo oo Mo Ho Mo NH MM oo MN Mo No Mo NN enanoH o>Aeeem onneAeA o>Huo< uneanHeuueunm noHueaHoHunem 0H>Ho manHeuH one nOAueonoM moooz oue>Hnm MnHamoLM nonoHHno ou oneo xnoz OHumoEoo NEAR mnAxnoB euneaunems oeHM nMHm n eoHuA>Huo< Neoxeoz eoneoenm MHzmHUHmmmou MOMMZHH MHIM mHmAeeem euneAeA e>Auu< oneEnAeuneunm noHueonHunem oA>Ao mnAnAenH one noAueonom mooez uue>Anm wanoonm nonoAAno ou oueu xnoz oHueoaoo oBAH wnAxuoz moeo: NAHeem eAMnHM a moHuH>Auo< Neoxooz eonenmnm mHzmHonmmoo mosszH oHno mammH .oH .M .M .s .M .N 15h tend to link shopping more often with itself. Conclusions which can be drawn from these differences remain elusive. On the surface at least it seems that differences in husband's life styles are more pronounced between locations as opposed to between housing types. For single family husbands, however, care to children is linked strongly with shopping and work, both heavily time and space fixed. High rise husbands have weak links to these sequences and prefer to combine child care with leisure activities. The patterns are in reverse of those which would normally be expected. The downtown wives in single family homes tend to link all activities with themselves more often than suburban wives, particularly educational activities, entertainment and active leisure (Table 6-19). This suggests that they devote larger blocks of time to these activities, perhaps as a result of their proximity to downtown Toronto. For the suburbanites there is a tendency to link entertainment more often with child care and private needs, evidence that these women are more tied to the home and must satisfy more of their leisure pastimes there (Table 6-17). Suburban women generally link all activities more strongly with care to children, an indication of the importance of children in the daily round of life. Wives living in high rise apartments, both downtown and suburban, show strong linkages for domestic work, care to children, and private needs with domestic work, but the links are significantly stronger for the suburban group (Tables 6-18 and 6-20). These women also show a stronger link between shopping and care to children. The downtown wives, however, demonstrate a greater propensity to combine child care with itself, an indication that they lack sufficient time to look after 155 soo W n oo sH NN oH MA MM NH Mo NA MA Mo oA Mo sH Mo oo oo No Mo No oo Mo Mo NM Mo oo Mo oH so No Mo Ao oo Ho MM 00 Ho Ho 00 oo oo oo oo oo oo NN oo oo Ho oo oo nenunnnm NHHaem eHMnAM a eoAuA>Auu< Neoxeez MM MN MH NA Mo MA Mo MA NM oA Mo Mo NH MH oo Mo Ms so so Mo MA NH Mo oo Mo MA Mo NN MA No MN MA 0N NH MA MM MA sM NN Mo MHzmHonmmou MUHeeem onneAeA e>Huo< uneEnAeunounm noAueerAunem 0A>Ho wnAnAenH one noAueonom mooez eue>Anm wnAmmonM nonoAAno on oneo xnoz oAueeaoo oaAH MnAxnoz eo>H3 .oH .M .M .N 156 soo w ... oo MH MM MO OO sA MH sH MO AN MO OH Mo Mo NO OO OO NO NO so so NO so OO NM OO OO MO Mo NO MO so OO oo OO oo OO OO OO OO OO OO NO OO oo oo Ms MO OO AO NO OO Ns on NN oo sN aN HH HN no MO so MO MO OO Mo Ns Ho Ho Mo Mo No OH oo OO MO OO MM MO oo ON OH MO OO OH sN MA NN MA M0 N0 MA NH OO oo NA M0 M0 Mo MN neonnnnm oeHM sMAM u eeHuA>Huo< Neoxous HZMHUHMMMOU MUMMZAH MHIM MAMMH enneAeA o>Aeeem onnerH o>Auo< unoanAeuneunm noAuenAUAunem oA>Ao mnAnAenH one noHueonoM ee>A3 eoeez oue>Aum wnAOmonm nouoHAno cu eneo xnoz UHueoaoo eaAH monnos .OA .M .M .s .M .N 157 soo W n oo MA NH HH MN NH MH Mo MA MO AH OH so OM sO OO OO OO No MO MO NO NO NO Ms MN NH so MO so MO NO OO OO OO MN OO HO OO AO OO OO NO OO NO OO Ms HO NO NO HO OO MM NN NO OO NH MN MO NH MM MO MO OO MO OO so MO Ms so so MO OH MA MO MN NH MN No AN MA MO MN NO MO OO OO MN MA MN MN MO MO OO OO OO MO MO HO HO ON euanoA o>Aeeem ennonH o>Auo< unoanAeunounM noAueerAuuem.0A>Ao mnAnAenH one noAueonoM eoeoz oue>Anm MnAmeonm nonoAAno on open sacs oAumoaon eEAH mnAxnoz nsounzoo NAAEem oHMnAM n eoAuA>Auo< Neoxmez ee>A3 MHZMHUHMMMOU musmzHH MAIM MHMAuo< Neoxeez MHZMHUHhmmoo MOMMZHH ONIM MHMAeeem onneAeA o>Auu< uneanAeunounm noAueaHoAunem 0A>Ao wnAnAenH one noAueonom mo>A3 moooz oue>Anm wnAnmonm nonoHAno on eneo xuoz oAueoEoO «BAH MnAxnoz It ”'1 159 children during various periods of the normal day. These women also link private needs strongly to work time, indicating a preponderance of working mothers. It may be that when this linkage is strong, so also will be that of child care to itself. There were no other significant differences between the two groups regarding other activity linkages. Any other differences that emerged were too minor to be considered meaningful. Wives living in single family suburban environments have a sig- nificantly higher probability of switching back and forth between and among care to children, private needs, and domestic work (Table 6-17). High rise housewives, however, demonstrate a much stronger linkage between private needs and work related activities, indicating the influence of work in organizing the daily schedule (Table 6-20). They also show a strong linkage between care to children and itself. These women dispense with responsibilities to their children at one time. This occurs perhaps as a function of pressures of work away from home and the limited amount of time spent in the home. In general, high rise wives who live downtown have stronger linkages between all activity categories and work, but they tend to be particularly strong for private needs, civic and collective participation activities, active leisure (which is usually performed away from home), and moderately strong for spectacles, entertainment, and social life. The pattern indicates that the working wives downtown are very active individuals who intermingle active leisure, social life, and voluntary participation with work. Wives who reside in single family suburban homes display signifi- cantly stronger linkages between care to children, private needs, 160 entertainment, and domestic work. Although the downtown wives in high rise apartments display strong linkages also, they are significantly lower than those for the suburbanites. These women also have strong linkages between both education and training and domestic work, and domestic work and itself, although less than the .10 difference. Another significant difference can be seen in the stronger link between private needs of the wife in the suburbs and care to children. This link is relatively weak for high rise women downtown. Such a pattern indicates that wives in the suburban single family home are closely tied to housework and caring for children, and must organize their discretionary activities, such as educational pursuits, voluntary associations, entertainment, and leisure, around these two more oblig- atory activities. For both groups of women all activities are strongly linked with themselves (main diagonals). This is to be expected since the activi- ties are repeated for every fifteen minute period when coded. However, some significant differences appear even in this case. For example, the single family suburban wives link civic participation activities tOgether more strongly than the high rise women. They do likewise with active leisure. The latter group, on the other hand, have stronger internal linkages for education and training. When given large blocks of discretionary time the suburban women seem to concentrate heavily on one form of activity. Finally, suburban wives show significantly higher linkages between care to children and passive leisure and education and passive leisure. The major differences in activity sequencing between single family downtown and high rise suburban wives parallel those for the single 161 family wives for both locations, with a few notable exceptions (Tables 6-18 and 6-19). Suburban wives in high rise apartment buildings exhibit a slightly greater work orientation than both single family groups in the suburban areas and downtown, although they tend to be slightly less work oriented than their downtown high rise counterparts. The high rise suburban group demonstrate higher linkage coefficients between most activities and domestic work, while the single family downtown group show a stronger linkage pattern with care to children. The limited number of husbands who responded to the time budget portion of the interview schedule precluded the breakdown of their linkage patterns beyond this stage. The number of husbands in downtown apartments and single family homes was too small to draw any meaningful comparisons. The linkage matrices at this level contained too many zero entry cells and displayed a highly irregular pattern, the result of a large sampling error. Clearly a larger sample of husbands was required. Linkage Dimensions In order to reduce the linkage matrices to a form whereby major patterns of activity sequencing could be identified, each matrix of linkage coefficients was subjected to factor analysis without rotation. This was done as a secondary form of analysis to the comparison of linkage matrices. Tables of factor loadings for each matrix are included as Appendix C. At the highest level of aggregation, where only the major cate- gories of location and housing type were compared, several major activity sequences could be identified for each group. The structures were quite similar for all four categories. In each case the first factor had high 162 positive loadings for domestic work, care to children, private needs, and passive leisure. High negative loadings were recorded for working time and entertainment, with the latter missing in the downtown category. This dimension describes a pattern of in home routine activities which follow themselves repeatedly. Factor scores indicate that these activities are highly linked among themselves. Interestingly, work and entertainment are most negatively associated with this factor. In fact, working time seems to form its own unique pattern of sequencing uncon- nected to most other activities. Shopping and entertainment form individual factors in all cases, and are preceeded by themselves most often. The factor structure for both downtowners and those in high rise apartments display certain similarities. The same is true for suburban and single family home groups. For those downtown and in apartments care to children links up with itself quite often, while for the latter two groups the activity is more closely associated with the first factor. In all four categories civic participation forms the nucleus of another factor and is associated negatively with work (indicating an evening activity) and positively with itself and shopping (except for downtowners). For all groups except downtown inhabitants the fourth factor identifies a leisure activities component. In this dimension active leisure is associated with itself and work for both those down- town and in high rise apartments. In summary, at the highest level of aggregation, the major dimen- sions of activity sequencing are quite similar across groups, except for those people living in central Toronto. The high rise apartment category (perhaps because of the number of inhabitants downtown) resembles the 163 downtown group along work and leisure lines. For each group, in home routine activities form a strong part of the daily pattern. This pat- tern dominates in both the single family homes and in the suburbs, most likely resulting from the influence of the wives' daily routine. At the second level of analysis tables of factor loadings have been constructed for the four possible combinations of housing type and location. Similar to the first level of analysis, all four combinations generated a first factor with positive loadings for domestic work, care to children, private needs, and passive leisure, and negative loadings for work and entertainment. The sole exception to this pattern was the group living in downtown high rise apartments. Both care to children and entertainment failed to load highly, and this factor accounted for a significantly lower percent of explained variance for this group than for any of the others. For people in downtown high rise apartments care to children does not form a part of the routine in-home sequence. In- stead, child related activities are related to civic participation and active leisure. This suggests that time devoted to children is reserved mainly for evening hours when both parents return from work or other outside the home activities. The priority of work in the lives of working couples dictates a pattern of activity sequences different from that where only one member of the family is employed. A factor common to all groups is associated positively with shop- ping and negatively with work. In all cases factor scores for work were highly negative and for shopping highly positive. Those in single family homes, both downtown and suburban, linked civic participation with this factor. For the members of this sample, then, shOpping activities were undertaken independently of the work trip (probably by wives) but 16b in conjunction with civic participation by single family home dwellers. Single family residents connect education and training differently than the high rise apartment residents, who linked educational activities with civic participation and active leisure. Factor scores indicate that these activities were strongly linked to themselves. The factor analysis of linkage matrices for both sexes revealed results which tended to support the previously determined activity sequences for men and women. For all husbands a strong work related factor was evident. Active leisure and work were related to this factor for suburban and high rise husbands in particular. In all cases for husbands the care to children, domestic work, private needs, passive leisure component was evident. This leaves no doubt as to the domin- ance of this pattern in everyday life. Cullen (1972) has concluded that routine personal and domestic chores held no structuring significance for the individual. Since people spent a considerable portion of their daily time at home, especially non-working wives in the suburbs, these activities are tied to a particular location. Evidence presented here from both the transition matrices and factor analysis tend to deny the conclusion reached by Cullen. Indeed, routine chores form a dominant part of the individual's life style, whether he or she realizes it or not. The wives' factor structure tended to substantiate this place fixed notion of activity sequencing. Both downtown and suburban wives placed in home routine activities on one factor and out-of- home, more discretionary activities on another. In summary, the factor analytic results do support the results gleaned from the tables of transition probabilities, but in some cases 165 are too gross to be interpretable. The level of aggregation of activ- ities may be too high to render factor analysis meaningful. The fol- lowing chapter attempts to sort out the meaning from the results of this analysis, and place the results in their proper theoretical context. Chapter 7 SUMMARY, CONCLUSIONS, AND PROSPECT Factorial Ecology and Social Area Analysis represent multi-dimen- sional descriptions of urban structure. Although Factorial Ecology constitutes an inductive approach to the fundamental structure of urban areas, three primary factors continue to emerge in the application of factor analytic techniques to urban data. They are: (1) economic status, (2) family status or urbanization, and (3) ethnic status. Social Area Analysis, on the other hand, postulates the existence of three similar constructs a priori, and states that urban regions can be dif- ferentiated by social rank, family status, and ethnic status. These structural approaches have had wide empirical testing and theoretical and conceptual articulation (Berry, 1971; Timms, 1971). Life style, as it has been conceptualized and operationalized in this research, relates theoretically only to the family status dimen- sion of Factorial Ecology and Social Area Analysis. In this sense, life style is a behavioral surrogate for family status or life cycle vari- ables (Bell, 1968). The application of time budgets within the spatial framework of central city-suburban differentiation served to test the social area hypothesis regarding variation in life style between city and suburb. An attempt has been made to operationalize the behaviors Bell (1968) claimed to be the determining factors in the selection of a residential location. 166 167 This dissertation has proceeded under the assumption that the residential structure of the city is incompletely understood in terms of the mechanisms which give it its present form. No one theory or hypothesis offers an adequate explanation or prediction of residential differentiation. The two major competing models which claim to explain the present situation in urban areas each offer partial explanatory power at different levels of scale. At the metropolitan wide level, the economic competition model sorts land uses into locations based on their ability to sustain costs at any location, with a fair amount of accuracy. The resulting land use profile is one of decreasing intensity with increasing distance from the urban core. The social choice model places increased emphasis on non-economic aspects of locational decision making, but focuses primarily on residential land uses. It posits a similar relationship between distance from the CBD and intensity of residential land use, but allows for deviation from this pattern on the basis of sentiment, status, life style preferences, and other conditions. Both hypotheses offer partial explanations of residential structure and therefore are inadequate in certain respects. Ultimately, a fussion of the two approaches into a unified theoretical system is desirable. Until such time as that evolves, research must continue into various aspects of the relationship between social structure and spatial structure. The isolation and study of particular variables is one means of accomplishing this. Both Social Area Analysis and urban Factorial Ecology result in an index of social organization relating to family status or stage in the life cycle, and by inference postulate a range of behavior. The behav- ioral categories can then be fit into different family status situations. 168 ‘What has never been fully Operationalized or tested is the existence of this range of behavior, commonly refered to as style of life. The methodologies of Social Area Analysis and Factorial Ecology represent structural approaches to an understanding of urban residential differentiation, whereas the research undertaken in this dissertation falls under the rubric of behaviorism. An investigation into human behavior has been used to test the validity of the life style dimension of the commonly accepted structural models of the ecological tradition. A discussion of the results is set forth below. But first, it seems useful to review the role that life style is thought to play in the process of residential selection. From the time that the University of Chicago ecologists first treated residential choice as a direct result of peOple's socioeconomic and ethnic identification to the determinants of residential selection treated in this dissertation the subject of where various groups live, and why, has undergone several major theoretical and empirical shifts of emphasis. According to the economic determinism model, groups of like income would reside in physical proximity (Park, Burgess, and McKenzie, 1925). The concentric zone, sector, and multiple nuclei models all incorporated this assumption to varying degrees. Later, housing came to be seen as a symbol of status. It was thought that housing was selected, not on the basis of space requirements, but as a means of improving status (warner £3,51L, 1957). Life cycle was introduced into the literature on residential selection by Rossi (1955), while Feldman and Tilly (1960) incorporated the educational component of status in the determination of what persons would live together in the same neighborhoods. Thus, the notion of ability to pay 169 became replaced by the broader notion of life style in research on neighborhood selection. This transformation led to the work of Wendell Bell (1958, 1968), who hypothesized that people select a location and housing type with some consideration of the patterns of behavior they desire to follow. The improvement of status became secondary to the desire that one be near like neighbors and live in a neighborhood which facilitated the practice of one's life style. Thus he postulated three broad categories of life style: familism, careerism, and consumerism, and linked them to specific locational and housing needs. Bell derived these life style preferences from the construct termed family status in Social Area Analysis. Bell failed to make clear the specific behaviors which would be found under each life style category, and directed his empirical tests to eliciting the preferences people had for moving to specific locations, not their actual behavior. Nevertheless, he has outlined a conceptual schema suitable for the study of residential selection and urban spatial structure. The focus of this dissertation has proceeded from the work of Bell, Michelson, and others by specifying that behavioral data can be employed in an alternative but supplementary analysis of urban residential structure. The purpose of this study has been to examine the relation- ship between the 1ife styles peOple practice and their locational and housing characteristics. Both Social Area Analysis and Factorial Ecology infer such a relationship, Bell hypothesized it, and this research attempted to test for it. The author sought to find support for or de- nial of the hypothesis that life style varied in metropolitan areas, for 170 both location and housing environment. The nature of the analysis assumed that persons living in similar locations followed similar life styles. The research question asked was whether differences were evident for major locational and housing categories. This study sought to relate life style patterns to the findings of Social Area Analysis regarding the spatial distribution of the family status index. Do families close to the center of the city behave dif- ferently from those located in suburban areas? What effect does housing type have on this location-life style relationship? Time budget data in the form of activity linkages were used as an operational measure of life style. The location variable was dichotomized into central city (downtown) and suburban, and housing type was limited to single family homes and high rise apartments. Differences by sexual gender were also assessed. The major conclusions drawn from this analysis now follow. Conclusions The conclusions drawn from the research conducted in this study can be divided into several categories. They deal with life style variation by location, by housing type, and by sex. A final set of conclusions relates to life style variation and urban residential structure and confirmation of the hypothesis connecting family status with location. When housing type and location are considered jointly the conclu- sion can be made that residents of single family homes in the suburbs and those in high rise apartments downtown exhibit the most extreme differences in life style. The former group most definitely practice the family oriented life style postulated in ecological theory, while 171 the latter tend toward a more cosmopolitan, diverse set of behaviors. Work and discretionary activities form distinctive sequence patterns for both husband and wife in downtown apartments. Routine domestic activities are linked closely together because of the press of time. Families in single family suburban homes typically include a non-working wife who remains at home with one or more children. This situation accounts for an activity sequence pattern that sees the husband linking many activities with work that the wife normally links with domestic and child care activities. The pattern for the wife is most dramatic in terms of lack of spatial mobility. Other data support these findings (Michelson, 1973). The pattern of activity sequencing exhibited by single family homeowners residing in Toronto proper can best be described as mixed. While both husbands and wives organize their homebound domestic behavior in a fashion similar to their suburban counterparts, their linkage patterns for such activities as civic participation, entertainment, and active leisure resemble that of the high rise apartment dwellers down- town. In some instances, then, their behavior represents a response to the central city location, and in others it reflects their fundamentally family oriented style of life. As far as high rise apartment residents in the suburbs are con- cerned, their patterns of behavior are more difficult to pinpoint., Since the differences in activity sequences for this group and single family suburban homeowners are minimal, it follows that their styles of living are somewhat similar. However, these people resemble their downtown counterparts to a degree, particularly with respect to work links. Factor analysis of activity linkages resulted in a simple 172 pattern of only three significant factors, a home based routine dimen- sion, an active leisure dimension, and a shopping dimension. The con- clusion must be made that the style of living of these people is more familistic than career oriented. Thus they give support to the thesis that location is the influencing factor. In terms of the fruitfulness of the various categories of location and housing type for pinpointing differences in life style, location seems to be a more sensitive differentiator. Variation in styles of living, as manifest in activity sequencing, is more readily seen when examining differences by location for both husbands and wives. The key variable operating to produce this pattern is family status, or life style as Wendell Bell termed it (Bell, 1968). Social Area Analysis and Factorial Ecology have demonstrated that family status varies concen- trically about the city center, with a trend toward larger families, smaller children, and a wife at home with increased distance from the city center. Although some differences in family status are evident by housing type as revealed by tables 5-2 and 5-3, differences in life style do not emerge as very pronounced between housing environments when location is controlled. This suggests that, at this level of aggregation, the home is not as important as the location in the realization of patterns of living. With regard to suburban homeowners, husbands enjoy a greater variety of different activity sequences between major categories, whereas wives display greater linkages between activities within categories. Thus, these husbands practice a more diverse life style, mixing up their activities more and enjoying greater spatial mobility as a result of 173 their place of employment being away from home. No wonder then that husbands express greater satisfaction than wives with suburban living. They travel outside the home more often, spend considerably less time alone, and tend to believe more strongly in the virtues of suburban residence as desirable for family life. Husbands suffer few of the negative aspects of location, while leaving the wife to cope with problems of isolation, large amounts of empty time, and boring activity patterns to fill the day. In support of this conclusion, Michelson (1973) has demonstrated that husbands make far more trips, travel further, and are more likely to visit the city center than wives. What makes these results more significant is that fewer husbands responded to the interview than wives. In general, husbands have stronger linkages between all activities and work than do wives. These differences are more pronounced for loca- tion than for housing type. The work linkage patterns of husbands and wives in high rise apartments tend toward more similarity, this due to the fact that the wife most likely works in addition to the husband. For wives, domestic work and care to children line up as important activities in the daily routine; the opposite is true for husbands. Since most husbands work during the day, they tend to link domestic work and child rearing activities more often to passive leisure, which is usually reserved for evening hours. The linkages between most activities and passive leisure are considerably lower for women than for men. women are able to spread passive leisure more uniformly throughout the day. Wives, as a rule, anchor their day's activities around care to children, domestic work, private needs, and passive leisure. Some do this more often than others, especially wives in single family suburban 17h homes. Wives in downtown high rise apartments seem to be less constrained by this pattern. The major difference is the low proportion of working women in the former category, and the high proportion in the latter. Husbands, on the other hand, order their day around two major forms of activity, work and passive leisure. Most activities lead into these two and to a lesser extent to private needs. Wives differ by location on which activities they link most strongly with domestic work, suburban wives being higher for private needs, education, and active leisure, and most other discretionary behavior. These women attempt to adapt their lifestyle more to loca- tional isolation and rely more heavily on each other for entertainment and stimulation. Downtown wives who most likely work and are less iso- lated trade off high linkages with domestic work for other linkages associated with outside the home activities. Husbands tend to differ with respect to the organization of their leisure time, especially in the areas of care to children, education and training, entertainment, and active leisure pursuits. Downtowners tend to have stronger outside the home links while suburbanites link up activities more often inside the home. The stratification seems to be along the lines of family orientation, with the more family oriented husbands residing in the suburbs. Life Style and Residential Structure The main theoretical question which has been addressed in this research report concerns the relationship between life style and the residential structure of the metropolis. Can the notion of residential differentiation be supported by time budgets of human behavior? As a 17S result of this research the answer is a qualified yes. An affirmative response is deemed appropriate in that family oriented life styles, as indicated by the high activity linkages among family oriented activities, tend to predominate in the suburbs. Within this locational category residents of both high rise apartments and single family homes demon- strate a family oriented pattern. In the central city where, according to ecological theory, more cosmopolitan life styles should prevail, residents of high rise apartments in Toronto subscribe to this pattern of living. Residents of single family homes who reside further out from the core, but still in the city, tend to divide their style of living between a familistic orientation and an urban-cosmOpolitan orientation. The reasons for this variation are several. The most important determinant of life style variation is the employment status of the wife. Behavior patterns differ significantly between families that contain a working wife and those that do not. Since the suburbs contain signi- ficantly fewer working wives the pattern of activity sequencing in that location tends toward a familistic orientation. In the central city, where the presence of working wives is high, behavior of both men and women demonstrates a more career oriented non-family life style. A second reason which is related to the employment status of the wife concerns the size of the family. Table 5-3 indicates that the number of children in each category of environment increases as one moves from apartment to house and from downtown to suburb. Suburbanites tend to have a higher percentage of younger children, while older children predominate in the central city. The number of families with no children exceeds 75 percent in downtown apartments but reaches only 22 percent 176 in the suburbs. The percentage of families with no children is low for all single family homes. Families with children, particularly small children, tend to live in the suburbs and in a single family home, and peg their life style around home, child care, and leisure. Families residing downtown, particularly in apartments, tend to be childless or have older children, and emphasize work, entertainment, and active leisure in their life style. Both single family homes downtown and high rise apartments in suburban areas house families that display a more familistic life style. The downtown homeowners enjoy the fruits of both single family living and a central city location. Those in suburban apartments have children in the younger years and a wife at home caring for them. A familistic behavior pattern follows from these conditions. Evaluation A qualified yes has been given to the confirmation of the major research hypothesis on the basis of the findings. The qualification enters the picture when the utility of the methodology employed comes under scrutiny. Certain aspects of the categorization of residential environments, activity sequencing, and the use of transition probabili- ties and factor analysis served to place limitations on the generali- zability of the findings. First, the use of behavioral data within a behavioral conceptual schema to test the validity of certain findings of Social Area Analysis proved fruitful. Too often researchers have made statements about human behavior on the basis of the characteristics of aggregates of people. By taking a strictly behavioral approach to the study of life style and 177 residential structure this research has strengthened the previously held views regarding the relationship between social and physical space. The use of time budget data in an analysis of this sort has not been reported in the literature, and so this dissertation represents a significant departure from traditional research in urban ecology. As a result, certain weaknesses in the methodology are now apparent. One of these concerns the aggregation of activity data into major behavior categories. As with all aggregate schemes much variation within groups of activities was lost with this procedure. 0n the other hand, this facilitated the handling of the activity sequencing matrices. Time budget data are very powerful in the analysis of behavioral patterns because they provide a complete picture of an individual's daily (or longer) activities. The problems generated by this type of data include the sheer number of data elements to be manipulated, the dif- ficulty of applying traditional modes of analysis to the data, and the tremendous variety of possible behaviors which can be recorded. As a result, some form of data reduction is required. In this study, since the author was restricted to only certain analytical procedures, the problem of activity sequencing seemed to be handled best by the transition probability matrix of Markov Chains. The modeling of activity sequences by means of Markov Chains was deemed inappropriate because of the necessity of assuming independence in decision making, an impossibility in this case. In so far as the ana- lytical procedure required a summary measure of activity sequencing, transition probabilities proved quite functional. However, the function was one of aggregation and description but the analytical approach was comparative. The power of statistical test was absent because of its 178 incompatibility with the matrix approach. Future research must resolve this dilemma. Factor analysis was employed as a means of reducing the data still further and of finding the underlying dimensions of activity sequencing for the several groups in the sample. The results were inconclusive, calling into question this approach. Factor analysis indicated very little in the way of differences in activity sequencing between various groups of Toronto residents. This may have resulted from the already highly aggregated nature of the input data. Since differences in life style may be more of degree than of kind, factor analysis may have been inappropriate as a differentiator. In summary it can be said that a behavioral approach to the under- standing of urban structure can be beneficial when coupled with existing generalizations about spatial form. It is only the possibility of combining these studies of spatial form with statements about behavior over space that gives promise of a distinctive and potentially rele- vant body of geographic theory. (King, 1969: 593) Speculations about the relationship between urban form and spatial process fill the literature in geography and related disciplines, but only in recent years has research in this vein been conducted on any scale. The testing of hypotheses of spatial form through the use of behavioral data is one means of verifying and clarifying this relation- ship. It is believed that the analysis of activity linkages conducted in this dissertation has provided some verification on the relationship between residential location and life style. That does not mean to say that this research has settled the issue once and for all. Indeed, much additional research is required, and I79 time-space budgets provide an appropriate data form upon which further research may be based. Although time budget research is at an early stage of develOpment significant findings are emerging as a result of the application of a variety of methodological approaches. Additional methodologies which could be and have been applied to time-space budgets include the mapping of human spatial movements by time of day. This would provide information on how people sequence their activities in both time and space. Bullock, Dickens, and Steadman (1971) have attempted to locate individuals with both time and space coordinates for a sample of students at Reading University, with a view toward planning for more efficient use of space. Hagerstrand's notion of time-space prisms constitutes a model of society with the purpose of guiding urban and regional planning and locational policies in general. He advocates simulation as a means of better understanding time-space relationships and their effect on human decision making. Currently, Cullen and Nichols (1970), Bullock, Dickens, and Steadman (1971), and Stephens (1973) are directing them- selves toward this effort. Within the context of urban residential structure a more precise breakdown of the location of individuals in the metrOpolitan area is desirable when searching for variation in life style. The categories of downtown and suburban are extremely general and therefore offer a lower level of generalization about the relationship between life style and location. Cell counting comes to mind as one means of determining additional discrete locational classes when dealing with sample survey data. As far as life style is concerned, the results of this research 180 indicate that a more comprehensive definition is required. Activity sequences should be complemented by other operationalizations of time and space use, such as the amount of time devoted to various activities, what the individual would rather be doing at any time, spatial movements and others. Michelson's (1973) approach incorporating expectations regarding time usage and environmental preferences as well as actual behavior present a better picture of an individual's value structure and hence life style. In any case, a first step has been made here to relate behavior patterns to urban residential structure. Life style forms a key axis of social differentiation, and therefore exerts a profound influence on spatial behavior and ultimately spatial form. The investigation con- ducted here reaffirms some notions about spatial form but suggests that further research is in order. APPENDIX A TIME BUDGET INSTRUMENT 181 APPENDIX A TIME BUDGET INSTRUMENT Now I'd like to get some idea of what an average day for you might in- clude. Let's take yesterday. (Last weekday) USE RESPONDEN'S OWN WORDS AS MUCH AS POSSIBLE. a. At what time did you get up? (CIRCLE TIME) Then what did you do? How long did it take you? Was anyone else with you? Where did you do it? Were you doing anything else at the same time? BEGIN BY ASKING THE RESPONDENT THE TIME HE GOT UP THAT MORNING AND WHAT HE DID FIRST. ASK HOW LONG THIS ACTIVITY TOOK AND THEN RECORD IT ON THE SHEET AT THE APPROPRIATE TIMES. ASK WHETHER OR NOT ANY- ONE WAS WITH THE PERSON WHEN HE DID IT: DETERMINE WHETHER THE PRESENCE OF THE OTHER INDIVIDUAL(S) WAS INCIDENTAL OR WHETHER HE WAS ASKED TO ACCOMPANY THE RESPONDENT SUCH THAT IT WAS REALLY AN INTERACTION PROCESS. ASK WHERE THE ACTIVITY TOOK PLACE AND WHETHER OR NOT THE RESPONDENT WAS DOING ANYTHING ELSE AT THE TIME. IF A PERSON GOES TO A STORE OR SOMEWHERE, BE SURE TO CHECK THE TIME TO THE STORE, SHOPPING AND HOME AGAIN. FINALLY, ASK THE RESPONDENT WHAT HE DID NEXT AND THEN REPEAT THE ABOVE LINE OF QUESTIONING. DO NOT ATTEMPT TO RECORD WHAT WAS BEING DONE AT EACH TIME GIVEN ON THE SHEET; RATHER USE THE SHEET SIMPLY AS A METHOD OF CALENDARING THE DAY'S ACTIVITIES. 182 APPENDIX A THE MINIMUM TIME SPAN TO BE CONSIDERED FOR AN ACTIVITY IS 15 MIN- UTES. RECORD THE START AND STOP TIMES TO THE NEAREST QUARTER HOUR. TIME SPENT TRAVELLING TO QR_FROM.AN ACTIVITY IS A SEPARATE ACTIVITY AND SHOULD NOT BE INCLUDED WITH THAT ACTIVITY. 183 APPENDIX A TIME BUDGET INSTRUMENT WEEKDAY___ Doing TIME What do? Who else? Where? anything Typical? else? 6:00 am 6:15 am 6:30 am A 6:45 am 7:00 am €§ 2:30 am 2:45 am 3:00 am APPENDIX B TIME BUDGET CODE 00 Ol 02 03 04 OS 06 07 08 O9 10 ll 12 13 14 15 16 184 APPENDIX B TIME BUDGET CODE Working:Time and Time Connected To It (00 - 09) Normal professional work (outside home) Normal professional work at home or brought home Overtime if it can be specifically isolated from 00 Displacements during work if they can be specifically isolated from 00 Any waiting or interruption during working time if it can be specifically isolated from work (e.g. due to supply shortage, breakdown of machines, etc.) Undeclared, auxiliary, etc. work, wives-children unpaid members to assist family Meal at the workplace Time spent at the workplace before starting or after ending work Regular breaks and prescribed non-working periods etc. during worktime Travel to (return from) workplace, including waiting for means of transport Domestic Work (10 - 19) Preparation and cooking of food, putting away groceries washing up and putting away the dishes Indoor cleaning (sweeping, washing, bed-making), general non- specific housework Outdoor cleaning (sidewalk, disposal of garbage) Laundry, ironing Repair or upkeep of clothes, shoes, underwear, etc. Other repairs and home operations, packing & unpacking, washing or repairing car 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 185 APPENDIX B Gardening, animal care, walking dog Heat and water supplies - upkeep Others (e.g. dealing with bills and various other papers usual care to household members, etc.) Care to Children (20 - 29) Care to babies, feeding baby Care to older children Supervision of school work (exercises and lessons) Reading of tales or other non-school books to children conversations with children Indoor games and manual instruction Outdoor games and walks Medical care (visiting the childrens' doctor or dentist, or other activities related to the health of children) Others Not to be used Travel to accompany children including waiting for means of transport Purchasing:of Goods and Service; (30 - 39) Purchasing of everyday consumer goods and products, shopping Purchasing of durable consumer goods Personal care outside home (e.g. hairdresser) Medical care outside home Administrative services, offices, bank, employment agency, customs, etc. Repair and other services (e.g. laundry, electricity, mechanics), car wash 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 S3 186 APPENDIX B Waiting, queuing for the purchase of goods and services, house or apartment hunting Others, signing lease or contract to buy Selling house or house contents; showing own house Travelling connected to the above mentioned activities, including waiting for means of transport Pgivate Needs: Meals and Sleep etc. Private and Non-described Activities (40 - 49) Personal hygiene, dressing (getting up, going to bed, etc.) Personal medical care at home Care given to adults, if not included in household work Meals and snacks at home Meals outside home or the canteen, essential other than 70-79 Night Sleep (essential) Daytime sleep (incidental), long time, e.g. 1 hr. 9 Nap or rest, 1 hr. or less Private activities, non-described, others. (using sauna alone) Travelling connected to the above mentioned activities, including waiting for means of transport Adult Education and Professional Training (50 - 59) Full time attendance to classes (undergraduates or post-graduate student), studies being the principle activity Reduced programs of professional or special training courses, driving lessons (including after work classes organized by the plant or enterprise in question) Attendance to lectures (occasionally) Programs of political or union training courses 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 187 APPENDIX B Home work prepared for different courses and lectures (including related research work and self-instruction) Reading of scientific reviews of books for personal instruction, specific to own profession Others No response, no further activity No secondary activity Travelling connected to the above mentioned activities, including waiting for means of transport Civic and Collective Participation Activities (60 - 69) Participation as member of a party, of a union, etc. Voluntary activity as an elected official of a social or political organization Participation in meetings other than those covered by 60 & 61 Non paid collective civic activity (e.g. volunteers) Participation in religious organizations Religious practice and attending religious ceremonies Participation in various factory councils (committees, commissions) Participation in other associations (family, parent, military, etc.) Others Travelling connected to the above mentioned activities, including waiting for means of transport §pectacles,_Entertginment, Social Life (70 - 79) Attending a sport event Circus, music-hall, dancing, show, night-club (including a meal in entertainment local), parade Movies 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 9O 91 188 APPENDIX B Theater, concert, Opera Museum, exhibition, library (ed. purposes) Receiving visit of friends or visiting friends, relatives at airport Party or reception with meal or snack offered to or offered by friends, relatives Cafe, bar, tearoom Attending receptions (other than those mentioned above) Travelling connected to the above mentioned activities, including waiting for means of transport Sports and Active Leisugg (80 - 89) Practice a sport and physical exercise Excursions or drive, hunting, fishing (if husband teaching wife to drive) Walks, browsing, window shOpping Technical hobbies, photography and developing, collections Ladies' work (confection, needlework, dress-making, knitting, etc.) Artistic creations (sculpture, painting, pottery, literature, writing poetry, etc.) Playing a musical instrument, singing Society games (cards, etc.), crosswords, board games, chess Others Travelling connected to the above mentioned activities, including waiting for means of transportation Passive Leisure (90 - 99) Listening to the radio, piped music Watching television 92 93 94 95 96 97 98 99 189 APPENDIX B Listening to records, tape recording Reading books Reading reviews, periodicals, magazines, pamphlets, etc. including proof reading done at home Reading newspapers Conversations, including telephone conversations Writing private correspondence, reading mail, writing in diary Relaxing, reflecting, thinking, planning, doing nothing, no visible activity (arrive home, use on Sunday if long interval between activities) Travelling connected to the above mentioned activities, including waiting for means of transport APPENDIX C TABLES OF FACTOR LOADINGS 190 H.a a.a o.oa o.sa a.ma ¢.sn «unusua> emaaaaaxm oqam. ouamqmq 0>wmmmm «can. unsmqoa m>wuo< oaks. “000.- uaaaaaauuoucm mmsm. coaumaaoauuam on>ao «nae. Name. mascaaua a coauuuaem Nook meomz ouu>aum aamm. mama. mcaaaoam «Nan.- monk. emueaaao as «use omen.- «was. sues oaummaoa swso.- ~mso.- «any sewage: H> > >H HHH HH H ocuumm cannonsm mmDHUDMHm MOHU woodmanxm mama. ousmumq m>ammmm mamm.u mooo. ousmqmq m>wuu< comm. unmacqwuumucm NNoo. aowuuaauauuam ua>ao meow. (onmo. wcgawuua a newuuosvm mqwm. mvmoz oum>aum ammo. mcqanonm Nwmo. memm. couvawno 0H sumo amen. ooqu. sup: owumoeon mm¢o.- ommm.- «any manage: H> > >H HHH HH H Hauomm asoussoo MMDHoDmHm MOHU voaaaaaxm cams. «unanwa o>ammmm woke. «uaaaoa o>wuu< mass. Hmsm.- uamacauuuoscm onus. consensuauuam ow>wo mmmn.- cana.- meanness a conuaoavm Boom. meooz usa>aum mama. mawaaoam News. swam. amuesaao 0a «use «mam. soo: unusuaoo Nwmn. skoo.- omom.- mafia manages H> > >H HHH HH H nouuum muamauuca< swam swam manhosmym moau vacaaaaxm «Nam. seek. «panama u>ammmm mack. «panama o>auu< cmoo. smmm.o unmasqsuuousm onom.- Nsmm. swam. consensuauuam ua>ao mmmm.- ENOQ. meanness a coauuoaum Esau. auouz uuu>aua sasm. manaaoem awn». couuaaso as «too mmwm.u noun. xuoz owumoaoa asan.- osmo.- mafia sedate; H> > >H HHH HH H pounce moaom hawewm onch MMDHUDMHm dOHomm #ao mqmqmmmm nook. «Cannon o>auo< mmom.u unmaoqmuuoucm maem.- mafia. coauaawuauuam oa>no some. wcaanuua a sensuous» HNQS. aeomz oua>wum momm. wcwnnonm samm.- «Oak. couuaaao 09 «use ammo.- mack. sup: oasmoaoa omom.- «smo.- mafia «aways: H> >H HHH HH H possum smousnam mafiamm oawcam MMDHUDMHw moao vuawaaaxm III came. muammwa u>ammam mooa.- «panama o>auo< moam.u uooacqmuumuam coaummauauumm uw>do ammo. meanness a coaumuaom Hun». avooz ous>wum mwms.- waaaaoam comm. omuvawno OH sumo Nmmw. xuos oaumoaoa moq5.n mafia wcuxuoz H> >H HHH HH H nouuam csoucson zdaamm mamcam amoeommam moao emaauaaxm mmmm. ousmwmg u>umamm unsound m>wuo< mmmm.u unmasquuuoucm coauuaHOMuuwm ow>wo wouowmua a coaumosvm mNMN. avooz ouu>aum weak. waaaaonm nook. canvanno as sumo mmmm. xuoz caumoaon memo.u omoo.u saga woqxuoz H> > >H HHH H uouuam canuansm «use swam MMDHUDme MOHU vochmem momm.u oHom. -¢m.u wmom. NHmm. mmwn. sack. mwxm. o¢0n. mmwm. ousmHoA o>Hmmam ousmwog o>Huu< unoaswauuuusm coHuunHoHuuum 0H>Ho wchHmuH a coausuavm «mam. avouz ous>Hum mcwamosm coupHHso 0H sumo man. xuoz oHuumaon n30 . - «a: 338: H> >H HHH HH :sousson swam smHm mmsuosmam moao voauoHaxm Hon. moss. muamHmH «>Hmnmm Home. ousmHmA o>Huo¢ Humm. usuaaHsuHouom cams. aoHuaaHuHuHuH uH>Ho mama. waHaHuHH a aoHuqonem sNNm. mummz uuu>HHH wswmnonm coHeHHno oH «one name. HmHm. any: oHusuEOH ooam. onm.- oaHH maquoz H> > >H HHH HH H Houumm :30uason u moHuH>Huo< mmvxowz mo>H3 mMDHUDMHm moaocm muo made 0 xHszmm< |’\1ll\'| I 199 NH.oH mm.NH oo.¢H a~.NN m¢.om mocmwum> moamemxm NHom. 50mm. mmHm.u son.u Nomm.. Hon. «com.u mace. ~mom.u Home. Howw. «HNm. qwmm.o ousmHoA o>HmmMM musmHma o>Huu< unmaonuHuucm coHumaHoHuHmm 0H>Ho moanuoua a couuaoanm moooz oum>HHm wcHamonw ooHvHHnu oH sumo xuoz oHumoaon mags wcHxHo3 H> > >H HHH HH ownusosm n.mmHuH>Huu< mmvxmmz mm>H3 mmDHUDMHm MOHU moaumHaxm Hnmm. moon. own“. noes. Hmom.u mwmo.n mums. mmmn.u mums. 0mm“. mmmo. oon. ousmuuq s>ummwm ouswwoq o>Huu< uooeaqmuuoucm coHquHouuumm 0H>Ho wnHaHmuH a :OHumosvm moooz mum>uum wcHAmonm couoHHno OH oumo xuoa UHHmoEon oEHH mcuxuoz H> > >H HHH HH omHH amHm - msHuH>Huo< Hmexmma mm>H3 MMDHUDMHm MOHU umaHuHaxm osoo. Hon. ousmHmH ~>Hmmmm HOHQ. Homm. ouamHoH u>Hso< noon. unoaouuuuoucm mm~u.- coHuaaHoHuumm oH>Ho sHHn.- Hana. waHaHauu a couuaoavu quH. mummz ou¢>HuH Name. maHaaoam cuss. asHuHHao 0H «use Hmmm.- Hams. sup: oHuuoEon AHsH.- mmmm.- oaHH ”cusses H> > >H HHH HH H Houomm mason mHHamm onch s moHuH>Huo< mmoxmoz mm>H3 mMDHUDmHm MOHo uocHuHaxm Jilwomm. muamHoH o>Hmmmm ousoHoA 0>Huo¢ mmoo.u unmasumuumuom mosm.- coHuoaHoHuuaH uH>Ho osmm. ~04m.- ammm. maHaHsHu a aoHuaosvm onH. «nowz ouu>HuH some. moo“. mowamoam HHHH. couvHHao cu ouau oomu. mean. sup: oHuooaon comm. oaHH maquoz H> > >H HHH HH H Houumm omHm ame - necesmsm mmsHoaeum Houoau MHno mqm euchHaxH News. muamHoH 0>memm mace. ousmHoA o>uuo< omoo.u unsecucuuoucm somn.- asmo.- :oHumaHoHuuaH uH>Ho nmmn.- HHHm. mchHsuu a aoHuaosvm Hmau. muumz oum>HHH ooHH. mcHaaoam Hmmm. HmHm. amueHHao 0H sumo coon. xuos oHummaoa Homn. mews wcuxuoz H> >H HHH HH H Houomu :mnunosm u mocmnmom MMDHODMHm MOHodm anU mqm osoHsHaxm [I meow. suSmHsA s>ummwm qqom. suSmHsH s>uuu< meow. ooHo NmHm.n moHcasuH w noHusosom Hamm. mossz sus>uum snum. wcuamonm asuoHHno os suso “Nmm. xuoz ouumsaon wmqo.u HHHn.u sBHH wcuxuos H> > >H HHH HH H uouomm czounsoa u mocsouom MMDHUDMHm MOHUdm mHno mqde O xHszmm< BIBLIOGRAPHY BIBLIOGRAPHY Abu-Lughod, Janet 1969 "Testing the Theory of Social Area Analysis: The Ecology of Cairo, Egypt," American Sociological Review, 34: 189-212. Alonso, William 1965 Location and Land Use: Toward A General Theory of Land Rent. (Cambridge: Harvard University Press). Anderson, T.R. and J.A. Egeland 1961 "Spatial Aspects of Social Area Analysis," American Sociologi- cal Review, 26: 392-398. Ansbacher, H.L. 1967 "Life Style: A Historical and Systematic Review," Journal 2: Individual Psychology, 23: 191-212. Bell, Wendell 1958 "Social Choice, Life Styles, and Suburban Residence," in William M. Dobriner (ed.), The Suburban Community, (New York: G.P.Putnam's Sons), pp. 225-247. Bell, Wendell 1968 "The City, The Suburb, and A Theory of Social Choice," in Scott Greer, Dennis L. McElrath, David W. Minar, and Peter Orleans, (eds.), The New Urbanization, (New York: St. Martin's Press), pp. 132-168. Bergen, Bennet M. 1966 "Suburbs, Subcultures, and The Urban Future," in Sam Bass Warner (ed.), Planning for A Nation of Cities. Cambridge: MIT Press, pp. 143-162. Berry, Brian J.L. 1965 "Internal Structure of the City," Law and Contemporary Prob- lems, 30: 111-119. Berry, Brian J.L. 1971 "Introduction: The Logic and Limitations of Comparative Factorial Ecology," Economic Geography, (Supplement) 47, 209-219. Boskoff, Alvin 1970 The Sociology 2; Urban Regions (Second Edition), (New York: Appleton-Century-Crofts). 205 206 Bourne, Larry S. 1967 Private Redevelopment gf the Central City, Research Paper Number 112, Chicago: Department of Geography, University of Chicago. Bourne, Larry S. 1971 "Apartment Location and the Housing Market," in L.S. Bourne (ed.) Internal Structure 2; £hg_City, (New York: Oxford University Press), pp. 321-328. Brail, Richard K. 1969 "Activity System Investigations: A Strategy for Model Design," Unpublished Ph.D. Dissertation, University of North Carolina at Chapel Hill. Britton, John N.H. 1971 "Methodology in Flow Analysis," East Lakes Geographer, 7: 22-36. Bullock, N., P. Dickens, and P. Steadman 1971 "The Modelling of Day to Day Activity Patterns," Architectural Desi n, 1: 292-298. Burgess, Ernest W. 1925 "The Growth of the City," in R.E. Park, E.W. Burgess, and R.D. McKenzie (eds.), Egg City, (Chicago: University of Chicago Press), pp. 47-62. Burgess, Ernest W., and Donald J. Bogue (eds.) 1964 Urban Sociology, (Chicago: University of Chicago Press). Burns, Thomas 1968 "Urban Styles of Life," Working Paper Number 6, London: Center for Environmental Studies. Butler, Edgar W., F. Stuart Chapin Jr., George C. Hemmens, Edward J. Kaiser, Michel A. Stegman, and Shirley F. Weiss.. 1969 Moving Behavior and Residential Choice, (washington: Highway Research Board). Carlstein T., B. Lenntorp, and S. Martensson 1968 Individars ngnsbanor i_Nagra Hushallstyper, (Lund: Lunds Universitet, Institutionen for Kulturgeografi och Ekonomisk Geografi). Chapin, F. Stuart, Jr. 1965 "The Study of Urban Activity Systems," in Urban Land Use Planning (2nd ed.), (Urbana: University of Illinois Press), Chap. VI. Chapin, F. Stuart, Jr. 1968 "Activity Systems and Urban Structure: A WOrking Schema," Journal g§_the American_Institutg.g£ Planners, 34: 11-18. 207 Chapin, F. Stuart Jr. and Richard K. Brail 1969 "Human Activity Systems in the Metropolitan United States," Environment and Behavior, 1: 107-130. Chapin, F. Stuart Jr., and Henry C. Hightower 1966 Household Activity Systems: A Pilot Investigation. (Chapel Hill: Center for Urban and Regional Studies, University of North Carolina). Chapin, F. Stuart, and T. H. Logan 1969 "Patterns of Time and Space Use, in H.S. Perloff (ed.), Th2 Quality 2; the Urban Environment, (Baltimore: Johns Hopkins Press), pp. 305-332. Clark, S.D. (ed.) 1961 Urbanism and the Changing Canadian Society, (Toronto: University of Toronto Press). Clark, S.D. 1966 The Suburban Sociegy, (Toronto: University of Toronto Press). Cullen, Ian G. 1972 "Space, Time and the Disruption of Behavior in Cities," Environment and Plannin , 4: 459-470. Cullen, Ian, Vida Godson, and Sandra Major 1971 "The Structure of Activity Patterns," Paper read at the European Meetings of the Regional Science Association, London. Cullen, Ian, and Vida Nichols 1971 "A Micro-Analytic Approach to the Understanding of Metropolitan Growth," (Unpublished Manuscript) Joint Unit for Planning Research, University College, London. Duncan, O.D., and B. Duncan 1955 "Residential Segregation and Occupational Stratification," American Journal gg Sociology, 60: 493-503. Fales, Raymond C. and Leon N. Moses 1972 "Land Use Theory and the Spatial Structure of the Nineteenth Century City," Papers, The Regional Science Association, 28: 49-820 Feldman, Arnold S., and Charles Tilly 1960 "The Interaction of Social and Physical Space," American Sociological Review, 25: 877-884. Firey, Walter 1945 "Sentiment and Symbolism as Ecological Variables," American Sociological Review, 10: 140-148. 208 Gans, Herbert J. 1962 "Urbanism and Suburbanism as Ways of Life: A Re-evaluation of Definitions," in Arnold M. Rose (ed.), Human Behavior and Social Processes: Ag Interactionist A roach, (Boston: Houghton Mifflin Company), pp. 625-648. Gans, Herbert J. 1968 "The Suburban Community and Its Way of Life," in H.J. Cans, People and Plans. (New York: Basic Books). PP. 132-140. Gerth, H., and C.W. Mills (eds.) 1958 From Max Weber: Essays in Sociolo , (New York: Oxford University Press). Glazer, N. and D.P. Moynihan 1963 Beyond the Melting Pot. (Cambridge: The M.I.T. Press). Goheen, Peter G. 1970 Victorian Toronto: 1850 £2 1900, Research Paper Number 127, Chicago: Department of Geography, University of Chicago. Golledge, Reginald G. 1969 "The Geographical Relevance of Some Learning Theories," in K. Cox and R. Golledge (eds.), Behavioral Problems i3 Geography: A Sygposium. Evanston: Northwestern University Studies in Geography, No. 17, pp. 101-145. , Greer, Scott 1962 The Emerging City. (New York: The Free Press). Hagerstrand, Torsten 1970 "What About People in Regional Science?" Papers, The Regional Science Associgtion, 24: 7-24. Hammer, Phillip, and Stuart F. Chapin 1972 Human Time Allocation: A Case Study 2; washington, D.C., (Chapel Hill, Center for Urban and Regional Studies, University of North Carolina). Hanson, Susan, and Duane F. Marble 1971 "A Preliminary Typology of Urban Travel Linkages," East Lakes Geo ra her, 7: 49-59. Harris, O.D., and E.L. Ullman 1945 "The Nature of Cities," Annals g; the American Academy 2; Political and Social Science, 242: 7-17. Harvey, David 1973 Social Justice and the City. (London: Edward Arnold). Havighurst, R.J. and K. Feigenbaum 1959 "Leisure and Life Style," American Journ§1_g§ Sociolo , 64: 396-4040 209 Hemmens, George C. 1966 The Structure 9; Urban Activity Linkages, (Chapel Hill: Center for Urban and Regional Studies, University of North Carolina). Hemmens, George C. 1970 "Analysis and Simulation of Urban Activity Patterns," Socio- Economic Planning Sciences, 4: 53-66. Hill, Frederick 1970 "Spatio-Temporal Trends in Population Density: Toronto, 1932- 1966," Research Paper Number 34, Center for Urban and Com- munity Studies, University of Toronto. Hitchcock, John R. 1969 "Urbanness and Daily Activity Patterns," Unpublished Ph.D. Dissertation, Department of City and Regional Planning, University of North Carolina at Chapel Hill. Horton, Frank E., and John F. Hultquist 1972 "Urban Household Travel Patterns: Definition and Relationship to Household Characteristics," East Lakes Geographer, 7: 37-48. Horton, Frank E., and Paul W. Shuldiner 1967 "The Analysis of Land Use Linkages," Highway Research Record, Number 165: 96-107. Horton, Frank E., and William E. Wagner 1968 "A Markovian Analysis of Urban Travel Behavior; Pattern Response by Socioeconomic-Occupational Groups," Highway Research Record, No. 283: 19-29. Hoyt, Homer 1939 The Structure and Growth 3; Residential Neighborhoodg_in, American Cities, (Washington: Federal Housing Administration). Hungarian Governemnt 1965 The Twenty-four Hours 9; the Day: Analysis 2f 12,000 Time Budgets, (Budapest: Hungarian Central Statistical Office). Kerr, 0., and J. Spelt 1965 The Changing Face 2; Toronto, (Ottawa: Geographical Branch, Department of Mines and Technical Surveys). King, Leslie J. 1969 "The Analysis of Spatial Form and Its Relationship to Geographic Theory," Annals, Assogigtiog_g§_American Gegggaphers, 59: 573-595. Kofoed, Jena 1970 "Person Movement Research: A Discussion of Concepts," Papers, 135 Regional Science Association, 24: 141-156. 210 Kumove, Leon 1966 "A Preliminary Study of the Social Implications of High Density Living Conditions," Toronto: Social Planning Council of Metro- politan Toronto, Research Paper. Lowry, Ira S. 1964 "A Model of Metropolis," Research Report RM-4035-RC, Santa Monica: The Rand Corporation. Lundberg, G.A., M. Komarovsky, and M.A. McInerny 1934 Leisure: §_Suburban Study, (New York: Columbia University Press). McElrath, Dennis C. 1962 "The Social Areas of Rome: A Comparative Analysis," American Sociolggical Review, 27: 376-391. McElrath, Dennis C. 1965 "Urban Differentiation," Law and Contegporary Problems, McElrath, Dennis C. 1968 "Societal Scale and Social Differentiation: Accra, Ghana," in Scott Greer, et a1., (eds.) The New Urbanization, (New York: St. Martin's Press), pp. 33-52. Meier, Richard L. 1939 "Human Time Allocation: A Basis for Social Accounts," Journal 2: the Americgg Institute 2f Planners, 15: 27-33. Meier, Richard L. 1962 A Communication Thqungf Urban Growth. (Cambridge: MIT Press). Michelson, William 1969 "The Physical Environment as Attraction and Determinant: Social Effects in Housing," Research Paper Number 22, Center for Urban and Community Studies, University of Toronto. Michelson, William 1970 Man and His Urban Environment: A Sociologicgl A roach, (Reading, Mass.: Addison-Wesley Publishing Company). Michelson, William 1972 "Discretionary and Nondiscretionary Aspects of Activity and Social Contact in Residential Selection," Paper read at European Coordination Center for Research and Documentation in the Social Sciences, Brussels, Belgium. Michelson, William 1973 "Environmental Change: A Report to the Central Mortgage and Housing Corporation on Results from the Project, The Physical Environment as Attraction and Determinant: Social Effects in Housing," University of Toronto. 211 Michelson, William, David Belgue, and John Stewart 1972 "Intentions and Expectations in Differential Residential Selection," Paper presented to Symposium on Effects of Resi- dential Mobility on the Wife, University of Indiana-Purdue University Medical Center at Indianapolis, June 12, 1972. Michelson, William, and Paul Reed 1970 "Theoretical Status and Operational Usage of Life Style in Environmental Research," Research Paper Number 36, Center for Urban and Community Studies, University of Toronto. Moore, Eric G. 1972 "Residential Mobility in the City," Commission on College Geography Resource Paper No. 13, Washington: Association of American GeOgraphers. Moriarty, Barry M. 1970 "Locational Preferences and the Pattern of Residential Change in the Lansing-East Lansing, Michigan Metropolitan Area," Unpublished Doctoral Dissertation, Department of GeOgraphy, Michigan State University. Murdie, Robert A. 1969 Factorial Ecology pf Metrppolgppp Toronto, 1951-1961, Research Paper Number 116, Chicago: Department of GeOgraphy, University of Chicago. Nourse, Hugh 0. 1968 "Regional Economics, (New York: McGraw-Hill Book Company). Nystuen, John D. 1967 "A Theory and Simulation of Intraurban Travel," in W.L. Garrison and D.P. Marble (eds.), Quantitative Geography Part I: Economic pdeCultural Topics, Evanston: Northwestern University Studies in Geography, Number, 13. Ogburn, William F., and Meyer F. Nimkoff 1958 Sociology (Third Edition), (Boston: Houghton Mifflin Company). Olsson, Gunnar 1969 "Inference Problems in Locational Analysis," in K. Cox and R. Golledge (eds.) Behgvioral Problems ip Geography: A Sypposium. Evanston: Northwestern University Studies in Geography, No. 17, pp. 14-34. Pahl, R.E. 1970a Whose City:‘ And Other Essays pp,Sociology and Planning, (London: Longman Group, Ltd.). Pahl, R.E. 1970b Patterns pg Urban Life, (London: Longman Group, Ltd.). 212 Pahl, P. 1968 "Time Allocation Study," Research Report Number 6, Athens: Athens Center of Ekistics. Park, Robert E., Ernest W. Burgess, and Roderick D. McKenzie 1925 The City. (Chicago: The University of Chicago Press). Quinn, J.A. 1950 Human Ecology. (New York: The Free Press). Parsons, Talcott, and Edward A. Shils (eds.) 1951 Toward 5 General Theory pg'Action. (Cambridge: Harvard University Press). Pred, Allan 1967 Behavior ppg_Location: Foundations for g Geoggpphic ppd Dyggmic Location Theory. Part I, Lund Studies in Geography, Ser. B. Human Geography, No. 27, The Royal University of Lund, Department of Geography. C.W.K. Gleerup. Rees, Phillip H. 1968 "The Factorial Ecology of Metropolitan Chicago," Unpublished Masters Thesis, Department of Geography, University of Chicago. Rees, Phillip H. 1970 "Concepts of Social Space: Toward an Urban Social Geography," in Berry and Horton (eds.), Geographic Perspectiyes pp Urban Systems, Englewood Cliffs: Prentice-Hall, Inc., Chapter 10, pp. 306-394. Riesman, David 1958 "The Suburban Sadness," in William M. Dobriner (ed.), The Suburban Communigy, (New York: G.P. Putnam's Sons), pp. 375- 408. Robson, Brian T. 1969 Urban Analysis: A Study 2; City Structure, (Cambridge: Cambridge University Press . Rogers, Andrei 1967 "Theories of Intra-Urban Spatial Structure: A Dissenting View," Land Economics, 63: 108-112. Rossi, Peter H. 1955 Why Families Move: A Study $3 the Social Psychology pf Urban Residential Mobility, (New York: The Free Press). Seeley, J.R., R.A. Sim, and E.W. Loosley 1956 Crestwood Heights, (Toronto: University of Toronto Press). Shevky, Eshref, and Wendell Bell 1955 Social Area Analysis: Theory, Illustrative Application egg Cogputational Procedures, (Stanford: Stanford University Press). 213 Simmons, James W. 1966 Toronto's Changing Retail Complex, Research Paper Number 104, Chicago: Department of Geography, University of Chicago. Sorokin, D.A., and C.Q. Berger 1939 Time Budgets pg Humpp'Behavior, (Cambridge: Harvard University Press). Stephens, John D. 1973 "Time-Space Paths and the Mechanics of Socio-Environmental Constraints," Ph.D. Dissertation Proposal, Department of Geography, Michigan State University. Szalai, Alexander 1966 "Trends in Comparative Time-Budget Research," The Amerggan Behavioral Scientist, 4: 3-8. Tallman and Morgenner 1970 "Life Style Differences Among Urban and Suburban Blue-Collar Families," Social Forces, 48: 334-348. Timms, Duncan W.G. 1971 The Urban Mosaic: Toward 5 Theory 2; Residential Differentiation, Cambridge: Cambridge University Press). Tonnies, Ferdinand 1963 Gemeinschaft Und Gesellschaft, Translated by Charles P. Loomis as Community and Society, (New York: Harper and Row, Inc.) wallden, Marja 1968 Aktivitetsfalt, Del 1, (Stockholm: Statens Institute for Byggnadsforsknina). Warner, W.L., g£_al. 1957 Social Class ip America, (Gloucester, Mass.: Peter Smith). Webber, Melvin M. 1963 "Order in Diversity: Community Without Propinquity," in L.M. Wingo (ed.) Cities and S ace, (Baltimore: Johns Hopkins Press), pp. 23-56. Webber, Melvin M. 1964 "Culture, Territoriality, and the Elastic Mile," Papers and Proceedings pf the Regional Science Associggion, 13: 59-69. Wheeler, James O. 1972 "Trip Purposes and Urban Activity Linkages," Annals, Associa- tion p§_American Geo ra hers, 62: 641-654. Whitelaw, J.S. 1972 "Scale and Urban Migrant Behavior," Australian Geographical Studies, 10: 101-106. 214 Williams, R.H. and C.G. Wirths 1965 Lives Through the Years: Styles p£_Life and Successful Aging. (New York: Atherton). Willmott, Peter 1969 "Developing Patterns of Urbanization: Some Social Trends," Urban Studies, 6: 286-308. Wilson, Godfrey, and Monica Wilson 1945 The Analysis 2; Social Changg, (Cambridge: Cambridge University Press). Wingo, Lowdon 1961 Transportation and Urban Land, (washington: Resources for the Future, Inc.). Wirth, Louis 1938 "Urbanism as a Way of Life," American Journal pf Sociology, 44: 1-240 Wolpert, Julian 1964 "The Decision Process in a Spatial Context," Annals, Associgtion pf American Geogrgphers, 54: 537-558. 1111' 49 111 5 8 o 3 0 3 9 2 1 3 "I u H " III III, " ml! HI