University of Southern California Dissertations and Theses

Familism, Suburbanization, And Residential Mobility In A Metropolis

(USC Thesis Other)

Familism, Suburbanization, And Residential Mobility In A Metropolis

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content This dlsssrtstton has been
microfilmed exactly as received ® 7-6502
JONES, Elizabeth Jensen, 1930-
FAMILISM, SUBURBANIZATION, AND RESIDENTIAL
MOBILITY IN A METROPOLIS.
University of Southern California, Ph.D„ 1967
Sociology, family
University Microfilms, Inc., Ann Arbor, Michigan
FAMILISM, SUBURBANIZATION,
AND RESIDENTIAL MOBILITY IN A METROPOLIS
by
Elizabeth Jensen Jones
A D isse rta tio n Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In P a r t i a l F ulfillm ent of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Sociology)
January 1967
UNIVERSITY O F SOU TH ERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LO S ANGELES, CALIFORNIA 9 0 0 0 7
This dissertation, written by
EJ_|_za_be.t:h_. Jensen__ Jones..........................
under the direction of hS.r....Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
for the degree of
D O C T O R OF P H IL O S O P H Y
<2-
Dean
Date February^J.967
DISSERTATION COM JHTT
Chairman
1
A C K N O W LED G EM EN TS
Appreciation is due the chairman of my guidance
committee, Maurice D. Van Arsdol, J r . , who f i r s t encouraged
me to undertake th is research and whose p o sitiv e c ritic is m
has contributed g re a tly to the form of the fin a l re p o rt.
Dr. Georges Sabagh, formerly co-chairman of my committee
and now at the U niversity of C a lifo rn ia at Los Angeles, was
very helpful in the e a r lie r stages of form ulating the prob
lem and the methods of study. As d ire c to rs of the re sid e n
t i a l m obility research from which my d ata have been drawn,
Drs. Sabagh and Van Arsdol have been responsible for making
th is study p o ssib le, as has the John Randolph and Dora
Haynes Foundation of Los Angeles, which supported the f ie ld
work fo r the r e s id e n tia l m obility research .
Completion of th is d is s e r ta ti o n has been f a c i l i t a t
ed by the other members of my guidance committee, Edward C.
McDonagh and Langdon E. L ongstreth. Without the varied and
p a tie n t assista n ce of my husband, G il, I would have been
able to accomplish very l i t t l e . To him and to Nancy, C hris,
Mike, Andy, and Don, who have had to liv e with me through
a l l the successive stages of research and w ritin g , go my
special thanks.
TABLE OP CONTENTS
C hapter Page
I. INTRODUCTION A N D STATEMENT OF PROBLEM . . . . 1
Introduction ................................................................ 1
Statement of the Research Problem ................... 11
Plan of the R e s e a r c h .............................................. 13
I I . THEORY CONSTRUCTION: THE HYPOTHESES A N D
THEIR RATIONALE............................................................ 15
An Axiomatic Theory of Urban Familism . . . 15
The Nature of T h e o r y ......................................... 15
Application of the Axiomatic Model to
the Research Problem ..................................... 17
C o n c e p t s ................................................................ 19
H y p o t h e s e s ........................................................... 20
P o stu la te s and Theorems ................................ 21
Empirical Support for the Hypotheses . . . 23
Summary.............................................................................. 39
I I I . TESTING THE THEORY: PROCEDURES............................ 40
O perational D efin itio n s ......................................... 40
Family O rien tatio n .............................................. 41
In te ra c tio n with K i n ......................................... 42
Age of Head of H o u s e h o l d ................................ 43
Size of H o u seh o ld ................................................... 43
Residence of K i n .................................................. 44
Area of R e s id e n c e .................................................. 44
R esidential S ta b il ity ......................................... 46
S t a t i s t i c a l Test of the Hypotheses . . . . 46
D escription of the Survey Design ................... 50
Summary.............................................................................. 52
IV. DATA ANALYSIS..................................................................... 53
Empirical Findings: P o stu la te s ....................... 53
Empirical Findings: Theorems ............................ 68
Summary.............................................................................. 96
i i i
C hapter Page
V. DISCUSSION OF FINDINGS.................................................... 97
Empirical Support for the Theory ........................ 97
A ssociation between V ariab les ........................ 100
Measures of Familism ........................................... 100
S tru c tu r a l C o rre la te s of Familism . . . 104
C r ite rio n V ariables ........................................... 106
Summary: Empirical Test of the
H y p o th e s e s ....................................................................... 106
P o s tu la te s and Theorems ................................. 107
Measurement of Familism ........................................... 109
I n te r a c tio n with K i n .....................................................109
Refinement of O perational Measure . . . 109
In tro d u c tio n of Control V ariables . . . 114
Family O rie n ta tio n .................................................... 115
Revision of O perational Measure . . . . 115
In tro d u c tio n of Control V ariab les . . . 119
Summary: Measurement of Familism . . . . 120
S tr u c tu r a l Measures of Family C h a ra cte r
i s t i c s ......................................................................................121
Revision of The T h e o r y ..................................... 126
Familism and Area of R e s id e n c e ............................. 126
P o s tu la te s and Theorems ................................. 127
Familism and R e sid e n tia l S t a b i l i t y . . . . 128
Age and Size of Household, Area of
Residence, and R e sid e n tia l S t a b i l i t y . . 130
S u m m a r y ......................................................................................133
VI. SU M M A RY AND CONCLUSIONS.....................................................138
The Extended Family System ...................................... 139
Age and G enerational D ifferen ces in
Extended Familism ............................................... 142
S u m m a r y .................................................................................145
Extended Familism and R e sid e n tia l
P ro p in q u ity and M obility ...................................... 146
Nuclear Familism and Suburban Residence . . 150
Extended Familism and Nuclear Familism . . . 152
E valuation of the Theory of Urban Familism . 156
Unresolved Q uestions .................................................... 160
Measurement of F a m i l i s m ........................................... 160
Kin Residence and R e sid e n tia l S t a b i l i t y . 162
C ontrol for Socioeconomic D iffe re n ce s . . 164
S u m m a r y ..................................................................................... 166
LITERATURE CITED ....................................................................... 169
i
iv
LIST OF TABLES
T able Page
1. Age of Head of Household by Area of Residence . 54
2. Age of Head of Household by Area of Residence . 56
3. Household Size by Age of Head of Household . . . 58
4. Household Size by Age of Head of Household . . . 59
5. Household Size by Age of Head of Household
(Married Couples Only) .................................................. 60
6. Family O rien tatio n by Household Size ......................... 62
7. Family O rientation by Household Size ......................... 63
8. Residence of Kin by Family O r i e n t a t i o n .................. 65
9. In tera c tio n with Kin by Residence of Kin . . . ; 66
10. R esidential S ta b ility by In te ra c tio n with Kin . 67
11. Household Size by Area of R e s i d e n c e ....................... 69
12. Household Size by Area of R e s i d e n c e ....................... 70
13. Household Size by Area of R e s i d e n c e ....................... 71
14. Family O rien tatio n by Age of Head of Household . 73
15. Residence of Kin by Household S i z e ........................... 74
16. In terac tio n with Kin by Family O rien tatio n . . . 76
17. R esidential S ta b il ity by Residence of Kin . . . 76
18. R esidential S ta b ility by Residence of Kin . . . 77
19. D irectio n of Mobility by Residence of Kin . . . 77
v
T able Page
20. Family O rientation by Area of Residence . . . . 80
21. Family O rien tatio n by Area of Residence . . . . 81
22. Residence of Kin by Age of Head of Household . 82
23. In te ra c tio n with Kin by Household Size . . . . 83
24. R esidential S ta b il ity by Family O rien tatio n . . 85
25. Residence of Kin by Area of R e s id e n c e ................. 86
26. In te ra c tio n with Kin by Age of Head of
H o u se h o ld .............................................................................. 88
27. R e sid en tial S ta b ility by Household Size . . . . 90
28. R esidential S ta b ility bv Household Size
(Married Couples Only; .............................................. 91
29. In te ra c tio n with Kin by Area of Residence . . . 92
30. R esid e n tia l S t a b i l i t y by Age of Head of
H o u se h o ld .............................................................................. 94
31. R esidential S ta b il ity by Area of Residence . . 95
32. R esid e n tia l S ta b il ity by Area of Residence . . 95
33. In te ra c tio n with Kin by Closeness of
R elationship ..................................................................... I l l
34. R esidential M obility Rates by In te ra c tio n
with Kin (by R e l a t i o n s h i p ) ..................................... 113
35. Extended Family O rien tatio n and Nuclear
Family O rien tatio n ....................................................... 117
36. Per Cent Extended Family O rientation by Area
of Residence and Age of Head of Household . . 123
37. Per Cent Extended Family O rientation by Area
of Residence and Size of H o u s e h o l d ................... 124
38. M obility Rates by Age of Head and Area of
R e s id e n c e .............................................................................. 132
vi
Table Page
39. Age D istrib u tio n of the Mobile, by Area of
R e s id e n c e .............................................................................. 134
40. M obility Rates by Size of Household and Age
of H e a d .................................................................................. 135
41. Comparative Findings: Per Cent of M etropolitan
Respondents Reporting Face-to-Face In te ra c
tion with K i n ................................................. 141
v i i
LIST OF FIGURES
F igure Page
1. Theory of Urban Familism: Matrix of
In te r r e la tio n s ................................................................ 24
2. Theory of Urban Familism: Matrix of
In te rre la tio n s--E m p iric a l Data ........................... 99
3. A ssociation of Categories of Family
O rien tatio n with Selected V ariables . . . . 103
4. Revised Theory of Urban Familism:
Matrix of I n te r re la tio n s (Predicted
and E m p i r i c a l ) ..................................................................... 129
v i i i
CHAPTER I
INTRODUCTION AN D STATEMENT OF PROBLEM
Introduction
The research to be reported in th is paper is con
cerned with developing and te stin g a theory of urban fa m il
ism. Impetus for the in v e stig a tio n has been provided by
d iscussions, in the so cio lo g ical l i t e r a t u r e , of family theo
ry and of stu d ie s of r e s id e n tia l m obility and suburbaniza
tio n . This introductory section w ill review the l i t e r a t u r e ,
concluding with a formal statement of the research problem.
The question, ". . . is there an emergent urban fa -
milism?” has recen tly been raise d by A. 0. H aller. The com
mon assumption of American s o c io lo g is ts - - th a t nuclear family
m obility, required by the occupational system, prevents de
velopment of an extended fam ily system in urban so c iety —may
be inaccurate, he suggests. The urban dweller need not
leave town to find an appropriate job, since p o sitio n s f i t
ting h is sp e c ia lty can be found w ithin the same urban com
plex in which he liv e s . Moreover, he probably rep resen ts
the second or th ird generation of c ity residence for h is
1
fam ily. During th e se g e n e ra tio n s a web of k in sh ip , which
c a r r i e s a complex s e t of mutual o b lig a tio n s , has had a
chance to form, through m arriage, c h ild re n , and c h i ld r e n ’s
m arriages. F u l f i l l i n g these o b lig a tio n s may be very s a t i s
fy in g . "Could i t b e ," asks H a lle r, " th a t the job--once so
c e n tra l to the u r b a n i t e ’s whole being—serv es now as an eco
nomic support fo r an enjoyable extended fam ily lif e ? " *
Sussman and B urchinal, drawing on s tu d ie s in other
modern n a tio n s as w ell as in the United S t a t e s , have pro
posed a re v is io n of fam ily theory r e l a t i n g to the is o l a t i o n
of the n u clear fam ily in modern u r b a n - in d u s tr ia l so c ie ty .
Because a la rg e urban area su p p lie s a l l kinds of s p e c ia liz e d
ed u catio n al and occupational tr a in i n g , the in d iv id u a l can
work a t h is s p e c ia lty while remaining in the midst of his
kin group. ". . . the extended kin fa m ily ," they s t a t e ,
c a r r i e s on m u ltitu d in o u s a c t i v i t i e s th a t have im p li
c a tio n s for the fu n ctio n in g of other s o c ia l systems
of the s o c ie ty . The major a c t i v i t i e s lin k in g the
network are mutual aid and s o c ia l a c t i v i t i e s among
kin r e la te d f a m i l i e s . 2
The kin fam ily network, w ith i t s p a tte r n s of mutual
a id , is a s tr u c t u r e composed of nuclear f a m ilie s bound to
gether by a f f e c tio n a l t i e s and by choice. Nuclear u n i t s
*A. O. H a lle r, "The Urban Fam ily," American Journal
of S o cio lo g y , LXVI (May, 1961), 621-622.
2
Marvin B. Sussman and Lee B urchinal, "Kin Family
Network: Unheralded S tru c tu re in C urrent C o n cep tu alizatio n s
of Family F u n c tio n in g ," Marriage and Family L iv in g , XXIV
(August, 1962), 235.
3
function autonomously, backed by leg al and c u ltu ra l norms;
the ro le of the kin network is supportive rather than coer
cive. In keeping with the demands of an u rb a n -in d u stria l
so c iety , the network does not req u ire geographic propinqui
ty , kin involvement in occupational placement and advance
ment, d ire c t in terv e n tio n as members of nuclear family
u n its achieve so c ia l s ta tu s , or a rig id h ie ra rc h ic a l au
th o rity s tru c tu re . These c h a r a c te r is tic s are la rg e ly ab
sen t. Instead, the network functions in in d ire c t ways to
f a c i l i t a t e achievement and m obility for i t s member fam ilies
and in d iv id u als.
Such re la tio n s h ip s are made fe a s ib le by: (1) mod
ern communication and tra n sp o rta tio n systems which f a c i l i
ta te in te ra c tio n among members; (2) a bureaucratic indus
t r i a l s tru c tu re , which removes the re s p o n s ib ility for job
placement from the fam ily while s t i l l perm itting i t to con
ce n tra te on a c t i v i t i e s intended to aid the so cial and eco
nomic achievement of members; and (3) the expansion of met
ro p o lita n areas, with th e ir range of educational, occupa
tio n a l, and s ta tu s o p p o rtu n itie s. R elationships can be
c a rrie d on even i f kin liv e some d istance from each other
w ithin the m etropolitan area. The tasks of the kin network
complement those of other social systems.
By achieving in te g ra tio n with other so cial systems,
concerned with the general goals of maintenance and
accomplishment of those systems, the extended fami
ly network cannot be considered as an iso la te d or
3
id io s y n c ra tic concept.
This co n cep tu alizatio n draws heavily on the w r i t
ings of Eugene Litwak, who has presented em pirical d ata in
support of h is p ro p o sitio n th a t , in the b u reau cratic middle
c la s s , extended fam ily r e la tio n s h ip s e x i s t in s p ite of geo
graphic m obility and even f a c i l i t a t e th a t m o b ility . The
’’modified extended fam ily” is described as providing eco
nomic, so c ia l and psychological support to the mobile nu
c le ar fam ily. In c o n tra st fo P a rso n s’ hypothesis th a t, in
a mature in d u s tria l so c ie ty , e ith e r extended fam ily s tr u c
tu re s w ill co llapse or nepotism w ill d estro y the in d u s tria l
o r d e r ,4 Litwak b eliev e s th a t the family has adapted for
survival with maximum e ffic ie n c y .
. . . . the extended family re la tio n s h ip which does
not demand geographical propinquity [not examined
by Parsons] i s a s ig n if ic a n t form of so cial behavior
. . . T h e o re tic ally the most e f f i c i e n t organization
combines the a b i l i t y of la rg e -s c a le bureaucracy to
handle uniform s itu a tio n s with the primary group’s
a b i l i t y to deal with id io s y n c ra tic s itu a tio n s .
These two th e o r e tic a l p o in ts suggest th at th ere is
both a need and a capacity for extended fa m ilie s to
3I b i d . , 234.
4T a lc o tt Parsons, ’’Revised A nalytical Approach to
the Theory of Social S t r a t i f i c a t i o n , ” C lass, S ta tu s , and
Power: A Reader in Social S t r a t i f i c a t i o n , ed. Reinhard
Pendix and Seymour M. t i p s e t (Glencoe, TTl.: The Free
P re s s , 1953), p. 116 f f.; and ’’The Social S tru ctu re of the
Fam ily,” The Family: I t s Function and D estin y , ed. Ruth N.
Anshen (New York: Harper B ros., 1^49), pp. 241-274.
e x is t in modern so c ie ty .^
Litwak does, however, lim it h is hypothesis of kin
f a c i l i t a t i o n of nuclear family m obility to selected stages
of the occupational cycle. The extended family encourages
moves only when they are occupationally rewarding. In a
large organization an individual ty p ic a lly moves up to a
peak, where he stay s u n til retire m e n t. Once the peak is
reached, the in d iv id u a l’s working e ffic ie n c y i s no longer
tie d to geographic moves; f u rth e r, i t is at th is point th a t
the nuclear family is in the best economic p o sitio n to sup
port moves of extended fam ily. The older family can thus
a s s is t married children to move, even fu rth e r away; but
there may also be some tendency for kin to move closer to
gether once the occupational peak is reached. Geographical
coalescence, Litwak s ta te s , occurs only when a family is at
peak earning capacity and when i t is le a s t lik e ly to d i s
rupt the in d u s tria l o rganization. At th is point the ca
r e e r i s t may find a comparable job near his extended fam ily,
or may encourage r e tir e d p aren ts to s e t t l e near him.
The extended kin network, then, has been described
by several w rite rs as c h a r a c te r is tic of the modern
^Eugene Litwak, "Geographic M obility and Extended
Family Cohesion," American Sociological Review, XXV (June,
1960), 394.
I b id . , 387. Litwak’s data support the tendency of
young fam ilies to move with kin a ssista n ce , but do not bear
d ir e c tly on the question of l a t e r geographical coalescence.
6
m etro p o lis. A second m etropolitan c h a r a c t e r i s t i c which may
also r e f l e c t an ’’emergent urban fam ilism ” is the suburbani
z a tio n which has been widely described as both r e f le c tin g
and f a c i l i t a t i n g an increased emphasis on family liv in g .
B ell, among o th e rs, has described the emergence of a fam il-
i s t i c l i f e sty le w ith in the nuclear fam ily. B e ll,
8 9
Mowrer, Jaco and Belknap, and other w rite r s have pointed
to nuclear familism and suburbanization as in te r r e l a te d
tren d s in the growth of the m etro p o lis.
In studying reasons fo r moving to the suburbs, Bell
found th a t over 80 per cent of h is respondents gave at
l e a s t p a r tly ’’f a m i l i s t i c ” reasons. These reasons r e f l e c t
p rim arily nuclear ra th e r than extended fam ilism , Bell
n o tes; most moves were made ’’because of the c h i ld r e n ,” and
only a small percentage of respondents moved to be close to
r e l a t i v e s not liv in g with them. In f a c t , se v eral respond
ents mentioned as an advantage of the move the fa c t th a t
the new lo c a tio n was a l i t t l e f a r th e r from th e ir
7Wendell B e ll, ’’S ocial Choice, L ife S ty le s , and
Suburban R esidence,” The Suburban Community, ed. William M.
Dobriner (New York: G. P. 'Putnam's Sons, i9 5 8 ) , pp. 225-
247.
^Ernest R. Mowrer, ’’The Family in S u b u rb ia,” The
Suburban Community, ed. William M. Dobriner (New York: G.
i?. Putnam's Sons, 1958), pp. 147-164; and ’’Sequential and
C lass V ariables of the Family in the Suburban A rea,” Social
F o rc es, XL (December, 1961), 107-112.
^E. G artley Jaco and Ivan Belknap, "Is a New Family
Form Emerging in the Urban Fringe?" American S o cio lo g ical
Review, XVIII (October, 1953), 551-537":-------------------------------
, 4.* 1 0
r e l a t i v e s .
Foote has characterized nuclear familism as involv
ing ". . . the self-co n scio u s recognition of family liv in g
as a d is t in c t iv e and desired a c t i v i t y . ’1 He sees th is as a
uniquely contemporary type of familism, quite apart from
tr a d itio n a l modes of operating a family business, v is itin g
kin, or ’’ra is in g a fam ily.Mowrer has s im ila rly de
scribed the suburban family, e sp e c ia lly in the newer sub
urbs, as centered in upon the raisin g of children and the
12
spending of time in the home.
Evidence concerning the r e la tio n s h ip between the
two types of urban familism is scarce. Some e f f o rts have
been made by family so c io lo g ists to e s ta b lis h extended kin
r e la tio n s and nuclear "family in te g ra tio n " as independent
1 3
v a ria b le s. Bell has suggested (see above) th a t
^■^Bell, op. c i t . , p. 239.
^ N e lso n N. Foote, "Family Living as P lay ," Mar
riage and Family L iving, XVII (November, 1955), 296.
^Mowrer, "The Family in Suburbia," lo c. c i t . , p.
157. Mowrer comments, with regard to the child-centerdness
of the suburban fam ily: "How much of th is child-centerdness
is a mirage is d i f f i c u l t to say for undoubtedly the adult
often finds much of his g re a te st enjoyment in those f e a
tu res of suburban l i f e which he a t tr ib u te s as contributing
to the w elfare of his ch ild re n ."
l^See E verett M. Rogers and Hans Sebald, "A D is
tin c tio n between Familism, Pamily In te g ra tio n , and Kinship
O rie n ta tio n ," Marriage and Family L iving, XXIV (February,
1962), 25-30.
14
ch ild-centered fam ilies may not want kin too close. One
study found an inverse re la tio n s h ip between size of family
of pro creatio n and involvement with kin, and concluded th a t
an abundance of re la tio n sh ip s w ithin the nuclear family may
tend to replace intensive re la tio n s h ip s with extended kin.*^
On the other hand, Litwak found th at h is measures
of extended- and nuclear-fam ily o rie n ta tio n together formed
16
a Guttman scale . Greer and Kube found some differen ces
in frequency of kin v is itin g by the level of urbanization
17
of urban neighborhoods, when fam ilies in four selected
census t r a c ts were compared. Kin v is itin g was the most im
portant single form of so cial in te ra c tio n in a ll four areas
and did not vary by occupational level or e th n ic ity , within
a middle-rank sample. However, frequent v is itin g was le ss
common in the most urbanized t r a c t , re fle c tin g a d ifferen ce
B ell, op. c i t . See also , Paul J. Reiss, "The Ex
tended Kinship System: C o rrelates of and A ttitudes on Fre
quency of In te ra c tio n ," Marriage and Family L iving, XXIV
(November, 1962), 333-339, who found older respondents, but
not younger respondents, wishing more in te ra c tio n with kin.
15:Lee N. Robins and Miroda Tomanec, "Closeness to
Blood R elatives outside the Immediate Family," Marriage and
Family L iving, XXIV (November, 1962), 340-346.
*6Eugene Litwak, "Occupational M obility and Extended
Family Cohesion," American Sociological Review, X X V (Febru
ary, 1960), 17.
1 7 A s measured by the Shevky-Bell typology of urban
sub-areas. Scott Greer and E lla Kube, Urban^Worlds: A
Comparative Study of Four Los Angeles Areas (Los Angeles:
Occidental College, 1955), pp. 5-13.
in lo c a tio n of r e la tiv e s . Fam ilies in the le ss urbanized
tr a c ts were more lik e ly to have r e la tiv e s present in the
18
m etropolitan area.
Mowrer has made some co n trib u tio n to th is question
in h is conceptualization of the T ,suburban cycle” as a v a r i
able affectin g suburban family p a tte rn s . He sees these
p a ttern s as a function of both the s ta tu s p o sitio n of the
neighborhood and i t s stage of development as a suburb. The
new suburb is characterized by home-centeredness and
primary-group neighborliness, which are replaced in the
course of the suburban cycle by in d iv id u a liz a tio n and wider
, 19
community re la tio n s .
Mowrer does not comment d ir e c tly on extended fam il
ism as i t varies by stage of suburban cycle. Indeed, hav
ing described the in te n siv e nuclear familism of the ”pio-
neering” stage of the cycle, in which "C hild-rearing and
18Ib id . , pp. 73-77.
^Mowrer, "Sequential and Class V ariables of the
Family in the Suburban A rea,” l o c . c i t . , 107. Elsewhere,
the same author has elaborated t h i s concept: ”... the
cycle of suburban l i f e is in microcosm the cycle from the
ru ra l to the urban both with resp e ct to the fam ily r e l a
tio n sh ip and community organization. The consequence is a
f l u i d i t y of s t r a t i f i c a t i o n in the new areas, r i g i d i t y in
the older areas; emphasis upon kinship asso cia tio n s in the
new areas, voluntary asso ciatio n s in the old; family unity
is higher in the new areas, lower in the old; the functions
of the family are enlarged in the new areas, contracted in
the old; intimacy of asso ciatio n in the new areas, imper
sonal and segmental asso cia tio n in the o ld .” Mowrer, "The
Family in Suburbia," l o c . c i t . , p. 163.
10
home maintenance become the focus of fam ily l i f e for hus-
20
band and wife . . . , ” he goes on to measure fam ilism as
’’v i s i t i n g the homes of the extended fam ily w ith in the met-
21
ro p o lita n a r e a . ” His data on th is item are re la te d not
to stage of cycle but to the husband’ s occupational s ta tu s .
Comparing craftsm en, s a le s , m anagerial, and p ro fe ssio n a l
workers, he fin d s craftsm en most f a m i l i s t i c in th is sense,
as w ell as most lik e ly to have had f a th e r s with the same
occupation and most li k e l y to regard t h e i r p rese n t homes as
permanent. These d a ta , Mowrer b e lie v e s , suggest th a t the
move to the suburbs re p re se n ts a move up the so c ia l scale
which tends to c re a te a lie n a tio n from non-mobile extended
fam ily members. The degree of extended fam ilism is de
scribed as a fu n ctio n both of s o c ia l rank (here measured by
occupational c la s s ) and of the stage of the suburban cy-
, 22
c l e .
The tendency of lo w e r-sta tu s neighborhoods, then,
to preserve t r a d i t i o n a l ro le s and fam ilism i s confounded by
the suburban cy cle--from more r u ra l to more urban. The
newer suburb is more l i k e l y to be f a m i l i s t i c , in both nu-
23
c le ar and extended terms, and to become in c re a sin g ly
20
Mowrer, ’’Sequential and Class V ariables of the
Family in the Suburban A re a,” lo c. c i t . , 109.
21Ib id . , 110. 22I b i d . , 111.
23
This g e n e ra liz a tio n is im plied but not d i r e c t l y
s ta te d by Mowrer.
11
’’urbanized” in the behavior of i t s r e s id e n ts as i t gets
o ld e r. The lo w e r-s ta tu s suburban fam ily i s more li k e l y to
be f a m i l i s t i c , at l e a s t in r e l a t i o n to the extended fam ily,
than is the u p p e r-s ta tu s fam ily.
Statement of the Research Problem
There is considerable evidence which suggests the
emergence of an urban fam ilism . The f a m i l i s t i c l i f e s ty l e ,
moreover, may well include both extended and n u clear-fa m ily
components, although on th is p o in t the evidence i s incon
c lu siv e and in some re s p e c ts c o n tra d ic to ry . For the p u r
poses of the p rese n t study, the d ec isio n has been made to
regard urban fam ilism , extended and urban fam ilism , nuclear
as p o in ts on a sin g le dimension of fam ilism . This decisio n
re p re se n ts an im p lic it hypothesis concerning the r e l a t i o n
ship between extended and n u clear fam ilism , which w ill be
te s te d in the course of the study.
The p resen t research re p re se n ts an e f f o r t to bu ild
on these varied sources in developing a theory of urban f a
milism, which w ill then be subjected to em p irical t e s t .
Hypotheses are drawn from two p rin c ip a l sources in the so
c io lo g ic a l l i t e r a t u r e : (1) fam ily theory, and (2) s tu d ie s
of m obility and suburbanization. The c e n tra l theme of the
theory is th a t , in the modern m etro p o lis, fam ilism (both
nuclear and extended) is (a) p o s itiv e ly asso c ia te d w ith
suburbanization and (b) n e g a tiv e ly asso ciated with
12
r e s id e n tia l m o b ility .
The theory to be constructed u t i l i z e s second-level
24
g e n e ra liz a tio n s based on a fu n c tio n a l approach. I t is
based on the conception th a t, w ith in the s e ttin g of the
m etropolitan community, fam ilism i s fu n c tio n a lly r e la te d to
both suburbanization and r e s id e n tia l s t a b i l i t y . F a m ilis tic
o r ie n ta tio n s and behavior are seen as f a c i l i t a t i n g , and f a
c i l i t a t e d by, suburban resid en ce and lower r e s i d e n t i a l mo
b i l i t y ; each may be p o t e n tia lly regarded as a fun ctio n of
the o t h e r .2'" ’
The hypotheses to be te s te d w ill be system atized in
24
"G en eralizatio n s on the interm ed iate le v e l of ab-
s tr u c tio n are what Merton c a l l s ’th e o rie s of the middle
r a n g e .’ These second-level g e n e ra liz a tio n s seem to take
two forms: those based on a fu n ctio n a l approach and those
th a t subsume se v eral f i r s t - l e v e l p ro p o sitio n s about several
s p e c if ic areas under a few th a t apply to a l l the areas. A
fu n c tio n a l study in v e s tig a te s the in te g ra tio n of p a rts into
a whole or the interdependence of the p a r ts of a group, a
community, or a s o c ie ty ." Harvey C. Locke, "A Model fo r
Small-Group Theory: G en e ra lizatio n s Derived from Empirical
R esearch," U n iv ersity of Southern C a lifo rn ia , mimeo., pp.
5-6. (Paper read before the F ifty -T h ird Annual Meeting of
the American S o cio lo g ica l S o ciety , at S e a t tle , Washington,
1958.)
^ R e la t e d fu n c tio n a l th e o rie s have been proposed by
Sussman and B urchinal, and Litwak; see d isc u ssio n on pp.
3-5 above. Note, however, th a t the r e la tio n s h ip between
fam ilism and r e s i d e n t i a l m o b ility hypothesized by Litwak is
contrary to th a t p re d ic te d in the p re se n t rese arch .
13
Of\
the form of an axiomatic theory. The postulates of the
theory, interrelating measures of familism, anticipated
structural correlates of familism, and resid en tial location
and mobility, are formulated with reference to the empiri
cal and theoretical litera tu re . Theorems are lo g ic a lly de
rived from the postulates and are also related to available
empirical findings.
Following operational d e f in itio n of terms, the hy
potheses w ill be te ste d with d ata from a sample survey of
the 1960 Los Angeles Standard M etropolitan S t a t i s t i c a l
Area. Analysis of these data w ill provide a b asis for
evaluating and, as necessary, revising the theory. Conclu
sions about the u t i l i t y for fam ily sociology of th is ap
proach to theory construction w ill be drawn on the basis of
the observed re la tio n s h ip s between the v a ria b le s.
Plan of the Research
This chapter has introduced the problem under in
v e stig a tio n . Chapter I I s ta te s the problem in the formal
terms of an axiomatic theory; p o stu late s are here presented
and theorems are derived. In addition, the em pirical and
th e o re tic a l l i t e r a t u r e relev an t to each p o stu la te and
^^See Hans Z etterberg, ”On Axiomatic Theories in
Sociology,” The Language of Social Research: A Reader in
the Methodology of so c ia l Research, ed. fraul F. Lazarsfeld
and Morris Rosenberg (Glencoe, 111.: The Free P re ss,
1955), pp. 533-540. Reprinted from On Theory and V e rific a
tio n in Sociology (The T ressler PressTJ 1^54), pp. 16-26.
14
theorem is summarized. In Chapter I I I the concepts are
given operational d e fin itio n s for the purpose of te stin g
the hypotheses. Chapter IV describes the sample survey of
the Los Angeles Standard M etropolitan S t a t i s t i c a l Area from
which data were secured, and p resen ts an analysis of the
data for each hypothesis in the theory. The data analysis
is elaborated in Chapter V, with a review of the empirical
asso ciatio n s between each pair of v ariab les measured, a
discussion of problems of measurement, and a revision of
the theory on the b a s is of the em pirical findings. Chapter
VI brings together the research data in terms of the ques
tions with which the development of the theory began: Is
there an urban extended family system associated with r e s i
d en tia l propinquity and s ta b ili ty ? Is nuclear familism as
sociated with suburban residence? Do extended and nuclear
familism occur together? Suggestions for fu rth e r in v e s ti
gation and a fin a l summary of principal findings of the r e
search conclude the paper.
CHAPTER I I
THEORY CONSTRUCTION: THE HYPOTHESES
A N D THEIR RATIONALE
The general su b ject matter of th is research, name
ly , the nature of urban familism and i t s re la tio n s h ip to
r e s id e n tia l m obility and suburbanization, has been d i s
cussed in the preceding pages. In th is chapter, formal
p resen tatio n of the research problem in the terms of an ax
iomatic theory w ill be undertaken. Basic concepts, postu
la t e s , and theorems are sta te d and the lo g ic al deriv atio n
of the theorems is noted. Empirical and th e o re tic a l sup
port for each p o s tu la te , and for each theorem as av a ila b le,
is then presented to conclude the statement of the theory.
An Axiomatic Theory of Urban Familism
The Nature of Theory
In discussions of urban familism and the v aria b les
which may be expected to occur in asso cia tio n with i t , su f
f ic ie n t observations and g e n e ra liz atio n s have been p re v i
ously made to ju s t if y an attempt to p red ic t the regular oc
currence in modern m etropolitan households of an
15
16
in te r r e l a te d complex of v a ria b le s which include f a m il is ti c
a ttitu d e s and behavior, s tr u c t u r a l c o r re la te s of familism,
suburban resid e n ce , and r e s id e n tia l s t a b i l i t y . C ontradic
tio n s in the em pirical evidence and in the th e o re tic a l con
clusions of others suggest th a t the form ulation of a theory
which form ally i n t e r r e l a t e s these v a ria b le s w ill almost in
e v ita b ly contain im perfections when subjected to em pirical
t e s t . Even an im perfect theory, however, may lead in the
d ire c tio n of b e t te r p re d ic tio n and explanation.
Schrag has id e n tif ie d ways in which theory may im
prove research as follow s: (1) By helping to id e n tify con
d itio n s th a t in te r f e r e with p re d ic tio n and explanation; (2)
By suggesting fu rth e r a p p lic a tio n s of p re d ic tio n and explan
ation formulas; (3) By helping to in te g ra te research f in d
ings in the i n t e r e s t of cumulative knowledge; (4) By as
s is tin g in the corroboration of em pirical knowledge by
c la rify in g the p ro v isio n al nature of t e s t and v e r if ic a tio n ;
and (5) By c la rify in g the ro le of law like assumptions in
the processes of p re d ic tio n and explanation.
Theories vary widely in the degree of lo g ic a l f o r
m ality which they a t t a i n . Z etterberg has argued th a t in
the development of ”middle-range" th e o rie s there are
^■Clarence Schrag, "Some Notes on Crim inological
Theory," Conference on Research Planning on Crime and De
linquency" ed. William R. Larson (Los Angeles: Youth Stud
ie s Center , U niversity of Southern C a lifo rn ia , 1962), pp.
2-4.
17
im portant advantages to be gained from the use of an axio-
2
m atic model. The axiom atic approach perm its the re s e a rc h
er to work sim ultaneously from an em p irical base by means
of in d u c tiv e a n a ly s is and an a p r i o r i base by means of de-
3
d u ctiv e a n a ly s is , with c l e a r ly defined r u l e s fo r lo g ic a l
deduction. Advantages of the approach have been id e n tif ie d
as fo llo w s; (1) The concepts and p o s tu la te s of an axiomat
ic theory o ffe r the most parsim onious summary of research
fin d in g s ; (2) The axiom atic theory has the h ig h e st p la u s i
b i l i t y per amount of supporting em pirical d a ta; (3) The ax
iom atic theory lo c a te s s t r a t e g i c rese arch problems; and (4)
The axiom atic theory provides a lim ite d area in which the
source of the f a i l u r e of a h ypothesis to meet the em pirical
4
t e s t can be lo c a te d .
A p p licatio n of the Axiomatic Model
to the Research Problem
The p resen t re se a rc h has begun with the general hy
p o th eses th a t, in the modern m e tro p o lis, fam ilism is (a)
2
Hans Z e tte rb e rg , ,T On Axiomatic Theories in S o c io l
o gy,” The Language of S ocial Research: A Reader in the
Methodology of S ocial R esearch, e d . Paul F. t a z a r s f e id and
M orris Rosenberg (G lencoe, i l l . : The Free P re ss, 1955),
pp. 533-540. R eprinted from On Theory and V e r if ic a tio n in
Sociology (The T re s s le r P re ss” 1954;, pp. 16-2&.
3
Schrag, op. c i t . , p. 9.
4
Z e tte rb e rg , op. c i t . , pp. 534-538.
18
p o s itiv e ly associated with suburbanization and (b) nega
tiv e ly associated with r e s id e n tia l m obility. Development
of a theory which e la b o ra te s on and perm its the te s tin g of
these hypotheses re q u ire s several assumptions, or im p lic it
hypotheses, concerning the v a ria b le s to be measured. The
assumptions made in th is research (which, as im p lic it hy
potheses, w ill be te sted in the course of te s tin g the th e
ory) are as follow s:
1. Familism has both a t titu d i n a l and behavioral com
ponents, which are p o s itiv e ly associated with each other.
2. Within the a t titu d i n a l component, extended familism
and nuclear familism are points on a sin g le dimension.
3. S tru c tu ra l c o rre la te s of familism can be id e n ti
fied.
The components of familism to be incorporated in
the theory as basic concepts are both a t titu d i n a l and be
h av io ral. The a t t i t u d i n a l component includes both extended
and nuclear familism. A nticipated s tru c tu ra l c o rre la te s of
familism are id e n tif ie d , and th ese, together with the f a
milism measures, are hypothesized to be associated with the
c r ite rio n v a ria b le s, r e s id e n tia l lo catio n and m obility.
The step s in formulating an axiomatic theory are
(1) li s t i n g of basic concepts and d e f in itio n of derived
concepts, (2) form ulation of hypotheses, and (3) se le c tio n ,
from among the hypotheses, of the p o stu la te s of the theory,
from which a ll other hypotheses, the theorems, are capable
19
of d e r iv a tio n .^
1. Concepts
The vocabulary of a th e o ry in clu d es two kinds of term s:
undefined and d e fin e d . The l a t t e r can always be e lim in a t
ed in favor of the former, which are used in th e i r accepted
or conventional meaning. Concepts in the p re se n t study,
l i s t e d below, are l e f t undefined in a s t r i c t sen se. C ir-
cum locations are in tro d u ced , however, to convey t h e i r i n
tended meaning.^
Family o r i e n t a t i o n : A range of d esig n ate d a t t i t u d e s
toward extended and nuclear fam ily behavior.
K in: Persons r e la te d by blood, m arriage or adoption to
the head of the household or his spouse.
I n t e r a c t i o n : S o cial c o n ta c t.
Household: The e n tir e group of persons occupying a
9
dw elling u n i t .
Head of household: The r e p r e s e n ta tiv e or a u th o rity
fig u re of the household, as defined by i t s members.
5 I b i d . , pp. 533-534.
I b i d ■ In the same co n tex t, Z e tte rb e rg uses the
terms b a s ic concepts and derived c o n c e p ts. The former are
defined through com binations of the l a t t e r .
7
Schrag, op. c i t . , p. 5.
g
Z e tte rb e rg , op. c i t . , p. 533.
9
Paul C. G lick, American F am ilies (New York: John
Wiley and Sons, 1957), Appendix A.
20
Area of resid e n ce: Location of the household w ithin
the m etropolitan area and i t s sub-areas.
R esid en tial s t a b i l i t y : Degree of permanence of lo ca
tio n of the household group in a dwelling u n it.
In form ulating hypotheses these concepts are r e l a t
ed by means of the following terms: (a) newer, (b) young
e r, (c) la rg e r, (d) c lo se r, and (e) g re a te r. All these are
comparative terms used to denote d iffe re n c e s on ordinal or
card in al sc ale s of measurement.
2. Hypotheses
The hypotheses of the theory p re d ic t asso cia tio n s
among v a ria b le s formulated in terms of the above concepts
and the comparatives used to r e la te them. The selected
v ariab les are: (a) measures of familism: family o r ie n ta
tio n and in te ra c tio n with k in ; (b) an tic ip a te d s tr u c tu ra l
c o rre la te s of familism: age of head of household, size of
household, and residence of k i n ; (c) c r ite r io n v a ria b le s:
area of residence and r e s id e n tia l s t a b i l i t y . In te rre la tin g
these v ariab les two at a time, twenty-one hypotheses can be
s ta te d . From these, the l e a s t necessary number of postu
la te s have been se le c te d , following the ru le s of c o n siste n
cy and independence.*^ The theorems are lo g ic a lly derived
from the p o stu la te s.
*^Z etterberg, op. c i t . , pp. 533-534.
P o stu la te s
1. The newer the area of residence, the younger the
head of the household.
2. The younger the head of the household, the larg er
the household.
3. The la rg e r the household, the greater the family
o rie n ta tio n .
4. The g reater the family o rie n ta tio n , the closer the
residence of kin.
5. The closer the residence of kin, the greater the
in te ra c tio n with kin.
6. The greater the in te ra c tio n with k in , the g reater
the r e s id e n tia l s t a b i l i t y .
Theorems
7. The newer the area of residence, the la rg e r the
household. (Derived from P o stu la te s 1 and 2.)
8. The younger the head of the household, the greater
the fam ily o rie n ta tio n . (Derived from P o stu la te s 2
and 3 .)
9. The larg e r the household, the closer the residence
of kin. (Derived from P o stu la te s 3 and 4 .)
10. The g reater the fam ily o rie n ta tio n , the g reater the
in te ra c tio n with kin. (Derived from P o stu la te s 4 and
5.)
11. The closer the residence of kin, the greater the
r e s id e n tia l s t a b i l i t y . (Derived from P o stu la te s 5 and
12. The newer the area of resid en ce, the greater the
fam ily o r ie n ta tio n . .(Derived from P o s tu la te 1 and The
orem 8.)
13. The younger the head of the household, the closer
the residence of k in . (Derived from P o s tu la te 2 and
Theorem 9.)
14. The larg e r the household, the g reater the in te r a c
tio n with k in . (Derived from P o stu la te 3 and Theorem
10. )
15. The g reater the family o r ie n ta tio n , the g reater
the r e s id e n tia l s t a b i l i t y . (Derived from P o stu la te 4
and Theorem 11.)
16. The newer the area of resid en ce, the closer the
residence of kin. (Derived from P o stu la te 1 and The
orem 13.)
17. The younger the head of the household, the greater
the in te ra c tio n with kin. (Derived from P o s tu la te 2
and Theorem 14.)
18. The la rg e r the household, the g re a te r the re sid e n
t i a l s t a b i l i t y . (Derived from P o stu la te 3 and Theorem
15.)
19. The newer the area of residence, the g reater the
in te ra c tio n with kin. (Derived from P o stu la te 1 and
Theorem 17.)
23
20. The younger the head of the household, the g re a te r
the r e s id e n tia l s t a b i l i t y . (Derived from P o stu late 2
and Theorem 18.)
21. The newer the area of resid en ce, the greater the
r e s id e n tia l s t a b i l i t y . (Derived from P o stu la te 1 and
Theorem 20.)
Figure 1 p resen ts the theory in tab u lar form. Pos
tu la te s with th e ir predicted re la tio n s h ip s between v a r i
ables appear in the diagonal, theorems in the remaining
c e lls .
Empirical Support for the Hypotheses
The hypotheses of a theory should be formulated so
th a t they do not co n tra d ict what is known to be true.**' On
the following pages research fin d in g s and/or informal ob
serv atio n s supporting the hypotheses of th is theory are
summarized. Where co n flic tin g evidence e x is ts in the l i t
e ra tu re i t is also presented.*^
11I b i d . , p. 533.
*^It w ill be noted that several theorems f a i l to
meet the c r ite r io n of em pirical support. These hypotheses
have been included in the theory in the in t e r e s t of lo g ical
consistency; i t is an ticip ated th a t, i f the findings of the
present study concur with those of previous research in
these in stan ce s, the introduction of control v a ria b le s in
the d ata analysis may o ffer p o s s i b i l i t i e s of re in te rp re tin g
findings co n siste n t with the theory.
FIGURE 1
THEORY OF URBAN FAMILISM: MATRIX OF INTERRELATIONS*
Area of
Residence
New
Old
Age of Head
of Household
Young
Old
Household
Size
Large
Small
Family
O rientation
High
Low
Residence
of Kin
Close
Far
Interaction
with Kin
High
Low
Age of
Head
Household
Size
Family
O rientation
Residence
of Kin
Interaction
with Kin
Residential
S ta b ility
Yg. Old Lg. Sm. High Low Close Far High Low High Low
N .
X
X X X
X
X X
X
X
X
X X
X X X
IJx_ X
X
*Arrows indicate re latio n sh ip s implied by the theory. Double arrows specify postulates;
single arrows specify theorems. (After Clarence Schrag, "Some Notes on Criminological The
ory," Conference on Research Planning on Crime and Delinquency, ed. William R. Larson (Los
Angeles: Youth Studies Center, University of Southern C alifo rn ia, 1962), p. 14.)
25
P o s tu la te 1 . The newer the area of resid e n c e , the
younger the head of the household.
The tendency for new suburbs to be in h a b ite d by
young couples with young c h ild re n has been widely observed.
D etailed d a ta s p e c i f i c a l l y re la te d to the hypothesis are
scarce, however.
Using 1950 census d a ta , Duncan and Reiss noted the
phenomenal growth of both old and newer suburbs. In con
t r a s t to c e n tra l c i t i e s , suburbs had a lower median age and
13
a preponderance of in d iv id u a ls under 45. Mowrer studied
a sample of Chicago suburbs, comparing old and new suburbs.
The l a t t e r were more homogeneous than the form er. F am ilies
tended to move to suburbs a f te r they had one c h ild ; couples
whose ch ild re n were grown tended to move out, or to remain
in older se c tio n s. Mowrerfs d isc u ssio n of the "suburban
cycle" im plies, w hile not d i r e c t l y s t a t i n g , a d i r e c t a s
so c ia tio n between the fam ily l i f e cycle and the suburban
14
cycle: new suburbs a t t r a c t young f a m ilie s . Meeker and
H arris have described r e la te d c h a r a c t e r i s t i c s fo r the o u ter
suburbs of Los Angeles: la rg e numbers of c h ild re n 0-4, and
13
O tis Dudley Duncan and A lbert J . R eiss, J r . , So
c i a l C h a ra c te ris t i c s of Urban and Rural Communities, 1951T
(.New York: John Wiley and Sons, 195b;, p. 120.
14E rnest R. Mowrer, "The Family in S ub u rb ia," The
Suburban Community, e d . William M. Dobriner (New York: (3.
P. Putnam’s Sons, 1958), pp. 147-164; and "S equential and
C lass V ariables of the Family in the Suburban Area," S ocial
F o rces, XL (December, 1961), 107-112.
26
few aged. These fin d in g s are based on the 1960 ce n su s.15
P o s tu la te 2 . The younger the head of the household,
the la rg e r the household.
The a s s o c ia tio n between age of head and household
siz e is a lo g ic a l one based on the fam ily cy cle, provided
age is ca teg o riz ed broadly enough to compensate for the
b a s ic a lly c u r v ilin e a r a s so c ia tio n between the v a r ia b le s .
The hypothesis has been s ta te d in lin e a r terms to conform
to the requirem ents of the th e o r e tic a l model, but i t is ex
pected th a t some m odification w ill be re q u ire d . Trends in
the d ir e c tio n of e a rly m arriages, of the completion of
ch ild -b e a rin g e a rly in a m arriage, and of increased f e r t i l
ity in re c e n t decades are expected to augment fam ily cycle
behavior in supporting the h y p o th e s is .15
P o s tu la te 3. The la rg e r the household, the g reater
the family o r ie n ta tio n .
17
The term fam ily o r ie n ta tio n is used in th is study
to denote an a t t i t u d i n a l measure of fam ilism , based on ex
pressed a t t i t u d e s toward both the nuclear fam ily and the
15Marchia Meeker with Joan R. H a rris, Background
for P lan n in g , Research Report No. 17, Research Department,
Welfare Planning Council, Los Angeles Region, 1964, p. 26.
16See Paul C. G lick, "The L ife Cycle of the Fami
l y ," Marriage and Family L iving, XVII (February, 1955),
3-9.
17
This term has been p rev io u sly used by Eugene L it-
wak, "Occupational M obility and Extended Family Cohesion,"
American S o cio lo g ical Review, XXV (February, I960), 9-21;
and "Geographic M obility and Extended Family Cohesion,"
American S o cio lo g ical Review, XXV (June, 1960), 385-394.
27
18
extended kin group. Most measures of th is type have not
reported c o rre la tio n s of a t titu d e s with household siz e .
Zimmerman has expressed the opinion th a t childbearing i s
the fundamental independent v a ria b le which determ ines the
development of other family fu n c tio n s. S o c ie tie s with nu
merous children are n e c e s s a rily f a m il is ti c ; ", . . the b a s is
19
of fam ilism is the b ir th r a t e . ,f Robins and Tomanec found
20
th at closeness to kin decreased as the siz e of the nuclear
family in c re a se d --a finding which throws some doubt on the
p o stu la te as o p e ra tio n a lly defined to include both extended
and n u clear fam ily o rie n ta tio n in a continuum (See page 41).
However, Litwak found no d ir e c t r e la tio n s h ip between i n t e r
action with kin and expressed a t titu d e s toward kin (extended
21
fam ily o r ie n ta tio n ) .
See Litwak, "Geographic M obility . . . , " lo c .
c i t . ; Panos D. B ardis, "A ttitu d es toward the Family among
College Students and Their P a r e n ts ,” Sociology and Social
Research, XLIII (May-June, 1959), 352-358; E v erett M.
Rogers and Hans Sebald, "A D is tin c tio n between Familism,
Family In te g ra tio n , and Kinship O rie n ta tio n ," Marriage and
Family L iving, XXIV (February, 1962), 25-30.
^ C a r l e C. Zimmerman, Family and C iv iliz a tio n (New
York: Harper B ros., 1947), p. 700.
20
As measured by the number of avenues used for
communication, the performance of se rv ic e s, and the f u l
fillm e n t of o b lig a tio n s between r e l a t i v e s . Lee N. Robins
and Miroda Tomanec, "Closeness to Blood R elativ es outside
the Immediate Fam ily," Marriage and Family L iv in g , XXIV
(November, 1962), 340-346.
2*Litwak, "Geographic M obility . . . , " lo c . c i t . ,
388.
28
P o stu la te 4 . The greater the family o rie n ta tio n , the
clo ser the residence of kin.
This p o stu la te is based on the assumption th a t
households with a strong family o rie n ta tio n w ill tend to
22
choose to liv e near r e la tiv e s . H aller and Sussman and
23
Burchinal have implied th a t the o p p o rtu n ities for employ
ment offered in the m etropolis make continued propinquity
with kin possible without undermining the occupational sys
tem. Litwak, however, has presented em pirical data sup
porting his statem ent th a t ". . . geographical distan ce be
tween r e la tiv e s does not n e c e ssa rily lead to a lo ss of ex-
24
tended fam ily i d e n t i f i c a t i o n . ” Extended-family oriented
respondents in h is sample were equally lik e ly to have r e l a
tiv e s present or absent in the m etropolitan area, and were
as lik e ly to plan a move away from th e ir kin as were the
non-family o rien ted . This finding tended to be lim ite d ,
however, to bu reau cratic m iddle-class respondents on a ca-
25
reer upswing, and cannot n e c e ss a rily be generalized to a
broadly re p re se n ta tiv e sample of a m etropolitan population.
22A. 0. H aller, ”The Urban Family," American Jo u r
nal of Sociology, LXVI (May, 1961), 621-622.
Marvin B. Sussman and Lee Burchinal, "Kin Family
Network: Unheralded S tru ctu re in Current Conceptualiza
tio n s of Family Functioning," Marriage and Family L iving,
XXIV (August, 1962), 231-240.
2^Litwak, "Geographic M obility . . . , " lo c. c i t .
25I b i d . , 389-392.
29
P o stu la te 5 . The clo ser the residence of kin, the
g reater the in te ra c tio n with kin.
Reiss found residence to be the most important f a c
tor in accounting for the frequency of fa c e -to -fa c e i n t e r
actio n among kin. Together with closeness of r e la tio n s h ip ,
i t accounted for n early a l l d iffe re n c e s in in te ra c tio n . He
regards i t as appropriate to refe r to the American kinship
system as "e c o lo g ic a lly s t r u c t u r e d .” Both Reiss and
28
Litwak reported th at frequent v i s i t s (weekly or more) oc
curred prim arily among kin re sid in g in the same community.
P o stu la te 6 . The g reater the in te ra c tio n with kin,
the g re a te r the r e s id e n tia l s t a b i l i t y .
No em pirical support or r e je c tio n for th is hypothe-
29
s is has been found in the l i t e r a t u r e . However, H aller
30
and Sussman and Burchinal have both suggested the e x i s t
ence of such a re la tio n s h ip . Sussman has w ritte n th a t p a r
ental support to newly married couples may help to reduce
m obility. " I t can be hypothesized th a t population m obility
is not higher than i t is because of parental support and
26Paul J . R eiss, "The Extended Kinship System:
C o rrela tes of and A ttitu d e s on Frequency of I n te r a c tio n ,"
Marriage and Family Living, XXIV (November, 1962), 335-336.
27I b i d . , 335.
28Litwak, "Geographic M obility . . . , " lo c . c i t . ,
388.
2< ^Haller , op. c i t .
Sussman and Burchinal, op. c i t .
31
other kin dependencies."
Theorem 7 . The newer the area of residence, the
larg er the household.
In comparison with c i t i e s , suburbs have la rg e r
32
households. Further, f e r t i l i t y r a te s are higher in the
33
suburbs and urban frin g e than in cen tral c i t i e s . Freed
man, reporting that wives in suburbs of the 12 la rg e s t c i t
ies in the United S tate s expect s lig h tly more b irth s than
city wives, suggested th at those wanting children tend to
move to suburbs and th a t suburban residence leads to want
ing more c h ild re n .3^ In Los Angeles County in 1960, areas
with la rg e households were also areas of large child popu
la tio n and were, with few exceptions, located in the outer
35
or newer suburbs.
Theorem 8. The younger the head of the household,
the greater the fam ily o rie n ta tio n .
No em pirical support or re je c tio n for th is theorem
Marvin B. Sussman, "The Isolated Nuclear Family:
Fact or F ic tio n ," Selected Studies in Marriage and the Fam
i l y , ed. Robert F. Winch, Robert McGinnis, and Herbert R.
Barringer (New York: H olt, Rinehart and Winston, 1962), p.
55. Adapted from Social Problems, VI (1959), 333-340.
32Robert F. Winch, The Modern Family (New York:
Holt, Rinehart and Winston, 1963), p. 1$9.
33Ib id . , p. 202.
34Ronald Freedman, Pascal K. Whelpton, and Arthur
A. Campbell, Family Planning. S t e r i l i t y , and Population
Growth (New York: McGraw-Hill Book Co., 1959), pp. 51i —
I T T .
35
Meeker and H arris, op. c i t . , p. 26.
31
has been found in the l i t e r a t u r e .
Theorem 9 . The la rg e r the household, the closer
the_residence of kin.
No em pirical support or r e je c tio n for th is theorem
has been found in the l i t e r a t u r e .
Theorem 10. The greater the family o rie n ta tio n , the
g re ater the in te ra c tio n with kin.
No em pirical support or r e je c tio n for th is theorem
has been found in the l i t e r a t u r e .
Theorem 11. The closer the residence of kin, the
g re ater the r e s id e n tia l s t a b i l i t y .
The lim ite d data av ailab le f a i l to support th is
theorem. Reiss reported th a t m obility is apparently random
3 6
with regard to residence of kin. Litwak measured lo c a
tio n of kin residence only in conjunction with extended
family o rie n ta tio n . Excluding respondents without r e l a
tiv e s present in the m etropolitan area, he c la s s ifie d 55
per cent of the remainder as p o te n tia l m igrants; of these
the la rg e s t number were extended-family orien ted . Those on
the upswing of a bu reau cratic career were at l e a s t as l i k e
ly to move away from r e la tiv e s if they were extended-family
oriented as if they were n o t. However, the extended-family
oriented on a career plateau were much le s s lik e ly to move
37
away from r e la tiv e s .
^ R e i s s , op. c i t . , 336.
37
Litwak, "Geographic M obility . . . , " l o c . c i t . ,
389-392.
32
Theorem 12 . The newer the area of re sid e n c e , the
g re a te r the fam ily o r ie n ta tio n .
Comparing Chicago suburbs on the b a s is of d a te of
se ttle m e n t, Mowrer g en eralized from h is fin d in g s th a t
” . . . fam ily u n ity is higher in the new a re a s , lower in
the old; the fu n c tio n s of the fam ily are enlarged in the
38
new areas, co n tracted in the old . . ." B ell found sub
urban r e s id e n ts more f a m i l i s t i c than c i ty r e s id e n ts , as
measured by responses to an a t t i t u d e scale designed to
e l i c i t commitment to one or more of sev eral l i f e s t y l e s :
39
fam ilism , c a r e e r , and consumership.
Theorem 1 3 . The younger the head of the household,
the c lo ser the resid e n ce of k in .
This r e la tio n s h ip between age and lo c a tio n of kin
is implied by R eiss, who observed th a t in te r a c tio n w ith kin
decreased with successive phases of the fam ily cycle, ap
p a re n tly as the cumulative e f f e c t of r e s i d e n t i a l m obility^
Theorem 14. The la rg e r the household, the g re a te r
the in te r a c tio n with k in .
Reiss*s fin d in g th a t kin tend to liv e fu rth e r apart
41
with the passage of time can be gen eralized to th is
O Q
Mowrer, "The Family in S uburbia," lo c . c i t . , p.
163.
39Wendell B e ll, "Social Choice, L ife S ty le s , and
Suburban R esidence," The Suburban Community, e d . William M.
Dobriner (New York: G. P. Putnam's Sons, 1958), pp. 225-
247.
^ R e i s s , op. c i t . , 336. ^ I b i d .
theorem. To the ex ten t th at younger fam ilies are larger
(cf. P o stu late 2), the theorem is expected to hold tru e. On
the other hand, Robins and Tomanec found more "closeness of
i n t e r a c t i o n " ^ with extended kin when the nuclear family was
sm aller, and suggested th a t apparently a constant amount of
time is devoted to in te ra c tio n with r e l a t i v e s . In tera c tio n
with secondary r e la tiv e s decreased when the respondent was
married. The smaller nuclear family in the middle c la ss,
the authors suggested, may explain Litwak*s finding that
there i s more contact between housewives and th e ir r e la tiv e s
43
as th e ir social sta tu s increases. The sample used by
Robins and Tomanec was composed prim arily of college s tu
dents. This f a c t lim its the age and so cial s ta tu s range
measured and thus may lim it the extent to which the findings
can be generalized to d iffe re n t populations.
Theorem 15. The g reater the family o rie n ta tio n , the
greater the re s id e n tia l s t a b i l i t y .
Findings reported by Litwak are contrary to those
predicted by th is theorem. Litwak found th a t for individu
als on the upswing of th e ir ca re ers, the extended-family
oriented were more lik e ly than respondents without th is or-
44
ie n ta tio n to be p o te n tia l m igrants. S im ilarly, the
42
See footnote 20, Chapter I I .
43
Robins and Tomanec, op. c i t . , 345.
44
Litwak, "Geographic M obility . . . , " loc. c i t .
34
h ig h e st p ro p o rtio n of respondents whose l a s t move took them
f a r th e r away from th e ir kin were extended-fam ily o rie n te d .
This d i s t i n c t i o n was c l e a r e s t for those in upper and middle-
c la s s b u re a u c ra tic o c c u p a tio n s.4^
Theorem 16. The newer the area of re sid e n c e , the
c lo se r the resid en ce of kin.
In the only re p o rt found of d ata r e la tin g to t h i s
h y p o th esis, Greer and Kube s ta te d th a t respondents in a low
46
urban (high fam ily) census t r a c t were more l i k e l y than
respondents in the other t r a c t s sampled to have kin p resen t
in the m etropolitan a r e a . 4^
Theorem 17. The younger the head of the household,
the g re a te r the in te ra c tio n with k in .
The fin d in g s rep o rted by Reiss th a t in te r a c tio n de-
48
creased with successive phases of the fam ily cycle sup
p o rt th is theorem.
Theorem 18. The la rg e r the household, the g re ate r
the r e s i d e n t i a l s t a b i l i t y .
R o s s i's fin d in g s concerning the r e la tio n s h ip of
r e s i d e n t i a l m obility and fam ily s iz e are contrary to those
p red ic ted by th is theorem. The la rg e r households in h is
45Ib id .
4^See footnote 17, Chapter I.
4^Scott Greer and E lla Kube, Urban Worlds: A Com
p a ra tiv e Study of Four Los Angeles Areas (Los Angeles: Ac
c id e n ta l C ollege, 1955J, p. 7 *5 "!
48
R eiss, op. c i t .
35
sample were the more mobile. Age and size were independent
ly re la te d to m obility, although age was somewhat more
49
strongly re la te d than household siz e .
Theorem 19. The newer the area of residence, the
g reater the in te ra c tio n with kin.
Mowrer, comparing old and new suburbs, described a
"suburban cycle" which is comparable to the cycle from the
ru ra l to the urban, both with re sp e ct to the family r e l a
tio n sh ip and community o rganization. In consequence, k in
ship asso cia tio n s are emphasized in the new areas, voluntary
asso ciatio n s in the old.'*0 Mowrer defined familism as " v is
itin g the homes of the extended fam ily w ithin the m etropoli
tan area."*^ Relating th is to age of subject and occupa
tio n a l s ta tu s , he concluded th a t the degree of familism is a
function of both the occupational c la ss and the stage in the
suburban cycle.
Using the Shevky measure of urbanization to c la s s i-
53
fy census t r a c t s , Bell and Boat found th a t kin were a
more important source of intim ate so cial contacts in low
^ P e t e r H. Rossi, Why Fam ilies Move; A Study in
the Social Psychology of Urban R esidential M obility (Glen
coe, t i l . : The Free P re ss, 1953), p . 71.
“ ^Mowrer, "The Family in Suburbia," lo c. c i t .
^Mowrer, "Sequential and Class V ariables of the
Family in the Suburban Area," loc. c i t . , 110.
52I b i d . , 111.
^2See footnote 17, Chapter I.
36
urban t r a c t s than in h i g h . ^ G reer, however, found no con
s i s t e n t d iffe re n c e in frequency of in te r a c tio n with k in in
four t r a c t s d i f f e r e n t i a t e d on the Shevky u rb an izatio n
s c a le , when a co n tro l fo r the presence of kin in the m etro
p o lita n area was introduced. (Respondents in the low urban
t r a c t were more l i k e l y to have r e l a t i v e s p rese n t in the
% 55
a re a .)
Jaco and Belknap have proposed the follow ing hy
p o th e sis fo r te s tin g :
A strengthening of s ib lin g as well as p a re n ta l
fam ily ro le s occurs in the frin g e fam ily compared
to the c e n tr a l c ity fam ily. The changed s i g n i f i
cance of the s ib lin g ro le s should extend the range
of kinship a s so c ia tio n among age peers and modify
c u rre n t urban voluntary a s so c ia tio n p r a c t i c e s . 56
Theorem 20. The younger the head of the household,
the g re a te r the r e s id e n tia l s t a b i l i t y .
This theorem is stro n g ly r e je c te d by the a v a ila b le
d a ta . On the b a s is of an an a ly sis of 1950 census d a ta ,
Glick reported th a t both frequency and d ista n ce of moves
were in v ersely re la te d to age. F o u r- f if th s of those who
married changed resid en ce at marriage or during the fo llo w
ing year; th e r e a f te r , the m o b ility r a te decreased sharply
54Wendell Bell and Marion D. Boat, "Urban Neighbor
hoods and Informal S ocial R e la tio n s ," American Journal of
Sociology, IX II (January, 1957), 391-398.
^ G re e r and Kube, op. c i t . , p. 75.
^ E . G artley Jaco and Ivan Belknap, " Is a New Fam
i l y Form Emerging in the Urban Fringe?" American S o cio lo g i
cal Review, XVIII (October, 1953), 557.
37
with number of years married. Only about o n e -fifth of cou
p les married 10 to 15 years moved during any given year. By
that time most of the changes needed to provide increased
liv in g space had already been made, and moving is more d i f
f i c u l t when the fam ily is at i t s maximum s iz e . With the
la te r decrease in family size as children leave home, how
ever, there was no increase in m obility. Rather i t co n tin
ued to decline during the l a t e r years of l i f e , suggesting
th a t older small fam ilies do not o rd in a rily move into sm all-
57
er q u a rte rs. S im ilarly , Rossi found a strong inverse r e
la tio n s h ip between age of head of household and in c lin a tio n
58
toward m obility.
Theorem 21. The newer the area of residence, the
g reater the re s id e n tia l s t a b i l i t y .
The presence of th is re la tio n s h ip in the urban
fringe was hypothesized though not te ste d by Jaco and Bel
knap. Defining the frin g e as the suburbs, s a t e l l i t e c i t
ie s, and adjacent ru ra l non-farm areas which have been
ch aracterized by recen t rapid growth, they proposed th a t
the frin g e , when contrasted to the older c e n tra l c ity , w ill
59
show le s s m obility.
Empirical d ata on m obility by r e s id e n tia l lo c atio n
57
G lick, op. c i t . , 8.
5 8—
Rossi, op. c i t .
59
Jaco and Belknap, op. c i t .
38
c l a r if y the re la tio n s h ip only in p a rt. All stu d ies agree
th a t m obility into suburbs is very high. Between 1950 and
I960, for example, in Standard Metropolitan S t a t i s t i c a l
Areas in the United S ta te s, the population of cen tral c i t
ies increased by an average of 10.7 per cent while popula-
60
tio n in outlying sections increased by 48.6 per cent.
Reporting on 1950 census d a ta , Glick found movement to ru
ra l non-farm areas (o fte n suburban areas) the g re a te st at
a l l ages. Reporting th a t in 1950 they lived in the same
house as they had in 1949 were 81.3 per cent of urban r e
spondents, 78.5 per cent of ru ra l non-farm respondents, and
61
86.3 per cent of ru ra l farm respondents. In Los Angeles
County in 1960, communities which had received the g re a te st
numbers of newcomers in recent years also had the la rg e s t
proportions of housing b u i l t between 1950 and 1960. In
c o n tra s t, communities with few newcomers showed r e la tiv e ly
l i t t l e new h o u s in g .^
All these studies are based on re tro sp e c tiv e or
past m obility and are thus p a r tic u la r ly su ited to measure
m obility into a given area. The data used for hypothesis-
te s tin g in the p resen t study rep o rt subsequent m obility
60Winch, op. c i t . , p. 158.
61
Paul C. Glick, American Fam ilies (New York: John
Wiley and Sons, 1957), p. 90.
6 2
Meeker and H arris, op. c i t . , p. 78.
39
during the year following the i n i t i a l survey in which r e
spondents were c la s s if ie d by area of residence, and thus are
able to focus on m obility out of an area as w ell.
Summary
This chapter has discussed the nature of axiomatic
theory in preface to the form ulation of a theory of urban
familism. The six p o stu la te s of the proposed theory, from
which f if t e e n theorems are derived, p re d ic t interrelation^-
ships between seven v a ria b le s se lected as measures of f a
milism ( a t t i t u d i n a l and b eh a v io ral), of an tic ip a te d s tr u c
tu ra l c o rre la te s of familism, and of r e s id e n tia l lo catio n
and m obility. The concluding se ctio n of the chapter has
presented em pirical support for the p o s tu la te s and theo
rems, noting several theorems for which support was not
available.
The chapter which follow s c l a r i f i e s the procedures
used in te s tin g the theory, providing operational d e f i n i
tio n s for the v a ria b le s and in d ic atin g the s t a t i s t i c a l c r i
te rio n used in te s tin g hypotheses. In ad d itio n , the sample
survey from which d ata were taken is described.
CHAPTER I I I
TESTING THE THEORY: PROCEDURES
The operational adequacy of a theory depends upon
i t s t e s t a b i l i t y . To be used in p red ic tio n and explanation,
a theory must be given an em pirical in te rp r e ta tio n through
o perational d e f in itio n s .^ In the present chapter the v a r i
ables of the proposed theory of urban familism are opera
tio n a lly defined. Further procedures for te s tin g the hy
potheses of the theory are also elu cid a ted : (1) a s t a t i s
t i c a l c r ite rio n of sig n ifican ce i s presented, and (2) the
sample survey from which d ata w ill be drawn is described.
O perational D efin itio n s
The '‘d ic tio n a ry ” of a theory connects i t s terms and
concepts with the data of observation and experience. The
2
connections are estab lish ed through operational definitions.
In the present theory, the d ic tio n ary includes seven
^Clarence Schrag, "Some Notes on Criminological
Theory,” Conference on Research Planning on Crime and De
linquency, ed. William R. Larson (.Los Angeles: Youth Stud
ie s C en ter, U niversity of Southern C a lifo rn ia , 1962), p. 2.
2I b i d . , p . 5.
40
41
concepts. Family o rie n ta tio n and in te ra c tio n with kin w ill
be used as measures of familism. A nticipated s tru c tu ra l
c o rre la te s of familism include age of head of household,
siz e of household, and residence of k i n . The c r ite r io n var
ia b le s , area of residence and r e s id e n tia l s t a b i l i t y , are hy
pothesized to be associated with familism and i t s s tr u c tu ra l
c o rre la te s in the modern m etropolis.
Family O rien tatio n
Family o rie n ta tio n , an a t titu d i n a l v a ria b le , com
p ris e s measures of both extended and nuclear familism and of
th e ir absence. I ts three components have been designated,
3
following Litwak, as extended family o rie n ta tio n , nuclear
family o rie n ta tio n , and non-family o rie n ta tio n . As p re v i
ously s ta te d , for the purposes of th is study familism has
been regarded as a single dimension. I t is an tic ip a te d that:
(1) extended family o rie n ta tio n and nuclear family o rie n ta
tio n w ill tend to occur to g eth er; (2) extended family o rien
ta tio n rep resen ts the p o sitiv e end of the continuum and w ill
not o rd in a rily be found apart from nuclear family o rie n ta
tio n ; (3) nuclear family o rie n ta tio n , when i t occurs alone,
rep resen ts a mid-point on the continuum from p o sitiv e to
^Eugene Litwak, "Occupational M obility and Extended
Family Cohesion," American Sociological Review, XXV (Febru
ary, 1960), 9-21; and ’Geographic M obility and Extended
Family Cohesion," American Sociological Review, XXV (June,
1960), 385-394.
42
negative (non-family o rie n ta tio n . These assumptions to g e th
er rep rese n t one of the im p lic it hypotheses of the research
and w ill be te ste d in the course of te stin g the theory.
Respondents w ill be sorted in to three mutually ex
clusive c a te g o rie s, representing extended, n u clear, and non
fam ily o rie n ta tio n re sp e c tiv e ly , on the b asis of th e ir r e
sponses to the following items in a survey interview : (a)
"People should always get together with r e la tiv e s on h o l i
days and other important sp ecial occasions." All respond
ents who "agree completely" with th i s statement are c l a s s i
fied as extended-family o rie n te d , (b) "A family should eat
together at l e a s t twice a day." Respondents who did not
agree with the preceding statement but who "agree complete
ly" with th is statem ent are c la s s if ie d as nuclear-fam ily
o rie n te d . The residual group of respondents--those who
"agree completely" with n e ith e r of the statem en ts--are
c la s s ifie d as non-family o rie n te d .
In te ra c tio n with Ki)n
In te ra c tio n with kin is measured on the b a sis of
inform ation given in response to the follow ing interview
questions: (a) "Which of your r e la tiv e s liv e in Los
Angeles or Orange Counties at present?" (b) "How often do
you v i s i t or see those re la tiv e s ? "
For a l l respondents w ith any r e la tiv e s liv in g out
side the household and within the m etropolitan area,
43
frequency of in te ra c tio n i s categorized as follow s: (1)
If fa c e -to -fa c e contact with any r e la tiv e is reported as
taking place weekly or more o fte n , in te ra c tio n is High.
(2) I f contact is reported as taking place le ss than once a
week but at le a s t once a month, in te ra c tio n is Medium. (3)
If co n tact is reported as taking place le s s than once a
month, in te ra c tio n is Low.
Age of Head of Household
Age of head of household is u til iz e d in th is study
as the most convenient single measure of stage of family
cycle. I t i s measured, on the b a s is of inform ation given
by the interview respondent, in fiv e age c a te g o rie s: Under
30, 30-39, 40-49, 50-59, and 60 and over. These ca teg o rie s
were judged most usefu l in providing approximations of fam
i l y cycle stages and had the ad d itio n a l advantage of p ro
viding, in the survey sample, f a i r l y comparable N’ s in each
category.
Size of Household
Household siz e is measured, on the basis of in f o r
mation given by the interview respondent, in f iv e ca teg o r
ie s : 1, 2, 3, 4, and 5 or more persons. All persons r e s i
dent in the household are included. Within the survey sam
ple these siz e categ o ries have the advantage of providing
f a i r l y comparable Nfs in each, as well as lending them
selves to combination as the fin d in g s w arrant.
44
At some stag es of the an a ly sis a co n tro l fo r fam ily
type w ith in the household w ill be introduced. This co n tro l
w ill serve to se p arate households which include a married
couple (with or without c h ild re n and with or w ithout other
persons p re se n t) from a l l those in which no married couple
is p re se n t. Such a co n tro l is expected to c l a r i f y c e rta in
family cycle phenomena as needed.
Residence of Kin
Residence of kin i s measured in terms of responses
to the following interview q u estio n s: (a) ’’Excluding r e l a
tiv e s and fam ily members liv in g here in th is place with
you, have you any other r e l a t i v e s or fam ily members liv in g
in Los Angeles or Orange Counties at p re se n t? ” (b) ’’Which
of your r e l a t i v e s liv e in Los Angeles or Orange Counties at
p re se n t? ” (c) ”Do these r e l a t i v e s liv e in th is neighbor
hood?”
I f any r e la tiv e s are p resen t in the neighborhood,
the respondent is c l a s s i f i e d as having kin in the Neighbor
hood . I f r e l a t i v e s liv e w ithin the Standard M etropolitan
S t a t i s t i c a l Area but none are re sid e n t in the neighborhood,
the respondent is c l a s s i f ie d as having kin In SMSA. All
other respondents are c l a s s i f ie d as having kin Outside
SMSA.
Area of Residence
In order to incorporate into the theory hypotheses
45
concerning the r e la tio n s h ip between familism and suburbani
z a tio n , a measure of suburbanization i s req u ired . The
search for such measures has e n lis te d widespread e f f o r t s ,
r e f le c te d in changes and c r itic is m s of census c a te g o rie s as
well as in v aried d e f in it io n s of the f rin g e , the suburb,
and the c i t y . Measures in use g e n e ra lly r e f l e c t d e n sity ,
d ista n ce from the c ity c e n te r, p o l i t i c a l boundaries, or a
combination of th ese.
The ’’date b u i l t up” measure proposed by Duncan,
Sabagh and Van Arsdol, and f i r s t applied to Los Angeles,
avoids many of the disadvantages of other measures, espe
c i a l l y when applied to a r e l a t i v e l y low -density, d e c e n tr a l
ized m etropolis. Analogous to the cohort a n a ly sis of f e r
t i l i t y , i t c h a ra c te riz e s sub-areas in the m etro p o litan area
on the b a s is of s t a t i s t i c s on year b u i l t and gross acreage
for census t r a c t s , plu s a d w e llin g -u n it d en sity c r ite r i o n
to id e n tify the b u ilt-u p areas a t p ast d a te s . A census
t r a c t is defined as " b u ilt-u p " when i t reaches a d e n s ity of
at le a s t two dw ellings per acre of gross land a r e a .4 In
the d ata a n a ly sis in the p re se n t study, census t r a c t s are
categorized as b u i l t up before 1940, 1940-49, 1950-59, and
1960 and a f t e r .
4Beverly Duncan, Georges Sabagh, and Maurice D. Van
Arsdol, J r . , ’'P a tte rn s of C ity Growth," American Journal of
Sociology, LXVII (January, 1962), 418-429.
R esidential S ta b ility
The survey from which d ata for te stin g the theory
are drawn was focused on r e s id e n tia l m obility. The o r i g i
nal sample, interviewed in 1961, were asked about th e ir
plans and d esire s to move: Ca) "Do you plan to move from
th is place w ithin the next year?" (b) " I f you had your
choice, would you stay here or move?” Responses to these
questions provide measures of planned and choice m obility.
Actual m obility is ascertained through the lo n g itu d in al de
sign of the study. Persons remaining, a year la te r in
1962, in th e ir 1961 residence are c la s s if ie d as non-movers;
those s h iftin g to another residence, w ithin or outside the
m etropolitan area, during the year are c la s s ifie d as mov
e rs.
For h y p o th e sis-testin g in the present research,
th is measure of actual m obility is used. Non-movers during
the one-year period are categorized as S ta b le ; movers are
categorized as Mobile.
S t a t i s t i c a l Test of the Hypotheses
There is considerable disagreement concerning the
use of s t a t i s t i c a l te s ts of sig n ifican ce in survey r e
search. S elv in , for example, has argued against the use of
v ir tu a lly a l l such te s t s , s ta tin g th a t s t a t i s t i c a l analysis
should focus on the d ire c tio n and size of asso cia tio n s
47
rath e r than on sig n ific a n c e .^ On the other hand, Gold, r e
plying to S elvin, has pointed out th at in d e s c rip tiv e r e
search, where c o rre la tio n not causation is the point at i s
sue, te s ts of sig n ifican ce are appropriate to indicate
whether a d ifferen ce does e x is t. Kish has also supported
the use of s t a t i s t i c a l te s ts in checking the e ffe c ts of ex
planatory against randomized v a ria b le s, while pointing out
the necessity of separating out the e ffe c ts of other con
founding v ariab les through the logic of the research de-
* >
7
sign.
The l a t t e r viewpoint is adopted in the present r e
search. At th is point in the development of the theory of
urban familism, the primary question is seen as the extent
to which te n ta tiv e re je c tio n of the n u ll hypotheses of no
asso ciatio n i s appropriate. Because i t i s not an ticipated
th a t many of the asso ciatio n s between v ariab les w ill be of
great magnitude, emphasis w ill be placed on te stin g for the
existence of a d ifferen ce ra th e r than on measuring the size
^Hanan C. Selvin, M A C ritique of Tests of S ig n if i
cance in Survey Research,” American Sociological Review,
XXII (October, 1957), 519-527.
David Gold, ’’Comment on ’A C ritique of T ests of
S ig n ific a n c e ,1” American Sociological Review, XXIII (Febru
ary, 1958), 8 5 -8 ^
7
Leslie Kish, ’’S t a t i s t i c a l Problems in Research De
s ig n ,” American Sociological Review, XXIV (June, 1959),
328-337.
48
of the a s so c ia tio n s.^
Of commonly used s t a t i s t i c a l t e s t s , only the X2
t e s t fo r independence appears to offer an appropriate model
for the p resen t research , in which the d ata for analysis
w ill be presented in contingency ta b le s of varying size.
Consequently, in examining observed re la tio n s h ip s between
v a ria b le s, the n u ll hypothesis of no asso ciatio n w ill be
tested in each instance by X2 . If the p ro b a b ility of ob-
taining a given value of X is .05 or le s s , the n u ll hy
pothesis w ill be re je c te d and the a lte rn a tiv e hypothesis
from the theory w ill be te n ta tiv e ly accepted, provided the
asso ciatio n between the v aria b le s is in the d ire c tio n p re
dicted by the theory.
In addition, a measure of asso ciatio n for each t a
ble w ill be given by the contingency c o e ffic ie n t (C). This
s t a t i s t i c is recognized as having lim ited usefulness; i t
w ill, however, provide a simple chi-square based measure of
asso cia tio n , which i s useful in comparing sizes of obtained
Q
In another study in which an axiomatic theory was
developed and te ste d , Schwirian and Prehn ju s t if i e d th e ir
use of an a rb itra ry minimum c o rre la tio n value, in place of
a t e s t of sig n ific a n c e , on the grounds th a t the size of the
population made s t a t i s t i c a l sig n ifican ce almost in e v ita b le
in every in stan ce. This argument does not apply to the
data of the present research. See Kent P. Schwirian and
John W . Prehn, "An Axiomatic Theory of U rbanization,” Amer
ican Sociological Review, XXVII (December, 1962), 812-525.
49
d if fe re n c e s among the hypotheses of the th e o r y .9 An im
p o rta n t l i m i t a t i o n of th e contingency c o e f f ic ie n t is i t s
f a i l u r e to a t t a i n u n ity ; in s te a d , i t s upper l i m i t i s a
fu n c tio n of the number of c a te g o r ie s . Two contingency co
e f f i c i e n t s are comparable only when yield ed by ta b le s of
the same siz e . Because the ta b le s in which d a ta of the
p re se n t rese arch are re p o rte d are of varying s iz e , a c o r
re c tio n has been introduced to render the c o e f f i c i e n t s more
d i r e c t l y comparable. Not only C, but also C/Cmax,10 is r e
ported fo r each ta b le .
Costner has recommended d isc o n tin u in g the use of
’'ch i-sq u a re based” measures of a s s o c ia tio n , on the grounds
th a t t h e i r in t e r p r e t a t i o n is e x c e ss iv e ly vague. Gamma
would probably be the most ap p ro p ria te measure of a s s o c ia
tio n fo r the d a ta of the p re se n t re se arc h ; i t has not been
used because of the complex computation req u ired and be
cause of the d e c is io n to focus s t a t i s t i c a l a n a ly s is around
a t e s t of s ig n if ic a n c e . See H erbert L. C ostner, ’’C r i t e r i a
for Measures of A s s o c ia tio n ,” American S o cio lo g ica l Review,
XXX (June, 1965), 341-353.
fk-1 ^Cmax, the upper lim it fo r C, is eq u iv ale n t to
v"Tc” , where k is eq u iv a le n t to the number of rows or c o l
umns. This form ula assumes a square ta b le with an equal
number of rows and columns. In the p re s e n t study not a l l
ta b le s are square; in such cases the sm aller number of rows
or columns has been taken as k to provide a con serv ativ e
e s tim a te . Cmax is as follow s fo r ta b le s of the s iz e s given
in t h i s paper: 2 X n = .707; 3 X n = .816; 4 X n = .866;
5 X n = .894. For a f u l l e r d isc u ssio n of th e contingency
c o e f f ic ie n t see Sidney S ie g e l, Nonparametric S t a t i s t i c s for
the B ehavioral S ciences (New York: Mc&raw-Hill Book Co.,
Inc.', 1<5'5'6), pp. I W - S t e .
50
D escrip tio n of the Survey Design
Data for te s tin g the theory are taken from a 1961-
62 study of r e s i d e n t i a l m obility in the Los Angeles Metro
p o lita n S t a t i s t i c a l A r e a .^ The study u t i l i z e d an area
p ro b a b ility sample of a l l ad u lt r e s id e n ts of housing u n its
w ithin the SMSA, which includes Los Angeles and Orange
12
C ounties. The sample, interview ed in the summer of 1961,
co n sisted of 981 r e s id e n ts in 300 of the 1,403 census
t r a c t s in the area . Interview s were completed with a ra n
dom sample of respondents, both male and female, in 90.3
per cent of the s e le c te d occupied housing u n its . Compari
sons with 1960 census d ata in d ic a te d th a t the sample is
re p re s e n ta tiv e of the to t a l m etro p o litan area population
with the exception of the Negro p o p u latio n , which was
The research was supported by the John Randolph
and Dora Haynes Foundation of Los Angeles and the Research
and P u b lic a tio n Fund of the U n iv ersity of Southern C a li
fo rn ia .
1 ?
The sample design involved s e le c tio n of a proba
b i l i t y sample of approximately 1,250 housing u n its a l l o
cated as follow s: one housing u n it per block, two blocks
per enumeration d i s t r i c t , and two enumeration d i s t r i c t s for
each of the a n tic ip a te d 300 census t r a c t s in the sample.
Census t r a c t s , enumeration d i s t r i c t s , and blocks were a l l
se le c te d with p r o b a b ili tie s p ro p o rtio n a l to s iz e . The de
sign included geographic s t r a t i f i c a t i o n of census t r a c t s ,
estim atio n of block siz e w ithin sampled enumeration d i s
t r i c t s , p r e - l i s t i n g of sampled blocks, and compensations
for p r e - l i s t i n g e r r o r s . Lack of 1960 Census d a ta for block
siz e n e c e s s ita te d the use of enumeration d i s t r i c t s for
which f i e l d inform ation was secured on the number of hous
ing u n its .
51
13
s e rio u s ly underrepresented.
A lo n g itu d in a l design provided fo r rein te rv iew in
the summer of 1962 of the panel of respondents o r ig in a lly
contacted in 1961. Through the use of telephone, mail
q u estio n n aire and interview procedures in the 1962 phase of
the study i t was p o ssib le to c la s s if y a l l respondents as
movers, non-movers, or deceased. The deaths of two r e
spondents re s u lte d in a f in a l sample of 979.
While the Los Angeles m etropolitan area has a high
er in tram etro p o litan r e s id e n tia l m obility r a te than other
m etropolises, and d i f f e r s in other re sp e c ts as w ell, i t
should not n e c e s s a rily be regarded as a ty p ic a l. Los
Angeles is sim ilar to other m etropolitan areas in the
United S ta te s in terms of i t s p a tte rn of so c ia l organiza
tio n and environmental u t i l i z a t i o n . Urban phenomena ac
companying growth are thus magnified and in te n s if ie d in Los
Angeles, and the m etropolis has come to symbolize metropol-
14
ita n development in the United S ta te s . S im ila r itie s be
tween Los Angeles and other m etropolitan areas suggest th a t
the fin d in g s presented J.n th is study may apply to the popu
l a tio n s of other m etropolitan areas as w ell.
13
A complete rep o rt of the research design is given
in Georges Sabagh and Maurice D. Van Arsdol, J r . , e t a l . ,
The R e stle ss M etropolis: R e sid en tial M obility in Tos
Angeles (.Los Angeles: U niversity o± Southern C a lifo rn ia ,
19o5.), mimeo.
^ D u n c a n , Sabagh, and Van A rsd o l, op. c i t .
52
Summary
This chapter has presented o p eratio n al d e f in itio n s
for the v a ria b le s of the rese arch , discussed the s t a t i s t i
cal c r ite r i o n adopted fo r te s tin g hypotheses, and described
the survey from which d a ta are taken. In the following
ch ap ter, an aly sis of the d ata in terms of the p o stu la te s
and theorems of the theory w ill be begun.
CHAPTER IV
DATA ANALYSTS
The hypotheses of the proposed theory of urban f a
milism were te s te d with d a ta from the sample survey of Los
Angeles m etropolitan area households described in the p re
ceding ch ap ter. In th is chapter the em pirical fin d in g s for
each hypothesis are presented in ta b u lar form and b r i e f l y
d iscu ssed . In Chapter V a more in te n siv e an a ly sis of the
fin d in g s in terms of the theory as a whole w ill be under
taken.
Empirical Findings: P o s tu la te s
P o stu la te 1 . The newer the area o f resid en ce, the
younger the head of the household.
Area of residence was o r ig in a lly analyzed in four
ca te g o rie s: pre-1940, 1940-49, 1950-59, and 1960 and a f
te r. This an aly sis produced the somewhat in c o n siste n t p a t
tern in g seen in Table 1.
Households with the head under 30 c o n s titu te d com
parable proportions of each area of residence category.
Households in which the head was between 30 and 49 appeared
to make up the m ajority of re s id e n ts in t r a c t s b u ilt up in
53
TABLE 1
Age of Head
of Household
AGE OF HEAD OF HOUSEHOLD BY AREA OF RESIDENCE
Area of Residence
(Date B u ilt Up)
Before
1940 1940-49 1950-59
1960
and a fte r
Under 30 12.9% 13 .5% 11.9% 14.3%
30-39 20.0 19.3 29.9 26.0
40-49 19.4 27.4 31.4 28.6
50-59 17.7 2 1 .6 15.3 19.4
60 & over 30.0 18.2 11.5 11.7
100.0 (351) 100.0 (171) 100.0 (261) 100.0 (196)
N = 979
Mdn. Age 48.8________________46.3______________42.6___________ 43.4__________
df = 12 X2 = 1.94 > .0 5 C = .045 C/C max = .052
55
the f i f t i e s and s i x t i e s ; however, t h i s a s s o c ia tio n was not
s t a t i s t i c a l l y s ig n i f ic a n t . Households with the head aged
50-59 were approxim ately equally li k e l y to be found in each
residence category. Households with the head aged 60 or
over were most l i k e l y to be found in the o ld e st are a s.
The d i s t r i b u t i o n of c e ll values in the ta b le sug
gested dichotom izing the area of resid en ce c a te g o rie s , as
has been done in Table 2. In th is ta b le the c u rv ilin e a r
p a tte rn emerges more c le a r ly , and the a s s o c ia tio n between
v a ria b le s is s ig n i f i c a n t .
The f a i l u r e of the under-30 group to be found in
g re a te r numbers in the newer areas probably r e f l e c t s a fam
i l y cycle phenomenon, namely, the la rg e prop o rtio n of un
married and newly married persons in th is group. Such p e r
sons tend to re n t ra th e r than to own; most new settle m en t
areas are inhabited by home owners. The hypothesis is sup
ported, modified as follow s: The newer the area of r e s i
dence, the younger the head of the household, with the ex
ception of households in which the head is under 30.
P o s tu la te 2 . The younger the head of the household,
the la rg e r the household.
The lo g ic a l b a s is of t h i s hypothesis is the fam ily
cycle. However, the family cycle would be expected to pro
duce a c u rv ilin e a r p a tte rn of fam ily s iz e , while the hy
p o th e sis is s ta te d in lin e a r term s. I t was a n tic ip a te d
th a t some m o d ificatio n of the hypothesis would be
TABLE 2
AGE OF HEAD OF HOUSEHOLD BY AREA OF RESIDENCE
Area of Residence
(Date B u ilt Up)
Before 1950 1950 and a fte r
Under 30 13.1% 1 2.9%
30-39 19.7 28.2
Age of Head 40-49 22.1 30 .2
of Household 50-59 19.0 17.1
60 & over 26.1 11.6
100.0 (522) 100.0 (457)
Mdn. Age 47.8 43.1
df = 4 X 2 = 40.85 < .001 C = .200 C/C i
= 979
.283
L n
O '
57
necessary. Table 3 in d ic a te s th a t m odification is required
to a lim ited ex te n t, with resp e ct to the under-30 age c a te
gory. This category contains s u b s ta n tia l numbers of the
unmarried and the newly married w ithout ch ild re n , and con
sequently r e l a t i v e l y larg e p ro p o rtio n s of one- and two-
person households. This category also includes a number of
couples who have not completed c h ild -b e a rin g , a f a c t r e
f le c te d in the r e l a t i v e l y low proportion of households with
four or more persons.
C u rv ilin e a rity at the beginning of the cycle ap
pears only in th is age category; beginning with the 30-39
age category the hypothesis holds throughout. The p e rc e n t
age of one- and two-person households in each successive
age category in creases c o n s is te n tly , while the percentage
of households with four or more persons d ecreases. Three-
person households, the interm ediate s iz e , remain as an ap
proximately constant percentage in each age category up to
age 60, when they decrease in number.
In Table 4 household size i s presented in three
ca teg o ries only, for g reat c l a r i t y . The same p a tte rn ap
p ears. Table 5 introduces a control for fam ily type. The
d ata are shown se p a ra te ly for households with a married
couple (with or w ithout c h ild re n ), om itting a l l other types
of households. The same p a tte rn appears, with a g re a te r
proportion of larger households. The o rig in a l hypothesis
may be r e s ta te d as follow s: The younger the head of the
TABLE 3
HOUSEHOLD SIZE BY AGE OF HEAD OF HOUSEHOLD
Age of Head
Household Size Under 30 30-39 40-49 50-59 60 & over
1 15.7% 5.6% 7.1% 16.3% 39.5%
2 26.8 15.5 21.3 37.0 45.7
3 20.5 17.7 19.7 18.0 9.5
4 21.3 25.0 23.2 16.9 11.6
5 or more 15.7 36.2 28.7 11.8 3.7
100.0 100.0 100.0 100.0 100.0
(127) (232) (254) (177) (189) N = 979
df = 16 X2 = 238.67 < .0 0 1 C = .442 C/C max = .494
U\
o o
TABLE 4
HOUSEHOLD SIZE BY AGE OF HEAD OF HOUSEHOLD
Age of Head
Household Size Under 30_______30-39_____ 40-49_______50-59_______60 & over
Small
(1 - 2) 42.5% 21.1% 28.4% 53.3% 85.2%
Medium
(3) 20.5 17.7 19.7 18.0 9.5
Large
(4 or more) 37.0 61.2 51.9 28.7 15.3
100.0 100.0 100.0 100.0 100.0
Cl27) (232) (254) (177) (189) N = 979
df = 8 X2 = 235.86 <.001 C = .440 C/C max = .546
Ul
v O
TABLE 5
HOUSEHOLD SIZE BY AGE OF HEAD OF HOUSEHOLD
(MARRIED COUPLES ONLY)
Age of Head
Household Size Under 30_______30-39_____ 40-49_______50-59_______60 & over
Small
(1 - 2) 26.7% 11.0% 20.6% 41.6% 76.2%
Medium
(3) 24.2 18.3 20.6 21.2 13.9
Large
(4 or more) 48.8 70.8 58.8 37.2 9.9
100.0 100.0 100.0 100.0 100.0
(90) (191) (204) (132) (101) N = 718
df = 8 X2 = 165.31 -C.001 C = .432 C/C max = .529
O '
o
61
household, the la rg e r the household, except for households
with the head under 30.
P o s tu la te 3 . The la rg e r the household, the
g re a te r the fam ily o r ie n ta tio n .
This hypothesis re ce iv e s lim ite d support from the
d ata. When percentages are computed on the b a s is of f iv e
household siz e c a te g o rie s , c e ll values f lu c tu a te to such an
ex ten t th a t i t is d i f f i c u l t to d isc e rn any p a tte r n . None
of the d iffe re n c e s is s t a t i s t i c a l l y s ig n i f ic a n t . (See Ta
ble 6 .)
When household siz e is dichotomized (Table 7 ), a
p a tte rn emerges, but i t is only p a r tly th a t p re d ic te d .
Contrary to the hypothesis, small households are more l i k e
ly than la rg e households to be extended fam ily -o rien ted .
The hypothesis holds, however, for nuclear family o r ie n ta
tio n , which appears more fre q u e n tly in la rg e than in small
households. Non-family o rie n ta tio n does not vary s i g n i f i
c an tly by size of household.
P o s tu la te 4 . The g re a te r the fam ily o rie n ta tio n ,
the closer the residence of kin.
The hypothesis is p a r t i a l l y supported, to the ex
te n t th a t respondents with r e l a t i v e s liv in g in the neigh
borhood were s ig n i f ic a n t ly more lik e ly to be extended
fa m ily -o rien ted than respondents without nearby r e l a t i v e s
(fo r th is d iffe re n c e alone, X = 7.27, < .0 5 ). No s i g n i f i
cant d iffe re n c e in fam ily o rie n ta tio n appeared, however,
TABLE 6
FAMILY ORIENTATION BY HOUSEHOLD SIZE
Household Size
1
2
3
4
5 or more
Extended
56.0%
50.7
54.8
36.9
44.1
Family O rien tatio n
Nuclear
23.0%
29.1
24.1
37.5
32.1
Non
21. 0%
20.2
20.8
25.6
23.8
100.0% (153)
100.0 (278)
100.0 (167)
100.0 (175)
100.0 (206)
(T o ta l: 979)
df = 8 X2 = 14.81 > .0 5 C = .122 C/C max = .150
TABLE 7
FAMILY ORIENTATION BY HOUSEHOLD SIZE
Family O rie n ta tio n
Household Size____________ Extended___________ Nuclear_______________Non
Small
(1-3)* 53.1% 26.3% 20.6% 100.0% (598)
Large
(4 or more) 40.8 34.6 24.6 100.0 (381)
(N = 979)
df = 2 X2 = 14.49 < . 0 5 C = .374 C/C max = .529
*Almost id e n tic a l values appear i f the Small category
is r e s t r i c t e d to 1-2.
64
between those with r e l a t i v e s lo cate d in the m etropolitan
area as a whole and those without r e l a t i v e s in the area.
Taken as a whole, the d iffe re n c e s in Table 8 are not s ig n i
f ic a n t.
P o stu la te 5 . The clo ser the residence of kin, the
g re ater the in te ra c tio n with kin.
This hypothesis was te ste d only fo r respondents
with kin liv in g in the m etropolitan area, or s l i g h t l y over
th re e -fo u rth s of the to ta l sample. The fin d in g s of other
stu d ies in d ic a te c le a rly th a t f a c e -to -fa c e in te ra c tio n with
r e la tiv e s outside the area w ill be le s s frequent than th a t
which occurs with kin in the are a .^ This finding should be
e s p e c ia lly applicable to a geographic area as la rg e as Los
Angeles--Orange Counties; consequently respondents with no
kin in the area were omitted from t h i s a n a ly s is . As Table
9 shows, the hypothesis i s stro n g ly supported.
P o stu la te 6 . The greater the in te ra c tio n with k in ,
the g re a te r the r e s id e n tia l s t a b i l i t y .
This hypothesis is not supported by the d a ta . No
d iffe re n c e s in s t a b i l i t y appear in Table 10 by frequency of
in te ra c tio n with kin. I t is important to note, however,
th a t in te r a c tio n has been measured only fo r those
^See Paul J . R eiss, ’’The Extended Kinship System:
C o rre la tes of an A ttitu d e s on Prequency of I n t e r a c t i o n ,”
Marriage and Family L iv in g , XXIV (November, 1962), 335-336;
and Eugene Litwak, "Geographic M obility and Extended Family
C ohesion,” American S ociological Review, XXV (June, 1960),
3 8 8 . --------------------------- --------------------
Family O rientation
Extended
Nuclear
Non
TABLE 8
RESIDENCE OF KIN BY FAMILY ORIENTATION
Kin liv e
In Neighborhood In SM SA_______ O utside SM SA
36.7% 43.0% 20.3% 100.0 (474)
32.7 46.6 20.7 100.0 (288)
26.4 47.7 25.9 100.0 (217)
(N = 979)
df = 4 I f ? - 8.15 >.05 C = .028 C/C max = .034
o
U\
TABLE 9
INTERACTION WITH KIN BY RESIDENCE OF KIN
Kin liv e
In te ra c tio n ______________In neighborhood______In SMSA outside neighborhood
High 84.5% 43.3%
Medium 11.6 29.8
Low 3.9 26.9
100.0 (325) 100.0 (441) N = 766
df = 2 X2 = 138.6 < .0 0 1 C = .387 C/C max = .547
O '
O '
TABLE 10
RESIDENTIAL STABILITY BY INTERACTION WITH KIN
I n t e r a c t i o n ________________Mobile______________________Stable
79.2% 100.0 (417)
80.3 100.0 (178)
79.5 100.0 (171)
N = 766
df = 2 X2 = 0.12 > .0 5 C = .000 C/C max =
High 20.8%
Medium 19.7
Low 20.5
.000
O '
68
respondents with kin in the a re a . Frequency of in te ra c tio n
i s c lo s e ly asso cia te d with lo c a tio n of resid e n ce of k in .
When the l a t t e r is measured, the expected a s so c ia tio n with
r e s i d e n t i a l s t a b i l i t y i s found. (See Theorem 11.)
Em pirical Findings: Theorems
Theorem 7 . The newer the area of resid e n c e ,
the la rg e the household.
The a n tic ip a te d r e la tio n s h i p between area of r e s i
dence and household s iz e appears in the d a ta . One-person
households were h ea v ily concentrated in the older area s and
were found in fre q u e n tly in the newer; two-person households
were ail so more l i k e l y to be found in older a re a s. Three-
person households, the in te rm ed ia te s iz e , were found in ap
proxim ately the same p ro p o rtio n in a l l a re a s. Larger
households were more h ea v ily concentrated in newer a re a s,
although th is c o n c e n tra tio n appears to be more marked for
fo u r-p erso n than fo r la rg e r households. Table 11 shows the
d i s t r i b u t i o n of household s iz e for four area of resid en ce
c a te g o rie s ; in Table 12 area of residence has been d ic h o to
mized. Collapsing siz e c a te g o rie s as w ell (Table 13) f u r
ther c l a r i f i e s the p a tte r n while masking no s ig n i f ic a n t
d if fe re n c e s . A nalysis of the d ata for m arried-couple house
holds only, yielded the same p a t te r n . (Table not shown.)
Theorem 8 . The younger the head of the household,
the g re a te r the fam ily o r ie n ta tio n .
This hypothesis is p a r t i a l l y supported and
TABLE 11
HOUSEHOLD SIZE B Y AREA OF RESIDENCE
Area of Residence
(Date B u ilt Up)
Household Size
Before 1940- 1950- 1960
1940 1949 1959 and a f te r
1 26.5% 12.3% 8.4% 8 .2%
2 30.0 34.5 29.3 24.0
3 17.2 15.8 16. 8 19.6
4 10.3 18.1 21.1 25.6
more 16.0 19.3 24.4 25.6
100.0 100.0 100.0 100.0 N = 979
(351) (171) (261) (196)
df = 12 x 2 = 74.26 < .001 C = .265 C/C max = .306
O'
o
TABLE 12
HOUSEHOLD SIZE BY AREA OF RESIDENCE
Area of Residence
(Date B u ilt Up)
Household Size
Before
1950
1950
and a fte r
1 21.9% 8.3%
2 31.4 27.0
3 16.9 16.8
4 12. 8 23.0
more 17.0 24.9
100.0 (522) 100.0 (457) N = 979
df = 4 X2 = 52.83 < .001 C = .224 C/C max .317
o
TABLE 13
HOUSEHOLD SIZE BY AREA OF RESIDENCE
Area of Residence
(Date B u ilt Up)
Before
1950
1950
and a fte r
Small (1-2) 53.3% 35.3%
Household Size Medium (3) 16.8 16.8
Large (4 or more) 29.9 47.9
100.0 (522) 100.0 (457) N = 979
df = 2 X2 = 38.66 < .001 C = .191 C/C max = .270
72
p a r t i a l l y re je c te d by the fin d in g s, depending on the type
of fam ily o rie n ta tio n being measured. Nuclear fam ily o r i
e n ta tio n decreased with increased age of the head of the
household, as p red icted by the hypothesis. Extended fam ily
o r ie n ta tio n increased c o n s is te n tly with increased age of
the head, in c o n tra s t to the re la tio n s h ip p re d ic te d . Non
family o r ie n ta tio n flu c tu a te d considerably with age, but no
c o n s iste n t p a tte r n could be seen. (See Table 14.)
Theorem 9 . The la rg e r the household, the
closer the residence of k in .
The hypothesis rece iv es p a r t i a l but inadequate sup
port from the d a ta . The only c o n s iste n t d iffe re n c e as p re
dicted is for sin g le-p erso n households. As Table 15 shows,
these are much le s s l ik e ly to have r e l a t i v e s in the metro
p o lita n area; and those with r e la tiv e s in the m etropolitan
area are le s s li k e l y to have them in the neighborhood.
This d iffe re n c e , however, probably r e f l e c t s fam ily type
ra th e r than siz e per s e , since the m ajority of a l l house
holds of two or more persons include a married couple.
Two-person household members are also le s s li k e l y than
those in la rg e r households to have r e l a t i v e s in the neigh-
borhood--a d iffe re n c e which holds i f a co n tro l for fam ily
type is introduced. (Table not shown.) However, they are
as lik e ly as la rg e r households to have r e l a t i v e s in the
m etropolitan area. While th e re is some flu c tu a tio n in the
proportions of la rg e r households with r e l a t i v e s in the
TABLE 14
FAMILY ORIENTATION BY AGE OF HEAD OF H OU SEH OLD
100.0 (127)
100.0 (232)
100.0 (254)
100.0 (177)
100.0 (189)
N = 979
df = 8 X2 = 35.52 .001 C = .187 C/C max = .229
Family O rien tatio n
Extended________Nuclear___________ Non
Under 30 39.8% 33.6 26.6
30-39 42.7% 36.6 20.7
Age of Head 40-49 46.3% 30.8 22.9
50-59 47.2% 24.7 28.1
60 & ever 65.0% 20.6 14.4
TABLE 15
RESIDENCE OF KIN 3Y HOUSEHOLD SIZE
Kin liv e
nei
in
ghborhood in SMSA
outside
SM SA
1 23.8% 43.5 32.7 100.0 (153)
2 27.6% 51.8 20.6 100.0 (278)
Household Size 3 41.8% 35.2 23.0 100.0 (167)
4 41.0% 43.8 15.2 100.0 (175)
5 or more 35.2% 45.1 19.7 100.0 (206)
N = 979
df = 8 X2 = 33.37 < .0 0 1 C = .185 c/c! max .227
<1
75
area, no c o n s is te n t p a tte rn appears. In s p ite of the s i g
n ific a n c e le v e l of chi square fo r the ta b le , the hypothesis
cannot be regarded as adequately supported.
Theorem 10. The g re a te r the fam ily o r ie n ta tio n ,
the g re a te r the in te r a c tio n with k in .
The hypothesis is not supported by the d a ta . While
high in te ra c tio n appeared to in c re ase with in c re a sin g fam
i l y o r ie n ta tio n , the d if fe re n c e s in Table 16 are not s t a
t i s t i c a l l y s i g n i f i c a n t .
Theorem 11. The c lo ser the resid en ce of k in , the
g re a te r the r e s id e n tia l s t a b i l i t y .
This hypothesis is supported fo r lo c a tio n of r e l a
tiv e s in sid e or o u tsid e of the m e tro p o lita n area, although
no d iffe re n c e appears fo r closeness of lo c a tio n w ithin the
area (in neighborhood/not in neighborhood). In Table 17
c e ll values in the th ir d column d i f f e r s i g n i f i c a n t l y from
values in the f i r s t two columns; in Table 18 the f i r s t two
columns have been combined in order to c l a r i f y the p a tte r n .
Further breakdown of the mobile group by d ir e c tio n
of move might be expected to emphasize the r e la tio n s h i p be
tween lo c a tio n of r e l a t i v e s in the SMSA and r e l a t i v e d i s i n
c lin a tio n to move o u tsid e the area . The g reat m ajo rity of
moves rep o rted were intram e t r o p o l i t a n , and t h i s tendency to
move w ith in , ra th e r than out o f, the m etropolis appeared
somewhat g re a te r fo r those with kin in the area than for
those w ithout, as shown in Table 19. However, the X for
Family O rientation
TABLE 16
INTERACTION WITH KIN BY FAMILY ORIENTATION
Interaction
High___________ Medium Low
Extended 57.3% 22.0 20.7 100.0
Nuclear 53.3% 24.0 22.7 100.0
Non 49.7%____________ 26.1________24.2 100.0
N = 766
df = 4 X2 = 2.87 > .0 5 C = .063 C/C max =
Mobile
Stable
TABLE 17
RESIDENTIAL STABILITY BY RESIDENCE OF KIN
Kin liv e
in neighborhood_________ in SM SA_________outside SM SA
19.2%
81.8
2 1 . 3%
78.7
32.3%
67.7
df = 2
100.0 (325)
2 _
= 11.21
100.0 (441)
< .02 C :
100.0 (213)
.105 C/C max
(376)
(229)
(161)
.077
N = 979
= .149
-< i
O'
Move was
TABLE 18
RESIDENTIAL STABILITY BY RESIDENCE OF KIN
Kin liv e
in SM SA outside SMSA
Mobile 20.5% 32.3%
Stable 79.5 67.7
100.0 (766) 100.0 (213) N = 979
df = 1 X2 = 7.79 < .01
TABLE 19
C = .089 C/C max =
DIRECTION OF MOBILITY BY RESIDENCE OF KIN
Kin liv e
in SMSA outside SMSA
w ithin county 86.7% 78.3%
out of county 13.3 21.7
100.0 (158) 100.0 (68) N = 226
df = 1 X2 = 1.70 > .0 5 C = .089 C/C max =
78
th is ta b le was not s ig n i f ic a n t a t the .05 le v e l ; small c e ll
values for o u t-o f-c o u n ty moves (20 and 13, r e s p e c tiv e ly )
l im it the g e n e ra liz a tio n s th a t might be made from th ese
d a ta .
I t is ev id en t from t h i s a n a ly sis and th a t of d a ta
on in te r a c ti o n with k in (P o s tu la te 6) th a t presence of kin
w ithin the m e tro p o litan area i s p o s it iv e l y a s so c ia te d with
r e s i d e n t i a l s t a b i l i t y , but clo sen ess of resid e n ce of kin
w ith in the area is n o t. Respondents with k in in the imme
d ia te neighborhood re p o rted g re a te r frequency of f a c e -to -
face in te r a c ti o n than respondents with kin in the SMSA but
not in the neighborhood (P o s tu la te 5); t h i s d iffe re n c e is
n o t, however, r e f l e c t e d in increased r e s i d e n t i a l s t a b i l i t y .
I n te r a c tio n s u f f i c i e n t to have a s t a b i l i z i n g e f f e c t appar
e n tly tends to occur with kin lo c a te d anywhere w ithin the
m etro p o litan area.
Theorem 1 2 . The newer the area of re sid e n c e ,
the g re a te r the fam ily o r ie n ta tio n
This h y p o th esis i s not supported by the d a ta .
Only f a c e - to - f a c e in te r a c ti o n was measured in th is
study. As Sussman and Burchinal have suggested, however,
other forms of in te r a c ti o n which r e l y on modern means of
communication may be eq u ally e f f e c t iv e in m aintaining a kin
fam ily network. W ithin the m e tro p o litan area telephones
are probably e f f e c t i v e l y used for t h i s purpose. See Marvin
B. Sussman and Lee B urchinal, "Kin Family Network: Unher
alded S tru c tu re in C urrent C o n c ep tu alizatio n s of Family
F u n ctio n in g ," M arriage and Family L iv in g , XXIV (August,
1962), 231-240.
79
While nuclear family o rie n ta tio n increased with newness of
area of residence, as p red ic ted , these d iffe re n c e s taken
alone were not q u ite s ig n if ic a n t. Extended family o rie n ta
tio n decreased s ig n ific a n tly with newness of area of r e s i
dence, in c o n tra st to the predicted r e la tio n s h ip . The mag
nitude of the l a t t e r negative asso c ia tio n is such th a t the
X2 values for both ta b le s are s ig n if ic a n t. In Table 21 the
four area of residence ca teg o ries of Table 20 have been
dichotomized, to conform to the p a tte rn followed in the an
a ly s is of the other p o stu la te s and theorems which include
th is v a ria b le .
Theorem 13. The younger the head of the household,
the clo ser the residence of kin.
No d iffe re n c e in closeness of kin residence by age
of head of household was found, with the exception th a t
households with the head aged 60 or over were le ss lik e ly
than younger households to have kin liv in g in th e ir immedi
ate neighborhood. Older households were, however, as l i k e
ly as others to have kin in the to t a l m etropolitan area.
(Data are shown in Table 22.) The hypothesis is not sup-
por ted.
Theorem 14. The larger the household, the
g re a te r the in te ra c tio n with kin.
The hypothesis is not supported by the d ata. While
Table 23 suggests a c u rv ilin e a r p a tte rn , with th re e - and
four-person households more lik e ly than la rg e r or smaller
TABLE 20
FAMILY ORIENTATION BY AREA OF RESIDENCE
Family O rien tatio n
Area of Residence
(Date B u ilt Up) Extended____________ Nuclear___________ Non
Before 1940 54.6% 26.4 19.0 100.0 (351)
1940-1949 52.1% 26.3 21.6 100.0 (171)
1950-1959 46.7% 30.3 23.0 100.0 (261)
1960 & a fte r 36.6% 36.5 26.9 100.0 (196)
N = 979
df = 6 X2 = 17.80 < .0 1 C = .424 C/C max = .520
o o
o
TABLE 21
FAMILY ORIENTATION BY AREA OF RESIDENCE
Area of Residence
(Date B u ilt Up)
Family O rien ta tio n
Extended Nuclear Non
Before 1950
1950 & a f te r
53.7%
42.3
26.4
33.0
19.9
24.7
df = 2 X2 = 12.66 < . 0 1 C = .361
100.0 (522)
100.0 (457)
N = 979
C/C max = .511
0 0
TABLE 22
Head
RESIDENCE OF KIN
in
neighborhood
BY AGE
Kin
OF HEAD OF
liv e
in SM SA
HOUSEHOLD
ou tsid e
SMSA
Under 30 34.9% 43.4 21.7 100.0 (127)
30-39 34.0% 44.3 21.7 100.0 (232)
40-49 34.4% 43.0 22.6 100.0 (254)
50-59 37 .4% 43.6 19.0 100.0 (177)
60 & over 25.9% 50.3 23.8 100.0 (189)
N = 979
df = 8 X2 = 7.02 > .0 5 C = .084 C/C max = .103
o o
C O
TABLE 23
INTERACTION WITH KIN BY HOUSEHOLD SIZE
Household Size High
In ter action
Medium Low
1 52.8% 16.0 31.2 100.0 (104)
2 52.2% 26.4 21.4 100.0 (217)
3 60.4% 17.5 22.1 100.0 (129)
4 57.0% 25.8 17.2 100.0 (149)
more 51.2% 24.7 24.1 100.0 (167)
N = 766
df = 8 X2 = 12.96 > .0 5 C = .126 C/C max = 154
00
OJ
84
households to report high frequency of in te ra c tio n with kin,
and single-person households more lik e ly than any other to
have low in te ra c tio n , the apparent d iffe re n ce s are not s t a
t i s t i c a l l y s ig n ific a n t.
Theorem 15. The greater the family o rie n ta tio n ,
the greater the r e s id e n tia l s t a b i l i t y .
The hypothesis is not supported by the d ata. Table
24 shows th a t no d iffe re n ce s were found in re s id e n tia l s ta
b i l i t y by family o rie n ta tio n ; respondents c la s s ifie d as ex
tended fam ily-oriented, nuclear fam ily-oriented, and non-
fam ily-oriented were equally sta b le.
Theorem 16. The newer the area of residence,
the closer the residence of kin.
This hypothesis is not supported by the data.
There was no d ifferen ce between older and newer neighbor
hoods in the proportion of respondents with no re la tiv e s in
the m etropolitan area. (See Table 25.) In addition, r e s i
dents of newer neighborhoods (prim arily tr a c ts b u ilt up
since 1959, a more d e ta ile d analysis not shown below r e
veals) were somewhat le s s lik e ly than others to have kin
resident in the neighborhood. This finding is contrary to
th at predicted.
Theorem 17. The younger the head of the household,
the g reater the in te ra c tio n with kin.
The hypothesis is not supported by the data. While
some su b sta n tia l d iffe ren ces in c e ll values appear in Table
Mobile
Stable
TABLE 24
RESIDENTIAL STABILITY BY FAMILY ORIENTATION
Family O rien tatio n
Extended Nuclear___________ Non
23.0% 20.9% 26.3%
77.0 79.1 73.7
100.0 (474) 100.0 (288) 100.0 (217)
N = 979
df = 2 X2 = 2.02 > .0 5 C = .045 C/C max = .064
00
U l
TABLE 25
RESIDENCE OF KIN BY AREA OF RESIDENCE
Kin liv e
Area of Residence
(Date B uilt Up)
Before 1950 1950 and a fte r
in neighborhood
in SM SA
outside SM SA
35.8%
42.0
22.2
30 .3%
48.3
21 .4
100.0 (522) 100.0 (457) N = 979
df = 2 X = 4.55 > .0 5 C = .022 C/C max = .031
00
O '
87
26, they are not in the d ir e c tio n p re d ic te d . The p rin c ip a l
v a r ia tio n which appears suggests a po ssib le c u rv ilin e a r r e
la tio n s h ip , with the 40-49 age group having le s s frequent
3
in te r a c tio n than any other.
Theorem 18. The la rg e r the household, the
g re a te r the r e s id e n tia l s t a b i l i t y .
When a l l households were analyzed to g e th e r, the
p red icted r e la tio n s h ip between household size and s t a b i l i t y
appeared only in the s ig n i f ic a n t ly greater m obility of
2
When age ca te g o rie s are trichotom ized (under 40,
40-49, and 50 and over) in order to bring out th is curvi-
l i n e a r i t y , the obtained of 10.04 is s ig n ific a n t at the
.01 le v e l. (Table not shown.) I t should be noted, th a t
households in the 40-49 age category tend to be r e l a t i v e l y
la rg e in siz e (though not as la rg e as those in the 30-39
category; see Tables 3, 4 and 5, pages 58-60 above). I t
may be speculated th a t the household focused on the rearin g
and launching of adolescent ch ild re n is le s s a v a ila b le for
in te ra c tio n with kin than at e a r l i e r or l a t e r sta g es. Such
a finding would be in keeping with th a t of Robins and To-
manec concerning an inverse r e la tio n s h ip between size of
nuclear fam ily and closeness to k in . Lee N. Robins and
Miroda Tomanec, ’'Closeness to Blood R elatives outside the
Immediate Fam ily,” Marriage and Family L iving, XXIV (Novem
b er, 1962), 340-346.
Other re la te d fin d in g s are those of Rohrer and
Schmidt in th e ir study of p a r tic ip a t io n in formal organiza
tio n s as r e la te d to fam ily l i f e cy cle. Married couples in
the p re -c h ild stage reported low a c tiv i ty in formal org an i
z a tio n s and a low number of in tra fa m ily a c t i v i t i e s ; those
with a preschool child also rep o rted low formal p a r t i c i p a
tio n but high in tra fa m ily a c t i v i t i e s . Couples whose young
e s t ch ild was aged six to eleven were activ e both in formal
o rg an izatio n s and w ithin the fam ily; while those whose
youngest child was a teenager reported high formal-
o rg an izatio n a c t i v i t y and low in tra fam ily a c tiv i ty . John
Frank Schmidt and Wayne C. Rohrer, M The R elationship of
Family Type to Social P a r t i c i p a t i o n , ” Marriage and Family
L iv in g , XVIII (August, 1956), 224-230.
TABLE 26
INTERACTION WITH KIN BY AGE OF HEAD OF HOUSEHOLD
In te ra c tio n with Kin
Age of Head_________High_________Medium___________ Low
under 30 57.8% 26.5 15.7 100.0 (102)
30-39 59.1% 22.1 18.8 100.0 (179)
40-49 46.8% 24.1 29.1 100.0 (197)
50-59 53.8% 26.2 20.0 100.0 (143)
60 & over 56.5% 18.3 25.2 100.0 (145)
N = 766
df = 8 X2 = 13.90 > .0 5 C = . 122 C/C max .150
o o
00
89
one-person households when compared to a l l others (See Ta
ble 27.) This d iffe re n c e i s p r in c ip a lly an a r t i f a c t of
fam ily type, since in the great m ajority of a l l except one-
person households a married couple is p resen t.
A sep arate an a ly sis lim ite d to households with a mar
ried couple p resen t (Table 28), in order to elim in ate the
e f fe c t of the c o n s is te n tly more mobile incom plete-fam ily
households, yielded apparent d iffe re n c e s in s t a b i l i t y , in
the d ir e c tio n p red ic ted by the hypothesis, between two- and
three-person and four- and fiv e-p erso n households. How
ever, chi-square did not reach the .05 le v el of s i g n i f i
cance. Because c o n tro llin g fo r family type elim in a tes the
d iffe re n c e s in r e s i d e n t i a l s t a b i l i t y by household s iz e , and
because v isu a l in sp ectio n makes i t clear th a t the only s ig
n if ic a n t d iffe re n c e in Table 27 is th a t between one-person
households and a l l o th e rs, the hypothesis cannot be regard
ed as supported.
Theorem 19. The newer the area of resid en ce, the
g reater the in te ra c tio n with k in .
This hypothesis is not supported by the d a ta . Ta
ble 29 shows no d iffe re n c e in frequency of in te ra c tio n with
kin, for those respondents with kin liv in g in the m etropol
ita n area, by area of residence of the respondent.
Theorem 20. The younger the head of the household,
the g re a te r the r e s id e n tia l s t a b i l i t y .
In keeping with the fin d in g s of other s tu d ie s , th is
TABLE 27
Household Size
STABILITY BY HOUSEHOLD SIZE
Mobile Stable
1 34.0% 66.0
2 22.3% 77.7
3 2 1.0% 79.0
4 19.4% 80.6
more 19.4% 80.6
100.0 (153)
100.0 (278)
100.0 (167)
100.0 (175)
100.0 (206)
N = 979
df = 4 X 2 = 13.60 < .0 1 C - .118 C/C max = .167
TABLE 28
RESIDENTIAL STABILITY BY HOUSEHOLD SIZE
(MARRIED COUPLES ONLY)
Household Size
Mobile Stable
2 2 2.8% 77.2 100.0 (218)
3 2 1.3% 78.7 100.0 (141)
4 16.7% 83.3 100.0 (167)
5 15.4% 84.6 100.0 (110)
more 23.5% 76.5 100.0 (81)
N = 717
df = 4 X2 = 4.38 > .0 5 C = .077 C/C max = .109
In te ra c tio n
TABLE 29
INTERACTION WITH KIN BY AREA OF RESIDENCE
Area of Residence
(Date B u ilt Up)
Before 1950_____ 1950 and a f te r
High 55.7% 53.0%
Medium 22.4% 24.2
Low 21.9 22.8
100.0 (402) 100.0 (364) N = 766
df = 2 X2 = 0.59 > . 0 5 C = .032 C/C max = .045
v O
co
93
hypothesis is stro n g ly re je c te d by th e d a ta . R e sid e n tia l
s t a b i l i t y increased c o n s is te n tly by age category. This r e
la tio n s h ip was found fo r a l l households and, when a co n tro l
for fam ily type was introduced, both f o r households which
included a married couple and fo r those which did n o t.
(See Table 30 fo r a l l d a t a .)
In m arried couple households, the percentage of
sta b le fa m ilie s w ith the head aged 30-39 was o n e-th ird
g rea te r than th a t fo r fa m ilie s with the head under 30. Put
another way, m o b ility in the under-30 category was double
th a t for households 30 and over. S t a b i l i t y continued to
in c re ase g rad u ally in successive age c a te g o rie s . House
holds without a m arried couple p re se n t were c o n s is te n tly
le s s s ta b le , e s p e c ia lly , by comparison, in the middle
y ears. In the 30 to 60 age range married couples were only
h a lf as mobile as other types of households; in households
with the head under 30 or over 60, th e re was l e s s d is c r e p
ancy between household types.
Theorem 21. The newer the a re a of resid e n ce , the
g re a te r the r e s i d e n t i a l s t a b i l i t y .
This hypothesis is supported by the d a ta . M obility
r a te s d eclin e c o n s is te n tly as newness of area of resid e n ce
in c re a se s. Tables 31 and 32, the l a t t e r w ith area of r e s i
dence dichotom ized, show th is r e la tio n s h i p c l e a r ly .
TABLE 30
RESIDENTIAL STABILITY BY AGE OF HEAD OF HOUSEHOLD
Family Type
Age of head All Married
of household households couple Other All Couple Other
Under 30 50 .4 52.2 43.2 (127) df 4 4 4
30-39 74.6 79.1 53.6 (232) X2 66.57 52.77 25.69
40-49 81.1 84.2 68.0 (254) < .001 < .0 0 1 C.OOl
50-59 83.0 86.3 73.9 (177) C .253 .263 .298
60 & over 86.2 88.1 84.1 (189) C/C max .358 .372 .421
(979) (717) (262)
v O
• P *
TABLE 31
RESIDENTIAL STABILITY B Y AREA OF RESIDENCE
Area of Residence
(Date B uilt Up) Mobile S table
Before 1950 27.7% 72.3 100.0 (351)
1940-1949 24.6% 75.4 100.0 (171)
1950-1959 2 1.9% 78.1 100.0 (261)
1960 and a fte r 15.3% 84.7 100.0 (196)
N = 979
df = 3 X2 = 11.19 ^ .02 C = .108 C/C max
TABLE 32
RESIDENTIAL STABILITY BY AREA OF RESIDENCE
Area of Residence
(Date B u ilt Up) Mobile Stable
Before 1950 26.6% 73.4 100.0 (522)
1950 and a f te r 19.0% 81.0 100.0 (457)
N = 979
df = 1 X2 = 7.49 -C .01 C = .089 C/C max = .126
vO
96
Summary
This chapter has presented em pirical fin d in g s for
each of the p o s tu la te s and theorems of the proposed theory
of urban fam ilism . In the follow ing chapter these fin d in g s
w ill be summarized and discussed in terms of the ex ten t to
which the theory as a whole is supported. O perational
measures of fam ilism w ill be rev ise d as in d ic a te d , co n tro l
v a r ia b le s w ill be introduced into the a n a ly s is in an e f f o r t
to c l a r i f y r e la tio n s h i p s , and a r e v is io n of the theory w ill
be undertaken.
CHAPTER V
DISCUSSION OF FINDINGS
The preceding chapters have stated the problem of
the study as the re la tio n s h ip of urban familism to re sid e n
t i a l m obility and suburbanization; developed an axiomatic
theory p red ictin g re la tio n s h ip s between v aria b le s (measures
of extended and nuclear familism, an tic ip a te d s tr u c tu ra l
c o rre la te s of fam ilism , and r e s id e n tia l lo c a tio n and mobil
i t y ) ; and presented the analysis of d a ta from an em pirical
t e s t of the theory. At t h i s point i t is necessary to take
stock: How does the theory hold up in the lig h t of the
s t a t i s t i c a l analysis of data?
Empirical Support for the Theory
Figure 2 re p re se n ts an em pirical rev isio n of Figure
1 (Chapter I I , page 24), in which the matrix of i n t e r r e l a
tio n sh ip s between v a ria b le s predicted by the theory was o r
ig in a lly presented. I t can be seen th a t, of the six postu
la te s of the theory, three are regarded as adequately sup
ported: P o stu late 1 (The newer the area of residence, the
younger the head of the household), P o stu la te 2 (The younger
97
98
the head of the household, the la rg e r the household), and
P o s tu la te 5 (The closer the residence of kin, the g reater
the in te ra c tio n with k in ).* For the two p o stu la te s involv
ing family o r ie n ta tio n , breakdown of th is v aria b le into ex
tended and nuclear family o rie n ta tio n y ie ld s a finding of
p a r t i a l support: P o stu la te 3 req u ires restatem ent as ’’The
la rg e r the household, the g reater the nuclear family o rien
t a t i o n ” ; and P o stu la te 4 as "The g re a te r the extended fam
ily o rie n ta tio n , the closer the residence of k in ." The
f i r s t em pirical gap in the theory thus appears between
these two p o s tu la te s ; as r e s ta te d , P o stu la te 4 should no
longer follow P o stu la te 3. P o stu la te 6 (The greater the
in te ra c tio n with kin, the greater the r e s id e n tia l s t a b i l
ity ) received no em pirical support.
The f if t e e n theorems were for the most p art sup
ported or not supported c o n s iste n t with the fin d in g s for
the p o stu la te s and theorems from which they were lo g ic a lly
derived. The exceptions which occurred were in the asso ci
atio n s between (a) in te ra c tio n with kin, family o rie n ta tio n
and r e s id e n tia l s t a b i l i t y , and (b) age of head of household
and r e s id e n tia l s t a b i l i t y . They w ill be discussed below.
*For the f i r s t two p o s tu la te s , however, modifica
tio n of the hypothesis to elim inate households with the
head under 30 has been necessary. I t was an ticip ate d that
such m odification would probably be necessary. (See d i s
cussion, Chapter I I , page 26.)
FIGURE 2
THEORY OF URBAN FAMILISM: MATRIX OF INTERRELATIONS*
EMPIRICAL DATA
A r e a o f
R esid en ce
Age o f Head
o f H o u s e h o l d
H o u s e h o l d
S i z e
E x t e n d -
F a m il y ed
O r i e n t a
t i o n N u c l e
a r
R e s i d e n c e
o f K in
I n t e r a c t i o n
w i t h Kin
R e s i d e n c e I n t e r a c t i o n R e s i d e n t i a l
w i t h K in S t a b i l i t y
H o u s e h o l d F a m i l y O r i e n
S i z e .E x te n d . N u c l e a r . o f K in
Age o f
Head
High Low High Low High Low Large Small lose Far
New
Old
Young
Old
Large
S m a ll
Low
High
Low
C l o s e
F a r
H ig h
Low
♦ R e f e r t o F i g u r e 1 , p a g e 2 4 . I n F i g u r e 1 , a r r o w s i n d i c a t e d r e l a t i o n s h i p s i m p l i e d b y t h e
t h e o r y ; a l l p o i n t e d f ro m t h e u p p e r l e f t t o t h e l o w e r r i g h t - h a n d c o r n e r o f t h e f i g u r e . I n
t h e a b o v e f i g u r e , a s b e f o r e , d o u b l e a r r o w s s p e c i f y p o s t u l a t e s , s i n g l e a r r o w s s p e c i f y t h e
o r e m s . A s o l i d a r r o w i n d i c a t e s e m p i r i c a l s u p p o r t f o r t h e h y p o t h e s i s ; a b r o k e n * ' * a r
row i n d i c a t e s a n o t - q u i t e - s i g n i f i c a n t e m p i r i c a l a s s o c i a t i o n i n t h e p r e d i c t e d d i r e c t i o n .
A a r r o w i n d i c a t e s a s i g n i f i c a n t a s s o c i a t i o n i n t h e d i r e c t i o n o p p o s i t e t o t h a t p r e d i c t
e d . A h o r i z o n t a l 4—> a r r o w i n d i c a t e s no s i g n i f i c a n t a s s o c i a t i o n b e t w e e n t h e v a r i a b l e s .
The e m p i r i c a l f i n d i n g s h a v e made i t n e c e s s a r y t o s u b d i v i d e f a m i l y o r i e n t a t i o n i n t o e x
t e n d e d and n u c l e a r c o m p o n e n ts .
Association between V ariables
100
Measures of familism
1. In te ra c tio n with kin, the behavioral measure of f a
milism adopted in th is research, was em pirically associated
only with residence of kin (P o stu late 5). With a l l other
variables of the theory no association appeared. The only
possible exception to th is statement occurred with regard
to age of head of household (Theorem 17; discussed in Chap
ter IV, page 87), where a s ig n ific a n t cu rv ilin e ar associa
tio n , in which in te ra c tio n was le s s frequent for households
with the head aged 40 - 49 than for those younger or older,
was found. This asso ciatio n was not, however, th a t p re
dicted by the theory.
These findings cast some doubt on e ith e r the e ffe c
tiveness of in te ra c tio n with kin as a measure of familism,
or the value of familism as a p red icto r of other v ariab les
in the theory. The only v ariab le with which in te ra c tio n
was associated as predicted was residence of kin, of which
i t re p re se n ts, in a sense, only a fu rth e r extension. Inter
action was measured only for those respondents with kin in
the m etropolitan area; and, as might be expected, i t was
strongly associated with closeness of residence w ithin the
area. Residence of kin was introduced into the theory as
an an ticip ated s tr u c tu ra l c o rre la te of familism. In the
analysis thus f a r , residence of kin has been shown not only
101
to be c o rre la te d w ith the behavioral measure of fam ilism ,
but also to be p r e d ic tiv e of other v a ria b le s where i n t e r a c
tio n i s not. The question of the n e c e s s ity fo r introducing
a measure of fam ilism may thus be a p p ro p ria te ly r a is e d ,
though not answered, at t h i s p o in t.
Before in te r a c tio n w ith kin is elim in ated as a
f r u i t f u l v a ria b le in b u ild in g the theory, fu rth e r an aly sis
is d e s ir a b le . This w ill take two forms: (a) re c o n s id e ra
tio n of the o p e ra tio n a l measure used CSee pages 109-114) and
(b) in tro d u c tio n of c o n tro l v a ria b le s (See pages 114-115).
2. Family o r ie n ta tio n , the a t t i t u d i n a l measure of f a
milism, was found to a s u b s ta n tia l e x te n t in s ig n i f ic a n t as
so c ia tio n with other v a ria b le s of the theory. However, an
a ly s is of d a ta involving a s so c ia tio n s of th is measure (See
P o s tu la te 3 and Theorems 8 and 12) in d icated the in a p p ro p ri
ateness of one of the im p lic it hypotheses underlying the op
e r a tio n a l d e f i n i t i o n of v a r ia b le s , namely, the assumption
th a t extended fam ily o r ie n ta tio n , nuclear fam ily o r ie n ta
tio n , and non-fam ily o r ie n ta tio n can be regarded as p o in ts
on a continuum. This assumption was based p rim a rily on
Litwak’ s rep o rt of the s c a l a b i l i t y of a comparable opera-
tio n a l measure of fam ilism ; i t was n o t, however, borne out
by the d a ta of the p re se n t rese arch .
2
Eugene Litwak, "Occupational M obility and Extended
Family Cohesion," American S o cio lo g ical Review, XXV (Febru
ary, 1960), 17.
102
As here measured, extended and nuclear family o r i
e n ta tio n tended to occur in inverse r e la tio n to each other
with resp e c t to the other v a ria b le s . Thus extended fam ily
o r ie n ta tio n was found, contrary to p re d ic tio n , in negative
a s so c ia tio n with siz e of household, age of head of house
hold, and area of resid en ce. (The g re a te r the extended
family o r ie n ta tio n , the smaller the household, the older
the head of the household, and the older the area of r e s i
dence.) In c o n tra s t, nuclear fam ily o rie n ta tio n occurred
in the p red icted p o s itiv e a s so c ia tio n with the same th ree
v a ria b le s (except for area of resid en ce, where the apparent
p o s itiv e asso cia tio n was not q u ite s ig n i f ic a n t ) . Non
family o rie n ta tio n , the re s id u a l category, was s i g n i f i c a n t
ly asso ciated with no other v a ria b le s and is apparently
not a meaningful c l a s s i f i c a t i o n .
Extended fam ily o rie n ta tio n was p re d ic tiv e of r e s
idence of kin in the neighborhood, though i t did not d i f
f e r e n t i a t e those with r e la tiv e s in the m etropolitan area
from those w ithout. Nuclear family o r ie n ta tio n , however,
was not associated with residence of kin.
Figure 3 p re se n ts those r e la tio n s h ip s in graphic
form. In summary, family o r ie n ta tio n was associated in
some way with the th ree a n tic ip a te d s tr u c t u r a l c o r re la te s
of familism: age of head of household, siz e of household,
and residence of kin. Contrary to p re d ic tio n , however,
extended family o rie n ta tio n was n eg ativ ely asso ciated with
103
the f i r s t two of these v a ria b le s , and only nuclear fam ily
o r ie n ta tio n was found in the p red icted a s s o c ia tio n . With
residence of kin, extended fam ily o rie n ta tio n was a s so c i
ated as p re d ic te d , but no a s so c ia tio n fo r nuclear fam ily o r
ie n ta tio n occurred.
FIGURE 3
ASSOCIATION OF CATEGORIES OF FAMILY ORIENTATION
W ITH SELECTED VARIABLES
Family O rien ta tio n
Extended Nuclear Non
Age of head of household -
+
0
Size of household -
+
0
Residence of kin
+
0 0
Area of residence -
+*
0
+ S ig n ific a n t a s so c ia tio n in p red icted d ire c tio n
- S ig n ific a n t a s so c ia tio n in opposite d ire c tio n
0 No s ig n if ic a n t a sso c ia tio n
* A ssociation not q u ite s ig n ific a n t
2
Level of s ig n ific a n c e : .05 (X )
Family o rie n ta tio n was not associated with one of
the c r ite r i o n v a ria b le s , r e s i d e n t i a l s t a b i l i t y . With the
o th e r, area of resid en ce, nuclear fam ily o rie n ta tio n ap
peared in the pred icted r e la tio n s h ip but the a s so c ia tio n
was not q u ite s t a t i s t i c a l l y s ig n if ic a n t; extended fam ily
o r ie n ta tio n , on the other hand, was n eg a tiv e ly associated
with area of resid en ce.
104
Further an aly sis of th is v aria b le w ill be undertaken
in order to c l a r i f y i t s proper place in the theory. The
f i r s t stage of the analysis w ill include an e x p lic it review
of the data re la te d to the im p lic it hypothesis th a t extend
ed, nuclear, and non-family o rie n ta tio n are po in ts on a s in
gle continuum. (Since the data have ca st doubt on th is as
sumption, the operational measurement of family o rie n ta tio n
w ill be examined in some d e t a i l ; see pages 115-119). In
the second stage control v ariab les w ill be introduced to
c l a r if y the nature of selected em pirical asso ciatio n s (see
pages 119-120).
3. The two measures of familism, in te ra c tio n with kin
and family o rie n ta tio n , were not associated with each other.
This asso ciatio n was one of the o rig in a l assumptions made;
i t was incorporated in the hypotheses of the theory. Taken
as a whole, the em pirical findings concerning familism, as
analyzed thus fa r, do not hold great promise for the pro
posed theory of urban familism.
S tru c tu ra l c o rre la te s of familism
As noted above, the an tic ip a te d s tr u c tu ra l co rre
la te s of familism were not em p irically co rrela ted with the
behavioral measure of familism, nor was th e ir asso cia tio n
with the a t t i t u d i n a l measure, family o rie n ta tio n , a c le a r-
cut one. Family o rie n ta tio n was s ig n if ic a n tly associated
in some fashion, though not c o n s iste n tly in the predicted
105
d ire c tio n , with each of the s tr u c tu ra l c h a r a c te ris tic s ; con
sequently some te n ta tiv e support for the use of the o rig in a l
lab el remains. The re la tio n sh ip s between the family s tr u c
tu ra l v a ria b le s and the c r ite r io n v a ria b le s, and of these
s tr u c tu ra l v aria b les among themselves, may be summarized as
follows:
1. Age of head of household was found, except for
households with the head under 30, in the predicted r e l a
tionship with area of residence. I t was also found in s ig
n ific a n t association with r e s id e n tia l s t a b i l i t y , but in the
opposite d ire c tio n to that predicted by the theory. This
association is comparable to th a t found in other stu d ies.
However, since the re la te d asso ciatio n s predicted by the
theory did occur, control v ariab les w ill be introduced into
the analysis in an attempt to c la r if y the re la tio n sh ip be
tween age and s t a b i l i t y . (See pages 130-135.)
2. Household siz e , lik e age, was found in the p r e d ic t
ed re la tio n sh ip with area of residence. An apparent asso
c iatio n as predicted between size and r e s id e n tia l s t a b i l i t y
was also found; however, th is was seen as a function of
family type (the s ig n ific a n t d ifferen ce was between one-
person households and a l l others) rather than of size per
s e .
3. Residence of kin was not associated with area of
residence. I t was, however, p red ic tiv e of r e s id e n tia l s t a
b i l i t y : s t a b i l i t y was higher for those with kin in the
m etropolitan area than for those w ithout.
4. Age of head of household was associated as p r e d ic t
ed, except where the head was under 30, with household siz e ;
i t was not associated with residence of kin. Some lim ited
asso cia tio n between size and residence of kin occurred, but
it appeared to be p rim arily a function of fam ily type and
not adequate to support the hypothesis. (See d iscu ssio n ,
Chapter IV, page 72.)
C rite rio n v aria b les
The e ffic ie n c y of measures of familism and s tr u c
tu ra l measures in p red ictin g the c r ite r io n v a ria b le s has
been discussed above. Here i t should be said only th a t
these two v a ria b le s, area of residence and r e s id e n tia l s t a
b i l i t y , appeared in p o s itiv e a sso cia tio n with each o th e r:
The newer the area of residence, the g reater the re s id e n
t i a l s t a b i l i t y .
Summary: Empirical Test of the Hypotheses
In summarizing th is sectio n , the o rig in a l po stu
la te s and theorems are re s ta te d in the l ig h t of the analy-
3
s is of em pirical d ata made thus f a r . Those in which no
^For the p o stu la te s and theorems as o r ig in a lly
sta te d , see Chapter I I , pages 21-23. A v isu al comparison
of the o rig in a l p o stu la te s and theorems with the p o stu la te s
and theorems re s ta te d to conform with the em pirical data
may be made by comparing Figure 1, page 24, with Figure 2,
page 99.
change has been necessary are s ta rre d (* ).
107
P o stu la te s
1. The newer the area of resid en ce, the younger the
head of the household (except when the head is under 30).
2 . The younger the head of the household, the la rg e r
the household (except when the head i s under 30).
3. The la rg e r the household, the g re a te r the nuclear
fam ily o rie n ta tio n and the le s s the extended fam ily orien
t a t ion.
4. The greater the extended family o r ie n ta tio n , the
clo ser the residence of k in .^
*5. The clo ser the residence of k in , the g re ater the
in te ra c tio n with kin.
6 . There is no a s so c ia tio n between in te ra c tio n with
kin and re s id e n tia l s t a b i l i t y .
Theorems
*7. The newer the area of resid en ce, the la rg e r the
household.
8. The younger the head of the household, the g reater
the nuclear family o rie n ta tio n and the le s s the extended
fam ily o rie n ta tio n .
9. There is no as so c ia tio n between household siz e and
residence of kin, except th a t single-person households are
le s s lik e ly to have r e la tiv e s p resen t.
10. There is no a s so c ia tio n between family o rie n ta tio n
and in te ra c tio n w ith kin.
*11. The closer the residence of k in , the g reater the
r e s id e n tia l s t a b i l i t y .
12. The newer the area of resid e n ce, the g rea te r the
4
Revision of the measure of familism elim in ates
th is re la tio n s h ip . See d iscu ssio n , pages 115-119 below.
108
n u clear fam ily o r ie n ta tio n ( ? ) “* and the l e s s the extended
fam ily o r ie n ta tio n .
13. There is no a s so c ia tio n between age of head of
household and resid en ce of k in .
14. There i s no a s so c ia tio n between household s iz e and
in te r a c tio n with kin.
15. There is no a s so c ia tio n between fam ily o r ie n ta tio n
and r e s i d e n t i a l s t a b i l i t y .
16. There i s no a s s o c ia tio n between area of residence
and residence of kin.
17. The a s s o c ia tio n between age of head of household
and in te r a c tio n with kin is c u r v ilin e a r .
18. There i s no a s s o c ia tio n between household s iz e and
r e s i d e n t i a l s t a b i l i t y , except th a t sin g le -p e rso n households
are le s s s ta b le .
19. There is no a s s o c ia tio n between area of residence
and in te ra c tio n with kin.
20. The younger the head of the household, the le s s
the r e s i d e n t i a l s t a b i l i t y .
*21. The newer the area of resid e n ce , the g re ate r the
r e s i d e n t i a l s t a b i l i t y .
R estated to conform with em pirical evidence, the
hypotheses have lo s t the s tr u c tu r e of an axiomatic theory.
Following c o n sid e ra tio n of questions of measurement of f a
milism and of c o n tro llin g fo r p o ssib le confounding v a r i
a b le s, an attem pt w ill be made to re v ise the o r ig in a l hy
potheses to achieve a more accurate axiomatic theory.
^A ssociation not q u ite s ig n i f ic a n t .
109
Measurement of Familism
I n te r a c tio n w ith Kin
Refinement of o p era tio n a l measure
Of the se v e ra l measures of in te r a c ti o n with kin
p o ssib le with the d a ta a v a ila b le , the sim p lest was chosen.
F a ce -to -fac e i n te r a c ti o n with any r e l a t i v e , w ithout reg ard
to clo sen ess of r e la tio n s h ip or to the number of r e l a t i v e s
with whom co n ta c t was m aintained, was categ o rized fo r f r e
quency. If co n tact w ith any r e l a t i v e was re p o rted as ta k
ing place weekly or more o fte n , in te r a c ti o n was categ o rized
as high; medium in te ra c tio n took p la c e , w ith any r e l a t i v e ,
le s s than once a week but at l e a s t once a month; low i n t e r
a c tio n occurred l e s s than once a month.
Because th is crude measure has f a ile d to p re d ic t
the other v a ria b le s of the th e o ry , an e f f o r t to r e f in e i t
is d e s ira b le . At the p resen t time, c o n s tru c tio n of an i n
dex based on the number of r e l a t i v e s with whom in te r a c tio n
occurs is not f e a s ib le , since the coding of d a ta has not
progressed th is f a r . There i s some reason to b e lie v e , how
ever, th a t closeness of r e la tio n s h ip is a more re le v a n t as
pect of in te ra c tio n w ith kin than is the number of d i f f e r
ent r e l a t i v e s involved. R eiss, for example, has commented
th a t two f a c to r s account for n early a l l d iffe re n c e s in kin
in te r a c ti o n : clo sen e ss of resid en ce and closeness of
' 110
re la tio n s h ip . The former has already been found s i g n i f i
cantly associated with in te ra c tio n in the present research;
the l a t t e r w ill be investigated here.
Table 33 shows the v a ria tio n in frequency of i n t e r
action by closeness of re la tio n s h ip . As before, frequency
is coded on the b asis of the g re a te st in te ra c tio n reported
with any r e la tiv e in each category. In other words, a r e
spondent whose family maintains weekly contact with a p ar
ent of e ith e r spouse w ill be categorized as high in i n t e r
actio n , even though v is itin g with another parent occurs
le ss often.
Percentages in th is ta b le , lik e a l l rep o rts on in
te ra c tio n with kin, are calculated only fo r households with
kin present in the m etropolitan area, in each category of
re la tio n s h ip . I t may be noted th a t 76.4 per cent of re
spondents reported no children present outside the house
hold; 67.7 per cent, no parents present; 47.7 per cent, no
sib lin g s present; and 64.8 per cent, no other r e la tiv e s
p resent. Only one-fourth to one-half of a l l sample house
holds have kin present in any given re la tio n s h ip category,
in other words; but, as previously reported, 766 (78.3%)
have kin present in some category.
^Paul J . R eiss, "The Extended Kinship System: Cor
r e la te s of and A ttitu d es on Frequency of In te ra c tio n ," Mar
riage and Family L iving, XXIV (November, 1962), 335-336.
TABLE 33
INTERACTION WITH KIN B Y CLOSENESS OF RELATIONSHIP
In te r action
Medium Low
Children 78.7% 1 2 .2 9.1 100.0
R elationship
(to head of Parents 60 .2% 27.8 1 2 .0 100.0
household
or spouse) S iblings 35.6% 27.6 36.8 100.0
Other 19.7% 23. 8 56.5 100.0
(230)
(316)
(511)
(344)
df = 6 X2 = 239.20 < .0 0 1 C = .387 C/C max = .474
112
In te ra c tio n with kin c le a rly v arie s with closeness
of re la tio n s h ip . Reported in te ra c tio n occurs most often
between grown children and th e ir p aren ts, when the l a t t e r
are respondents, and somewhat le ss often when the former
are respondents. Nearly f o u r - f i f th s of parents with grown
children present in the area, compared with th r e e - f if th s of
grown children with parents p rese n t, rep o rt weekly v is itin g
with them. This d iffe re n ce suggests the likelihood that
some parents spread th e ir frequent v is itin g among several
children p resen t in the area. Frequent in te ra c tio n with
sib lin g s and, e s p e c ia lly , other kin is le s s common; nearly
tw o-thirds of respondent households with sib lin g s present
do, however, v i s i t with one or more sib lin g s at le a s t once
a month.
Table 34 p rese n ts r e s id e n tia l m obility ra te s by in
te ra c tio n with kin, broken down by closeness of r e la tio n
ship. Refinement of the measure by addition of closeness
of re la tio n s h ip does not change the o rig in a l findings. For
respondents with kin present in the m etropolitan area, f r e
quency of in te ra c tio n with kin is not p red ictiv e of r e s i
d en tial s t a b i l i t y .
I t should be noted th a t the associated v aria b le,
residence of kin, was p re d ic tiv e of d iffe re n ce s in s t a b i l
ity between households with kin present in the m etropolitan
area and those without kin p resen t. Introduction of r e l a
tionship categ o ries fo r residence of kin does not change
TABLE 34
RESIDENTIAL MOBILITY RATES BY INTERACTION WITH KIN (BY RELATIONSHIP)
In te ra c tio n
High Medium Low df x2 C C/C max
Children 19.9 14.3* 10.5* (230) 2 3.23 > .05 .114 .161
R elationship
( to head of P aren ts 21.6 18.2 25.6 (316) 2 0.98 >.05 .055 .078
household
or spouse) Sibling s 23.1 18.5 18.5 (511) 2 1.48 >.05 .055 .078
Other 22.1 23.2 20.0 (344) 2 0.38 >.05 .032 .045
*N for c e ll i s le s s than 10
113
114
the o rig in al fin d in g s. (Table not shown.) Regardless of
which r e la tiv e s were present in the m etropolitan area,
households with r e la tiv e s present were more stable than
those with no r e la tiv e s present.
Because n e ith e r in te ra c tio n with kin nor residence
of kin was found in asso ciatio n with area of residence, the
other c r ite rio n v aria b le, i t has not seemed necessary to
introduce consideration of closeness of re la tio n sh ip into
th is asso ciatio n .
Introduction of control v ariab les
In a f in a l e f f o r t to discover associations between
in te ra c tio n with kin and other v ariab les of the theory,
controls were introduced into measurement of the associa
tions between in te ra c tio n , the two c r ite r io n v aria b les and
the a ttitu d in a l measure of familism, with the following
negative r e s u lt s :
1) When residence of kin (in or outside neighborhood,
w ithin SMSA) i s co n tro lled , there is no asso ciatio n between
in te ra c tio n with kin and r e s id e n tia l s t a b i l i t y .
2) When size of household is co n tro lled , there is no
asso ciatio n between in te ra c tio n with kin and r e s id e n tia l
s t a b i l i t y .
3) When age of head of household is co n tro lled , there
i s no association between in te ra c tio n with kin and re sid e n
t i a l s t a b i l i t y .
115
4, 5, 6) When the same three v a ria b le s are co n tro lle d ,
there is no a s so c ia tio n between in te ra c tio n with kin and
area of resid en ce.
7) When residence of kin is c o n tro lle d , there i s no
a s so c ia tio n between in te ra c tio n w ith kin and extended fam ily
o r ie n ta tio n .
Family O rientation
Revision of operational measure
7
The o rig in a l assumption made was that i t would be
p o ssib le to d is tin g u is h three p o in ts on a fam ily o r ie n ta
tio n continuum--extended family o r ie n ta tio n , nuclear fam ily
o r ie n ta tio n , and non-family o rie n ta tio n --a n d th a t using mu
tu a lly exclusive measures would be ap p ro p riate . This a s
sumption was not borne out by the d a ta a n a ly s is . Non
family o r ie n ta tio n occurred in s ig n if ic a n t a sso cia tio n with
none of the v a ria b le s and can th e re fo re be reje cte d as a
u seful category. Where extended and nuclear family o r ie n
ta tio n both occurred in a s ig n ific a n t a s s o c ia tio n , they
were c o n s is te n tly r e la te d in opposite d ire c tio n s to the
v a ria b le in q uestion. These fin d in g s in v a lid a te the a s
sumption of a continuum and ra is e the question of the pos
s ib le r e la tio n s h ip between extended and nuclear fam ily
7
On the b a s is of Litw ak's previous fin d in g s. L it-
wak, op. c i t .
116
o rie n ta tio n . In v estig atio n of th is r e la tio n s h ip re q u ire s
rev isio n of the measures used, since the use of mutually
exclusive measures has appeared to be a r b itra ry and perhaps
• * 8
inappropri a t e .
Table 35 shows th e re la tio n s h ip between extended
family o rie n ta tio n and nuclear family o rie n ta tio n when the
same items as before are used to measure each but when
these are not regarded as mutually exclu siv e. In the o rig
inal a n a ly sis, a respondent who expressed agreement with
the measure of extended family o rie n ta tio n was c la s s if ie d
as extended fam ily -o rien ted re g a rd le ss of h is response on
the measure of nuclear fam ily o rie n ta tio n . In Table 35 the
measures have been tre a te d independently, thus d is tin g u is h
ing respondents who agreed w ith both the extended and nu
clear family items from those who agreed with e ith e r or
n e i th e r .
As o r ig in a lly assumed, the extended fam ily -o rien ted
Q
In the design of th is research the im p lic it hy
potheses, or underlying assumptions, were not te ste d p rio r
to the te s tin g of the hypotheses of the theory i t s e l f .
Since the assumptions concerning a sso cia tio n s between fa -
m i l i s t i c a t titu d e s and behavior and between familism and
i t s a n tic ip a te d s tr u c tu ra l c o rr e la te s were incorporated i n
to the hypotheses of the theory, th is approach has caused
no d i f f i c u l t y in these in stan ces. The assumption concern
ing family o r ie n ta tio n , however, was incorporated into the
o perational d e f in itio n of th a t variable and not into the
hypotheses. Since the in v a lid a tio n of the assumption w ill
now req u ire re v isio n of the operational d e f in itio n and r e
te s tin g of the hypotheses in terms of the revised measure,
i t may be suggested th a t te s tin g th is assumption e a r l ie r
would have been more e f f i c i e n t .
TABLE 35
EXTENDED FAMILY ORIENTATION AND NUCLEAR FAMILY ORIENTATION
Extended Family O rien ta tio n
P resen t Absent
Nuclear
Family
O rien tatio n
Present 56.9%
376
79.1%
43.1%
288
56.7%
664
100.0%
Absent 68.6%
98
20.9%
31.4%
217
43.3%
315
100.0%
474
100.0%
505
100.0%
N = 979
df = 1 X2 = 54.64 < .001 C = .223 C/C ma: .315
117
118
tend to be nuclear fa m ily -o rie n te d as w e ll, with 79.1 per
cent of the extended fa m ily -o rie n te d in th is category. Fur
th e r, n early h a lf (43.1%) of the n u clear fa m ily -o rie n te d are
not extended fa m ily -o rie n te d . However, the twenty per cent
of the extended fam ily -o rie n ted who are not a lso nuclear
fam ily -o rien ted re p rese n t s u b s ta n tia l e r ro r in the assump
tio n .
As o r ig in a l ly measured, 474 respondents (both c e l l s
in column 1) were extended fa m ily -o rie n te d , 288 were n u cle
ar fa m ily -o rie n te d , and 217 were n o n -fa m ily -o rien te d . A
p o ssib le a l te r n a ti v e measure, viewing extended and nuclear
fam ily o r ie n ta tio n as overlapping c a te g o rie s , would d e s ig
nate the same 474 respondents as extended fa m ily -o rie n te d ,
but would d esig n ate a l l 664 respondents in row 1 as nuclear
fa m ily -o rie n te d .
I t is p o ssib le th a t the apparent inverse r e l a t i o n
ship with o th er v a ria b le s which has been rep o rted fo r ex
tended and nuclear fam ily o r ie n ta tio n is an a r t i f a c t of the
measures used o r ig in a l ly . By these measures more than h a lf
of the la rg e number of respondents agreeing w ith the n u c le
ar fam ily item were omitted from t h i s category by being
c l a s s i f ie d w ith the extended fa m ily -o rie n te d . A fu rth e r
data an a ly sis s u b s titu tin g the independent measure of nu
c le a r o r ie n ta tio n fo r th a t p re v io u sly used yielded s i g n i f i
cant re la tio n s h ip s w ith none of th e v a r ia b le s . This f in d
ing supports the above suggestion and le a d s to the
119
conclusion th a t nuclear family o rie n ta tio n as measured is
not a useful v a ria b le , and th a t the inverse r e la tio n s h ip s
o rig in a lly found simply re fle c te d the operational d e f i n i
tion which lim ited the nuclear fam ily-oriented to those not
extended fam ily -o rien ted .
R ecalculation of the o rig in a l fin d in g s for family
o rie n ta tio n with th is v aria b le dichotomized as Extended
Family O rie n ta tio n : Present/A bsent ( i . e . , with the former
categ o ries of nuclear and non-family o rie n ta tio n combined)
a lte re d previously reported findings in only one in stan ce.
For P o s tu la te 4, there was no d iffe re n c e in family o r ie n ta
tion by residence of kin. (Previously, greater extended
family o rie n ta tio n was reported for respondents with r e l a
tiv e s in the neighborhood.)9
Introduction of control v a ria b le s
A lim ited number of control v a ria b le s were in tr o
duced in an attempt to discover an a sso cia tio n between ex
tended family o rie n ta tio n and r e s id e n tia l s t a b i l i t y . These
v a ria b les were not found in the predicted asso cia tio n with
each other, yet both occurred in some type of asso cia tio n
with age of head of household, siz e of household, and area
of residence. R esults of th is analysis were as follow s:
9
On the other hand, the asso cia tio n between the
v a ria b les of P o stu la te 3--fam ily o rie n ta tio n and size of
household—appeared more c le a rly in the revised a n a ly sis.
120
1) When age of head of household is c o n tro lle d , there
is no asso cia tio n between extended fam ily o rie n ta tio n and
r e s id e n tia l s t a b i l i t y .
2) When siz e of household is c o n tro lle d , th ere is no
a s so c ia tio n between extended fam ily o rie n ta tio n and r e s i
d e n tia l s t a b i l i t y .
3) When area of resid en ce is c o n tro lled , there is no
a s so c ia tio n between extended fam ily o rie n ta tio n and r e s i
d e n tia l s t a b i l i t y .
Summary: Measurement of Familism
The two measures of familism, in te ra c tio n w ith kin
and family o rie n ta tio n , have been fu rth e r analyzed in an
attem pt to improve t h e i r e ffic ie n c y as p r e d ic to rs . However,
n e ith e r refinem ent of the measure of in te r a c tio n with kin,
in terms of closeness of r e la tio n s h ip , nor the in tro d u c tio n
of se lec ted control v a ria b le s in the a n a ly sis has improved
the p re d ic tiv e value o f th is measure.
The measure of family o rie n ta tio n has been revised
on the b a s is of em pirical fin d in g s, w ith nuclear and non
family o rie n ta tio n omitted and only the presence or absence
of extended family o r ie n ta tio n in d ic a te d . The in tro d u ctio n
of co n tro l v a ria b le s in an attempt to discover an a s so c ia
tio n between family o rie n ta tio n and r e s id e n tia l s t a b i l i t y
had no e f f e c t .
In the continued an a ly sis of the theory which
1 2 1
follow s, in te ra c tio n with kin w ill be elim inated from con
s id e ra tio n . The re v ise d measure of extended fam ily o rie n ta
tio n w ill be c o n s is te n tly used in place of the former more
general fam ily o rie n ta tio n measure.
S tru c tu ra l Measures of Family
C h a ra c te ris tic s
Thus f a r , measures of fam ilism have not dem onstrat
ed g reat promise as p re d ic to rs of the c r i t e r i o n v a ria b le s ,
area of residence and r e s id e n tia l s t a b i l i t y . E ffo rts to
r e fin e the behavioral measure of fam ilism , in te ra c tio n with
k in , have rendered i t no more u se fu l as a p r e d ic tiv e v a r i
able. Revision of the a t t i t u d i n a l measure, fam ily o r ie n ta
tio n , has c l a r i f i e d the v aria b le i t s e l f and maintained i t s
a s s o c ia tio n with area of residence; but no r e la tio n s h ip be
tween family o r ie n ta tio n and r e s id e n tia l s t a b i l i t y could be
discovered even with the in tro d u c tio n of c o n tro ls .
In s h o rt, of the two measures of familism pred icted
to be asso ciated with two c r i t e r i o n v a ria b le s , only one is
a c tu a lly asso ciated with one c r i t e r i o n . Could t h i s one as
so c ia tio n also be elim inated, for the purposes of e f f i c i e n t
p re d ic tio n , by s u b s titu tin g one or more of i t s s tr u c tu r a l
c o r re la te s ? Is th e re , in other words, any need to i n t r o
duce f a m i l i s t i c so cial-p sy c h o lo g ical v a ria b le s in order to
p re d ic t r e s id e n tia l lo c a tio n and m o b ility ? Age of head of
household, siz e of household, and resid e n ce of kin have a l l
proved to be s tr u c tu r a l c o rre la te s of fam ilism , as
1 2 2
o rig in a lly assumed (although the d ire c tio n of the asso cia
tio n , for the f i r s t two s tru c tu ra l v a ria b le s, is not th at
p re d ic te d ). Of these, age and size (though not residence
of kin) are p red ic tiv e of area of residence. Is the ob
served asso ciatio n between extended family o rie n ta tio n and
area of residence merely a function of the association of
extended family o rie n ta tio n with age of head and/or size of
household? If controls for these v ariab les are introduced,
w ill the association between extended family o rien ta tio n
and area of residence disappear?
As Tables 36 and 37 show, the introduction of con
tr o ls does larg ely elim inate th is asso ciatio n . When age is
co n tro lled , re sid e n ts of older neighborhoods s t i l l appear
to be more extended fam ily-oriented, but the differen ces
are not s t a t i s t i c a l l y sig n ific a n t (with the exception of
that in the 30-39 age category). When household size is
co n tro lled , extended family o rie n ta tio n s t i l l appears con
s is te n tl y more c h a ra c te ris tic of older neighborhoods, but
again d iffe re n c e s are not s t a t i s t i c a l l y s ig n ific a n t. Be
cause c e ll values are r e la tiv e ly small, making the a t t a i n
ment of acceptable le v e ls of sig n ifican ce le ss lik e ly , and
because the d ifferen ces which do appear are very co n sisten t
ly in the predicted d ire c tio n , conclusions must be te n ta
tiv e . N evertheless, i t appears that extended family o rie n
tatio n may not be necessary as a p red icto r of area of r e s i
dence; the observed association between these two v ariab les
TABLE 36
Age of Head
of Household
PER CENT EXTENDED FAMILY ORIENTATION BY AREA OF
RESIDENCE A N D AGE OF HEAD OF HOUSEHOLD
Area of Residence
(Date B u ilt Up)
Before 1950 1950 and a fte r df X2 C C/C max
Under 30 47.1% 32.2% 1 2.32 > .0 5 .134 .190
30-39 50.4 36.4 1 4.01 < .05 .130 .184
40-49 49.9 43.4 1 0.83 > .0 5 .055 .078
50-59 48.5 46.2 1 0.03 > .0 5 .000 .000
60 & over 66.9 62.4 1 0.20 > .05 .032 .045
N = 979
123
TABLE 37
PER CENT EXTENDED FAMILY ORIENTATION BY AREA OF
RESIDENCE AND SIZE OF HOUSEHOLD
Household Size
Area of Residence
(Date B u ilt Up)
Before 1950 1950 and a fte r df X2 C C/C max
1 58.8% 51.3% 1 0.40 > .05 .055 .078
2 53.5 47.1 1 0.90 ^ .05 .055 .078
3 61.4 48.1 1 1.96 > .05 .110 .156
4 46.3 31.5 1 3.24 > .05 .134 .190
more 51.7 38.5 1 3.08 > .05 .122 .173
125
i s probably adequately accounted fo r by other v a ria b le s to
which .both are r e l a t e d , namely, age of head of household
and siz e of household. To t h i s e x te n t, the in tro d u c tio n of
a s o c ia l-p s y c h o lo g ic a l measure of fam ilism is not req u ired
to p r e d ic t th is aspect of fam ily behavior. This te n ta t iv e
conclusion f a i l s to support th a t made e a r l i e r by such w r i t
e rs as F ava,10 in regard to the importance of s o c ia l-
psychological measures in accounting for suburban re sid e n c e .
N e ith e r, however, does i t c o n tra d ic t Fava’s conclusion,
which was based on d i f f e r e n t measures.
The present re se a rc h has, th e re fo re , e s ta b lish e d no
c le a r r a tio n a le for introducing f a m i l i s t i c v a ria b le s in the
p r e d ic tio n of e ith e r area of resid en ce or r e s i d e n t i a l s t a
b i l i t y . In p re d ic tin g the l a t t e r only the s t r u c t u r a l v a r i
ab les have proved e f f e c t iv e . In p re d ic tin g area of r e s i
dence the e f fe c tiv e n e s s of extended fam ily o r ie n ta tio n ap
p ea rs to be la rg e ly accounted fo r by i t s a s so c ia tio n with
the s tr u c t u r a l v a r ia b le s , age of head of household and siz e
of household.
10Fava has p resen ted d a ta showing th a t neighboring
d i f f e r s by area of resid e n ce even when c o n tro ls for seven
other v a ria b le s are introduced. S ylvia F le is Fava, ”Con-
t r a s t s in N eighboring: New York City and a Suburban Coun
t y , ” The Suburban Community, ed. William M. Dobriner (New
York: G. P . Putnam’ s Sons, 1958), pp. 122-131.
126
Revision of the Theory
Familism and Area of Residence
A re v is io n of the theory of urban familism, based
on the em pirical fin d in g s, w ill n e c e ss a rily focus on the
asso c ia tio n between extended family o rie n ta tio n and area of
residence, since th is was the only s ig n ific a n t asso cia tio n
found between a measure of familism and a c r ite r io n v a r i
able. This rev isio n w ill be undertaken in s p ite of the
fin d in g , ju s t discussed, th a t the a sso cia tio n between ex
tended fam ily o rie n ta tio n and area of residence may be
la rg e ly elim inated by the in tro d u ctio n of the s tr u c tu r a l
v a ria b le s as c o n tro ls . Since th i s conclusion remains te n
t a tiv e , i t also seems worthwhile to complete the lo g ic a l
development of the theory at th is p o in t.
In the introductory section of th is paper, review
of the l i t e r a t u r e led to a d ecisio n to in v e s tig a te the r e
la tio n s h ip between suburban residence and familism. Most
p rio r research has emphasized nuclear familism in suburbia;
inform ation concerning the re la tio n s h ip between extended
familism and area of residence was found to be scanty and
inconclusive. In developing the theory, i t was hypothe
sized th a t familism, including extended familism, is p o s i
tiv e ly associated with suburban residence. This hypothesis
was based on (1) e a r l i e r findings of a re la tio n s h ip between
extended and nuclear familism, and (2) the assumption th a t
127
extended and nuclear familism would d if fe r only in degree,
not in d ire c tio n , in th e ir association with other v aria b les.
The l a t t e r assumption was not, however, borne out by the
data an aly sis. Instead, extended family o rie n ta tio n was
found to be associated with non-suburban residence.
For th is reason, hypotheses of the revised theory
which include the v a ria b le , extended family o rie n ta tio n ,
are stated as d ir e c t opposites of hypotheses in the o r ig i
nal theory. Except for this change, the rev isio n is simply
the f i r s t h a lf of the o rig in al theory; i t s p o stu lates cor
respond to P o stu la te s 1, 2 and 3 of the o rig in a l. The
three variab les, in te ra c tio n with kin, residence of kin,
and r e s id e n tia l s t a b i l i t y , are not included in th is r e v i
sion of the theory.
P o stu la tes
Al. The newer the area of residence, the younger the
head of the household.
A2. The younger the head of the household, the larger
the household.
A3. The la rg e r the household, the le s s the extended
family o rie n ta tio n .
Theorems
A4. The newer the area of residence, the larg er the
household.
A5. The younger the head of the household, the le ss
the extended family o rie n ta tio n .
A6. The newer the area of residence, the le s s the ex
tended family o rie n ta tio n .
128
All the hypotheses were em pirically supported. F ig
ure 4 thus re p re se n ts both p red ic ted and observed r e l a t i o n
ships between the v a ria b le s. Households in newer areas of
residence, when compared with those in o lder“ areas, are
younger, la rg e r, and ch aracterized by le s s extended family
o rie n ta tio n .
Familism and R esidential S ta b il ity
In ad d itio n to the discussion of familism and sub
urban residence, the introductory sectio n of th is paper con
sidered the extended kin network in the modern m etropolis.
C haracterized by in te ra c tio n (not n e c e ssa rily fa c e -to -fa c e )
among kin, th e network may or may not, according to the l i t
e ra tu re , encourage geographic propinquity and r e s id e n tia l
s t a b i l i t y . Is i t p o ssib le, given the em pirical fin d in g s of
the present research, to form ulate a sub-theory describing
the extended kin network in re la tio n to residence of kin
and r e s id e n tia l s t a b i l i t y ?
The v a ria b le s in th is research most d ir e c tly r e l a t
ed to d e s c rip tio n s of the extended kin network are the
measures of familism, in te ra c tio n with kin and extended
family o rie n ta tio n ; the s tr u c t u r a l v a ria b le , residence of
kin; and the c r i t e r i o n , r e s id e n tia l s t a b i l i t y . However,
n e ith e r in te ra c tio n with kin nor extended family o r ie n ta
tio n was found in s ig n ific a n t a sso cia tio n with re s id e n tia l
s t a b i l i t y . In te ra c tio n with kin, but not extended family
FIGURE 4
REVISED
Area of
Residence
Age of Head
of Household
Household
Size
THBQRY OF URBAN FAMILISM: MATRIX OF INTERRELATIONS*
(PREDICTED A N D EMPIRICAL)
New
Old
Young
Old
Large
Small
Age of
Head
Household
Size
Extended
Young Old Large Small High Low
■ N .
*Arrows in d ic a te re la tio n s h ip s implied by the theory. Double arrows specify postu
l a t e s ; sin g le arrows sp ecify theorems.
130
o rie n ta tio n , was associated with residence of kin. Resi
dence of kin was p re d ic tiv e of r e s id e n tia l s t a b i l i t y .
The lim ited associations found between these v a ri
ables do not permit axiomatic form ulation. I t can be said
only th a t (1) In te ra c tio n with kin is associated with r e s i
dence of kin, and (2) Residence of kin is associated with
r e s id e n tia l s t a b i l i t y . The th ird possible asso ciatio n , be
tween in te ra c tio n with kin and r e s id e n tia l s t a b i l i t y , was
not found.
Age and Size of Household, Area of
Residence, and R esidential S ta b ility
I t may be noted th a t the remaining associations
found among v ariab les of the theory may be te n ta tiv e ly
sta ted as an additional minor sub-theory of the o rig in al
axiomatic theory. A co n sisten t re la tio n s h ip , as o rig in a lly
predicted, occurred among area of residence, size of house
hold, and re s id e n tia l s t a b i l i t y , a s follow s:
1. The newer the area of residence, the larger the
household.
2. The larg er the household, the greater the residen
ti a l s t a b i l i t y .
3. The newer the area of residence, the greater the
AiThe association between household size and r e s i
d e n tial s t a b i l i t y , however, appeared only in the differen ce
between one-person and other households.
131
re s id e n tia l s t a b i l i t y .
Age of head of household, which was also associated
as predicted with area of residence and household siz e , was
inversely associated with r e s id e n tia l s t a b i l i t y . Older
heads of households move le s s freq u en tly than younger, and
they tend to liv e in older neighborhoods; n ev erth eless,
older neighborhoods are le ss sta b le than newer re s id e n tia l
a r e a s .
I t appears th at area of residence and age of head
of household are independently re la te d to r e s id e n tia l s t a
b i l i t y . Introduction of a control for age of head of house
hold does not reverse the d ire c tio n of the asso ciatio n be
tween area of residence and r e s id e n tia l s t a b i l i t y ; at each
age le v e l, m obility ra te s are higher in old than in new
re s id e n tia l areas. (See Table 38.) A s t a t i s t i c a l l y s ig n i
fic a n t association occurs only in the 30-39 age category,
however, again re fle c tin g the r e la tiv e ly small c e ll values
which r e s u lt from the introduction of co n tro ls. At the
same time, the d is trib u tio n of the mobile by age level is
very sim ilar for old and new areas of residence; in both,
12
the younger re sid e n ts are s ig n ific a n tly more mobile.
1 P
I t should be recalled th a t there is some curvi-
li n e a r i t y in the re la tio n s h ip between area of residence and
age of head of household; the older the area the older the
head, except for households in which the head is under 30.
There i s , then, a su b stan tial proportion of the youngest
households in older r e s id e n tia l areas.
TABLE 38
Age of Head
of Household
MOBILITY RATES BY AGE OF HEAD AND AREA OF RESIDENCE
Area of Residence
(Date B u ilt Up)
Before 1950 1950 and a fte r df X2 C C/C max
Under 30 56.5 41.0 1 2.50 > .0 5 .138 .195
30-39 34.9 17.7 1 8.14 < .0 1 .184 .260
40-49 24.2 15.1 1 2.75 > .0 5 .105 .149
50-59 18.8 15.2 1 0.19 > .0 5 .032 .045
60 & over 13.1
*
1 0.18 > .0 5 .032 .045
*N for c e ll is l e s s than 10.
133
(See Table 39.) The a sso c ia tio n s of area of residence and
age of head of household w ith r e s id e n tia l s t a b i l i t y do n o t,
th e re fo re , appear to be a p p ro p riate ly pred icted by an ax io
matic model.
In sim ila r fash io n , siz e of household and age of
head of household may be independently re la te d to re s id e n
t i a l s t a b i l i t y . Although older households are ty p ic a lly
smaller than younger (except fo r those in which the head is
13
under 30), la rg e r households are at l e a s t as s ta b le . In
tro d u ctio n of a co n tro l for age suggests th a t a t l e a s t in
the 30-39 age category, la rg er households are s ig n if ic a n tly
more s ta b le than sm aller households. (See Table 40.)
Again, th i s asso c ia tio n is not p red icted by the axiomatic
model.
Summary
This chapter has discussed the r e s u lts of the empir
ic al t e s t of the theory of urban fam ilism . P a tte rn s of a s
so c ia tio n between v a ria b le s have been reviewed, and they
have been summarized in a restatem ent of the 21 p o s tu la te s
and theorems to conform to the em pirical evidence. As r e
s ta te d , the hypotheses no longer f i t the axiomatic model.
Consequently, d e ta ile d co n sid eratio n has been given to the
1 Q
The o rig in a l fin d in g was of no a s so c ia tio n be
tween household siz e and r e s id e n tia l s t a b i l i t y , except th a t
one-person households were le s s s ta b le .
Age of Head
of Household
TABLE 39
AGE DISTRIBUTION OF THE MOBILE, BY AREA OF RESIDENCE
Area of Residence
(Date B u ilt Up)
Before 1950 1950 and a f te r
Under 30 27.8% 27.8 %
30-39 25.8 25.6
40-49 20.0 23.3
50-59 13.6 13.3
60 & over 1 2 .8 1 0 .0
100.0 (138) 100.0 (88)
df 4 4
X2 51.45 21.01
< .001 < .0 0 1
c .298 .207
C/C max .421 .293
Age of Head
of Household
TABLE 40
MOBILITY RATES BY SIZE OF HOUSEHOLD AND AGE OF HEAD
Size of Household
Small Cl-3) Large (4 or more) df X2 C C/C max
Under 30 54.5 42.7 1 1.46 ' T .05 .105 .149
30-39 40.0 16.2 1 15.20 < .001 .247 .349
40-49 20 .5 17.5 1 0 .2 0 > .05 .032 .045
50-59 18.9
*
1 0.80 > .0 5 .063 .089
60 & over 13.3
*
1 0.01 > .05 .000 .000
*N fo r c e ll i s l e s s than 10.
u >
U l
measures of familism used; measures of in te r a c tio n w ith kin
and fam ily o r ie n ta tio n have been re fin e d , and c o n tro l v a r i
ables introduced, in an e f f o r t to improve th e ir e ffic ie n c y
as p r e d ic to r s . This e f f o r t met with no success in the case
of in te r a c tio n with k in . The e ffic ie n c y of extended fam ily
o r ie n ta tio n as a p re d ic to r of area of residence was la rg e ly
accounted fo r by the in tro d u c tio n of the s tr u c t u r a l v a r ia
b le s , age of head of household and siz e of household, as
c o n tro ls . N either of the measures of fam ilism was e f f e c
tiv e in p re d ic tin g r e s id e n tia l s t a b i l i t y .
An e m p irica lly supported re v is io n of the theory,
c o n s istin g of three p o s tu la te s and th re e theorems, has been
form ulated on the b a s is of the a s s o c ia tio n s between extend
ed fam ily o r ie n ta tio n , age of head of household and siz e of
household, and area of resid en ce. Households in newer areas
of resid e n c e , when compared w ith those in older area s, are
younger, la r g e r , and c h a ra c te riz e d by le s s extended fam ily
o r ie n ta t io n .
The remaining re la tio n s h ip s found, w ith the l i m i t
ed exception of those between area of re sid e n c e , siz e of
household, and r e s id e n tia l s t a b i l i t y , did not lend them
selves to refo rm u latio n in axiomatic terms. I t could be
said only th a t (a) in te ra c tio n with kin was asso c ia ted with
residence of kin, and residence of kin with r e s i d e n t i a l
s t a b i l i t y ; and Cb) age of head of household was r e la te d to
r e s id e n tia l s t a b i l i t y , but app aren tly independent of i t s
137
r e la tio n s h ip with size of household and area of resid e n ce .
In the following concluding chapter, the an a ly sis
of fin d in g s concerning the proposed theory of urban fa m il
ism w ill be f u r th e r summarized, and conclusions drawn. Sug
g estio n s fo r the d ir e c tio n of fu rth e r rese arch in t h i s area
w ill also be made.
CHAPTER VI
SU M M A RY A N D CONCLUSIONS
The e f f o r t made in th is research to develop a theo
ry of urban familism grew out of a review of the l i t e r a t u r e
focused on the following questions:
1. Is there evidence of an extended family system in
the modern metropolis?
2. Is the extended family system associated with r e s i
d e n tia l propinquity and r e la tiv e ly low r e s id e n tia l mobil
ity ?
3. Is there evidence of nuclear familism associated
with suburban residence?
4. Are extended familism and nuclear familism p o si
tiv e ly associated with each other?
Measures of familism, selected c rite rio n v aria b les—
r e s id e n tia l s t a b i l i t y and area of residence, and a n tic ip a t
ed s tr u c tu ra l c o rre la te s of fam ilism --in which residence of
kin was included, together with age of head of household
and size of household--were re la ted to one another through
the model of an axiomatic theory. The theory was form ulat
ed to permit d ir e c t te s tin g of questions 2 and 3 above. A
138
139
p o s itiv e re la tio n s h ip between the v a ria b le s of question 4
was assumed in the o p eratio n al d e f in itio n of an a t titu d i n a l
measure of familism, and provision was also made fo r t e s t
ing th is assumption in the course of te s tin g the hypotheses.
Data bearing on question 1 were in c id e n ta lly gathered in
the process of te s tin g the theory.
The discussion which follows considers each of
these questions in tu rn , summarizing the em pirical evidence
which c o n trib u te s toward answers for each. Subsequently,
a tte n tio n w ill be turned to an evaluation of the theory of
urban familism. I ts stren g th s and weaknesses w ill be in d i
cated, and d ire c tio n s for fu rth e r in v e stig a tio n designed to
c la rify remaining questions w ill be suggested.
The Extended Family System
1. Is there evidence of an extended fam ily system in
the modern m etropolis?
In our m etropolitan Los Angeles sample, 78.3 per
cent of a l l respondents reported the presence of kin in the
m etropolitan area; 33.2 per cent of the to t a l had kin l i v
ing in th e ir immediate neighborhood. Of respondents with
kin in the m etropolitan area, 54.4 per cent in te ra c te d with
one or more re la tiv e s * at le a s t once a week; a to ta l of
^•Only fa c e -to -fac e in te ra c tio n was measured. Addi
tio n a l contact by mail and, e s p e c ia lly , by telephone can be
assumed.
140
77.6 per cent v is ite d kin at le a s t once a month. When kin
lived in the neighborhood, frequent in te ra c tio n was the
ru le ; 84.5 per cent saw each other at le a s t weekly. R e iss’s
g en eralizatio n regarding the ecological stru c tu rin g of kin
in te ra c tio n is well supported by these data.
The extent of presence of kin in the area and in t e r
action with kin reported is comparable to th at described in
other stu d ies which have pointed to the frequency of kin in
te ra c tio n as an in d icato r of the existence of sustained fam
i l y re la tio n s h ip s in the m etropolis. For a summary of com
p arativ e fin d in g s, see Table 41.
In addition to the data presented in Table 41, the
present research included a t titu d i n a l d ata on extended kin
re la tio n s h ip s . Characterized as extended fam ily-oriented,
on the basis of an a ttitu d in a l measure, were 48.4 per cent
of survey respondents. This measure was based on complete
agreement with the statem ent, "People should always get to
gether with r e la tiv e s on holidays and other special occa
sio n s ." Extended family o rie n ta tio n , the a ttitu d in a l meas
ure, was not associated e ith e r with the behavioral measure
of familism, in te ra c tio n with kin, or with the presence of
kin in the m etropolitan area. These findings are comparable
^ a u l J . Reiss, "The Extended Kinship System: Cor
r e la te s of and A ttitudes on Frequency of In te ra c tio n ," Mar
riage and Family L iving, XXIV (November, 1962), 335-336.
141
TABLE 41
COMPARATIVE FINDINGS: PER CENT OF
METROPOLITAN RESPONDENTS REPORTING
FACE-TO-FACE INTERACTION W ITH KIN
Frequency of In te ra c tio n
Survey Weekly Monthly
Los Angeles 1962 54% 78%
Los Angeles 1954?^ 35-55 65-80
4
San Francisco 1953 30-45 56-72
D etro it 1952?^ 49 74
Buffalo 19526 52 —
to L itw ak's, supporting h is theory th a t extended family or-
ie n ta tio n is not dependent on geographic propinquity and
•a
Scott Greer and E lla Kube, Urban Worlds: A Compar
a tiv e Study of Four Los Angeles Areas (tos Angeles: Occi
d en tal College, 19$5), p. 74. In th is study and in the San
Francisco study, in te ra c tio n ra te s were reported se p arate ly
for four social area types, and no single fig u re is a v a il
able.
4Wendell Bell and Marion D. Boat, ''Urban Neighbor
hoods and Informal Social R e la tio n s," American Journal of
Sociology, LXII (January, 1957), 391-398.
^Morris Axelrod, "Urban S tru ctu re and Social P a r t i
c ip a tio n ," American S ociological Review, XXI (February,
1956), 1 3 - l T .
Eugene Litwak, "Geographic M obility and Extended
Family Cohesion," American Sociological Review, XXV (June,
1960), 388.
142
7
fa c e -to -fa c e in te ra c tio n with kin.
Other s tu d ie s have gathered more d e ta ile d inform a
tio n concerning types of so c ial a c t i v i t i e s and mutual aid
Q
among kin. Such d ata were not secured in th is rese arch .
However, fu rth e r in s ig h t into c e rta in asp ects of extended
kin r e l a t i o n s , s p e c i f i c a l l y those r e f le c tin g age and gener
a tio n a l d iffe re n c e s in a t titu d e s and behavior, was provided
by th is re se a rc h and w ill be discussed here.
Age and G enerational D ifferences
in Extended Familism
Several types of evidence suggest th at extended kin
r e la tio n s are a more ac tiv e concern of p aren ts than of
th e ir grown c h ild re n . In the f i r s t p la ce, the p resen t r e
search has found th a t extended family o rie n ta tio n tends to
c h a ra c te riz e older rath e r than younger households. The
g re a te r the extended family o r ie n ta tio n , the older the head
of the household, the sm aller the household, and the older
7Ib id .
O
For d a ta on mutual aid between kin in other metro
p o lita n a re a s, see Harry Sharp and Morris Axelrod, "Mutual
Aid among R e lativ es in an Urban P o p u la tio n ," P rin c ip le s of
Sociology, ed. Ronald Freedman e t a l . (New York: Henry
Holt ana C o ., 1956), pp. 433-43^7 Rarvin B. Sussman, "The
Iso la te d Nuclear Family: Fact or F ic tio n ," Selected Stud
ie s in Marriage and the Family, ed. Robert F. Winch, Robert
McGinnis, and H erbert R. Barringer (New York: H olt, Rine
h a rt and Winston, 1962), pp. 49-57 (Adapted from S ocial
Problems, VI [1959], 333-340); and Sussman, "The Help P a t
te rn in th e M iddle-Class Fam ily," American S o cio lo g ical Re
view, 1CVIII (February, 1953), 22-21T
the area of residence.
Second, d ata on in te ra c tio n by r e la tio n s h ip in d i
cate th a t, among respondents with kin in the m etropolitan
area, the g re a te s t frequency of in te ra c tio n occurs between
parents and th e ir grown c h ild re n . However, p aren ts re p o rt
more frequent in te ra c tio n with grown ch ild re n than i s r e
ported by grown children with p a re n ts. This finding is not
in c o n s is te n t. I t probably r e f l e c t s the fa c t th a t some p a r
ents in te r a c t with more than one grown ch ild in the metro
p o lita n area; and thus the sum of th e ir in te ra c tio n with
th e ir c h ild re n , ra th e r than in te ra c tio n with one child only,
is r e fle c te d in the reported frequencies.
These fin d in g s may be regarded as co n siste n t with
other stu d ie s which have suggested a g re a te r investment in
extended kin r e la tio n s by the older generation than by the
younger. R eiss, who asked respondents whether they had
been able to see r e l a t i v e s as much as they would lik e ,
found th a t older respondents tended to wish for more con
ta c t, but younger respondents did not. P arents of grown
children were the only category of respondent in which a
s u b s ta n tia l segment expressed the wish to have kin liv in g
9
in the same neighborhood. F u rth er, older respondents
c la s s if ie d th e ir married children as "fam ily ," while
9
Half of a l l respondents wished to have kin c lo se r,
liv in g in the same larg e c i t y , but most did not want them
in the neighborhood.
144
younger respondents c l a s s i f i e d th e ir p are n ts as " r e la -
u1 0
t i v e s . "
At the same tim e, Reiss found t h a t in te r a c tio n de
creased with su ccessiv e phases of the fam ily cy c le, appar
e n tly as the cum ulative e f f e c t of r e s id e n tia l m o b ility on a
k in sh ip system. As the years go by, kin tend to liv e f u r
ther a p a r t . I n the p rese n t re se a rc h , on the o th er hand,
n e ith e r residence of kin nor in te ra c tio n w ith kin was found
to vary d i r e c tly with age. Older respondents were as l i k e
ly as younger respondents to have kin p re se n t in the m etro
p o lita n area and to i n t e r a c t fre q u e n tly with them.
B e l l 's d e s c rip tio n of the young f a m ily 's move to
the suburbs as m otivated by conjugal r a th e r than extended
fam ilism ( p r in c ip a lly by a concern fo r c h ild re n ) is also
re le v a n t to th i s d isc u ssio n . Few of his respondents moved
in order to be close to kin; r a th e r , sev eral commented th a t
the move was d e s ira b le in th a t i t took them f a r th e r away
12
from r e l a t i v e s . While no inform ation i s a v a ila b le on the
a t t i t u d e s of the kin moved away from, i t may be speculated
th at i t was the l a t t e r ' s i n t e r e s t in freq u en t in te ra c tio n
^ R e i s s , op. c i t . , 337-338.
11I b i d . , 336.
^W endell B ell, "Social Choice, L ife S ty le s , and
Suburban R esidence," The Suburban Community, ed. William M.
Dobriner (New York: G. P. Putnam^s Sons, 1958), p. 239.
145
which led the young fam ilies to wish to liv e at a g reater
di stance.
Robins and Tomanec*s finding of an inverse r e l a
tio n sh ip between size of fam ily of p ro creatio n and involve
ment with kin led them to conclude th a t an abundance of r e
la tio n s h ip s w ithin the nuclear family tends to replace in -
13
tensive re la tio n sh ip s with extended k in . This hypothesis
may provide additional support for findings th a t postpar-
en tal couples seek more closeness with kin than do young
fam ilies with ch ild ren . In the present research, an asso
c ia tio n between in te ra c tio n with kin and size of household
was not found; however, extended fam ily o rie n ta tio n was a s
sociated with household siz e . Since frequency of in te r a c
tion is at le a s t as lik e ly to r e f l e c t opportunity as d e sire
for closeness, these findings may be regarded as c o n siste n t
with those of Robins and Tomanec, whose ’fcloseness to k in ’1
measure was not based on frequency of i n t e r a c t i o n . ^
Summary
The great m ajority of survey respondents in the
13
Lee N. Robins and Miroda Tomanec, "Closeness to
Blood R elatives outside the Immediate Family," Marriage and
Family L iving, XXIV (November, 1962), 345.
*^I b id . Rather, i t was based on the number of ave
nues used for communication, performance of se rv ic e s, and
fu lfillm e n t of obligations between respondent and kin.
Each in d icato r of these categories was scored as p rese n t,
absent d esp ite opportunity, or no opportunity.
146
m etropolitan area stu d ied in t h i s research rep o rted the
presence of kin in the area. Of those with kin p re s e n t,
over h a lf v i s i t e d w ith one or more r e l a t i v e s as o ften as
once a week. These fin d in g s suggest th a t a c tu a l i n t e r a c
tio n is somewhat more widespread than is a c le a r fe e lin g of
o b lig a tio n to get together with r e l a t i v e s . S lig h tly le s s
than h a lf of a l l respondents agreed com pletely th a t people
always should get to g e th e r with r e l a t i v e s on sp e c ia l occa
sio n s; n e v e rth e le s s , the m a jo rity d_o get to g eth er with r e l
a tiv e s a good deal more fre q u e n tly than s p e c ia l occasions
would re q u ire . The households of our sample appear to p a r
t i c i p a t e , fo r the most p a r t, in extended kin networks; at
the same tim e, th ere i s evidence th a t a good many regard
th e ir p a r tic ip a t io n as e s s e n t i a l l y v o lu n tary .
The extended kin network ty p ic a lly extends across
two or more g en e ra tio n s, with the most in te n s iv e i n t e r a c
tio n th at between p are n ts and t h e i r grown c h ild re n . Of
these two g en e ra tio n s, the older tends to be more a c tiv e ly
involved, both in e x p e c ta tio n s and in ac tu al behavior, in
extended kin r e la tio n s h i p s .
Extended Familism and R e sid e n tia l
P ropinquity and M obility
2. Is the extended fam ily system a s so c ia te d with r e s i
d e n tia l pro p in q u ity and r e l a t i v e l y low r e s i d e n t i a l m obil
ity ?
There is general agreement in the l i t e r a t u r e with
147
Litwak’s view th a t contemporary extended fam ily r e l a t i o n
ships do not req u ire geographic propinquity for th e ir main
t e n a n c e .^ However, H a l l e r ^ and Sussman and B u rc h in a l,^
while agreeing with Litwak th a t propinquity is not re q u ire d ,
have suggested that i t may be p referre d by fam ilies with
commitment to an extended kin network. F u rth er, living
near r e la tiv e s may be po ssib le without occupational s a c r i
fic e for many fa m ilie s, since the m etropolitan area makes
av ailab le a wide range of occupational choice. Litwak,
while emphasizing the occupation-based m o b ility of the bu
re a u c ra tic middle c la s s , has hypothesized a cy c lic a l p a t
tern of m obility. While the extended family may support
moves, even moves away from kin, made by the young family
on a career upswing, l a t e r ’’geographical coalescence” r e
su ltin g in r e s id e n tia l closeness w ill tend to occur when a
career plateau is reached.
In the present study as in Litwak’s , an a t titu d i n a l
measure, extended fam ily o rie n ta tio n , was not associated
with residence of kin nor with r e s id e n tia l m obility. The
^ L itw a k , op. c i t . , 385-394.
16A. O. H aller, "The Urban Fam ily,” American Jour
nal of Sociology, LXVI (May, 1961), 621-622.
17
Marvin B. Sussman and Lee Burchinal, ”Kin Family
Network: Unheralded S tru c tu re in Current Conceptualiza
tions of Family F unctioning,” Marriage and Family L iving,
5CXIV (August, 1962), 231-240.
18
Litwak, op. c i t . , 387.
148
extended fam ily-oriented were no more lik e ly than oth ers to
have kin nearby, and they were as lik e ly to move as those
respondents not so o rien te d .
The behavioral measure of familism, in te ra c tio n
with k in , was applied only to respondents with kin present
in the m etropolitan area. Consequently, only propinquity
within the m etropolis could be compared with frequency of
in te ra c tio n . Face-to-face in te ra c tio n took place more f r e
quently for respondents with kin liv in g in th e ir own neigh
borhood than for those w ith kin elsewhere in the m etropoli
tan area. A fu rth e r decrease in in te ra c tio n frequency for
19
those with no kin in the m etropolis can be assumed.
D ifferences in in te ra c tio n w ith kin, w ithin the
m etropolitan area, were not r e fle c te d in d iffe re n c e s in
r e s id e n tia l m ob ility . However, the presence or absence of
kin in the area, which tends to f a c i l i t a t e or hinder f r e
quent in te ra c tio n , was p re d ic tiv e of d iffe re n c e s in r e s i
d e n tia l m obility. Thirty-two per cent of respondents w ith
out kin in the area, but only 20 per cent of those with kin
p re se n t, moved during the year (1961-62) of the research .
This d iffe re n ce was s t a t i s t i c a l l y s ig n ific a n t.
Whether or not kin liv e near each other within the
metropolitan area, in the same neighborhood, helps deter
mine frequency of interaction but is unrelated to m obility.
19
I b i d . ; R eiss, op. c i t .
149
But when the measured d iffe re n c e is r e s t r i c t e d to th a t be
tween respondents with kin p rese n t in , or absent from, the
m etro p o lita n area, i t is^ p re d ic tiv e of m o b ility . As Suss
man and Burchinal have suggested, clo sen ess of residence
w ithin a m etro p o litan a rea , given modern communication and
tr a n s p o r ta tio n f a c i l i t i e s , is probably not of g re a t impor-
20
tance to the extended kin network. Whether or not r e l a
tiv e s are p re se n t in the m etro p o lita n area does, however,
appear to have im p licatio n s fo r r e s id e n tia l s t a b i l i t y .
I t should be noted th a t fo r respondents both with
and w ithout kin p re se n t in the area, the m ajority of moves
which took place were in tra -c o u n ty . (In the sample as a
whole, 84.4 per cent of the moves made in the study year
21
took place w ith in the county.) There was a somewhat
g re a te r p ro p o rtio n of o u t-o f-co u n ty moves made by those
w ithin kin in the area; however, c e l l values were small and
the d iffe re n c e was not s i g n i f i c a n t . (See Table 19, page
77.) These fin d in g s are somewhat unexpected. To the ex
te n t th a t d e s ire to l i v e near kin in flu en c es r e s i d e n t i a l
m o b ility , i t seems reasonable th a t the presence of kin in
the m etropolis w ill be accompanied by decreased m o b ility
out of the area. Why, however, should i t also be
20
Sussman and B urchinal, op. c i t .
21
The design of the survey was such th a t m obility
into the m etropolitan area, which is high in Southern C a li-
f o rn ia , is not r e f le c te d in these f ig u re s .
150
accompanied by decreased i n t r a -m etro p o litan m o b ility ? This
question aw aits a s a t i s f a c t o r y answer.
In summary, expressed commitment to kin was not a s
so ciated with decreased r e s i d e n t i a l m o b ility , nor was a c tu
al frequency of in te r a c tio n with kin liv in g in the m etro
p o lita n area. However, the o pportunity for in te r a c tio n
with k in , as measured by the presence or absence of kin
w ithin the m etropolitan a re a , was s i g n i f i c a n t l y a sso c ia ted
22
with decreased r e s i d e n t i a l m o b ility , even though the
great m ajo rity of the m obility which occurred did not take
the movers out of the area.
Nuclear Familism and Suburban Residence
3. Is th ere evidence of nuclear fam ilism a sso cia ted
with suburban residence?
Data on reasons fo r moving to the suburbs, which
23
have been presented in other stu d ie s in support of a
fam ilism -suburban a s s o c ia tio n , have not been analyzed in
the present re se a rc h . R eliance has been placed on
22This finding would not be expected on the b a s is
of L itw ak's theory, though he did not present d ata s p e c if ic
to t h i s p o in t. I t supports the hypotheses of H aller and of
Sussman and B urchinal.
23Wendell B ell, op. c i t . , pp. 225-247; E rnest R.
Mowrer, ’’The Family in Suburbia, ” The Suburban Community,
ed. William M. Dobriner (New York: G. P. Putnam^s Sons,
1958), pp. 147-164; Mowrer, "Sequential and C lass V ariables
of the Family in the Suburban A rea,” Social F o rc e s, XL (De
cember, 1961), 107-112.
151
measurement of the a sso c ia t ion between familism and present
suburban re sid e n c e , using the a t t i t u d i n a l measure, fam ily
24
o r ie n ta tio n , to in d ic a te n u clear fam ilism . While the in
i t i a l a n a ly s is of fin d in g s yielded an a lm o s t-s ig n ific a n t
a s s o c ia tio n between n u clear fam ily o r ie n ta tio n and suburban
resid en ce (see Chapter IV, pages 79-81), f u r th e r a n a ly sis
of the measurement of fam ily o r ie n ta tio n led to the conclu
sion th a t nuclear fam ily o r ie n ta tio n , as d efin ed , simply
rep resen ted the absence of extended fam ily o r ie n ta tio n and
was not adequate to stand alone as a measure (see d is c u s
sion Chapter V, pages 115-119). Consequently, no d i r e c t
measure of nuclear fam ilism is a v a ila b le at t h i s stage of
the a n a ly s is .
Extended fam ily o r ie n ta tio n , however, was n e g a tiv e
ly asso cia ted with suburban resid e n c e . The extended
fa m ily -o rie n te d in t h i s sample tended to be o lder persons
liv in g in sm aller households in older neighborhoods. This
fin d in g co rro b o ra tes th a t of Robins and Tomanec concerning
an in v erse r e la tio n s h ip between household s iz e and c lo s e -
25
ness to k in , as well as R e is s ’s re p o r t that persons in
l a t e r sta g e s of the l i f e cycle d e s ire more kin in te r a c tio n
In r e tr o s p e c t i t appears u n fo rtu n a te th a t a be
h a v io ra l measure of nuclear fam ilism , comparable to the
in te ra c tio n -w ith -k in measure of extended fam ilism , was not
included in the survey in terv iew . See d isc u ssio n of meas
urement of fam ilism , page 152f.
25
Robins and Tomanec, op, c i t .
152
. - . 26
than do younger persons.
As a n tic ip a te d , suburban households were g e n e ra lly
younger and la rg e r in s iz e than those in older c ity neigh
borhoods, and they included a g re a te r p ro p o rtio n of i n t a c t
f a m ilie s . This rese arch has, however, as a r e s u l t of prob
lems of measurement, provided no d ata bearing d i r e c t l y on
the q uestion of the a s so c ia tio n between fam ilism , regarded
as an a t t i t u d i n a l or behavioral v a ria b le , and suburban r e s
idence .
Extended Familism and Nuclear Familism
4. Are extended familism and nuclear familism p o s i
tiv e ly asso c ia te d with each other?
As in d icated above, the intended d i r e c t measure of
n u clear fam ilism in th i s rese arch did not prove s a t i s f a c t o
ry. Analysis of the a s so c ia tio n between the a t t i t u d i n a l
measures of extended and nuclear fam ily o r ie n ta tio n o r ig in
a l l y employed (see Chapter V, pages 115-119) yielded a pos
i t i v e a s so c ia tio n between the two. However, when extended
and nuclear fam ily o rie n ta tio n were defined as m utually ex
c lu s iv e , th e ir a s s o c ia tio n w ith other v a ria b le s was in op
p o s ite d ir e c tio n s ; and when nuclear family o rie n ta tio n was
taken as a measure se p arate from extended fam ily o r ie n ta
tio n , i t was asso cia ted with none of the other v a ria b le s in
26n . .
R eiss, op. c i t .
153
the theory. For th is reason the decision was made to elim
in a te nuclear fam ily o rie n ta tio n from fu rth e r co n sid eratio n .
I t can be concluded, then, e ith e r th a t (1) nuclear familism
is not p re d ic tiv e of the other v a ria b le s of th is research,
or (2) the nuclear family o rie n ta tio n measure used does not
adequately in d ic a te the presence or absence of nuclear f a
milism.
The c r itic is m of o th e rs 1 operational d e f in itio n s of
familism which has been made by Rogers and Sebald is of in
t e r e s t in the present d iscu ssio n . While i t is g en erally
agreed, they s ta te , th a t familism is properly defined as
the subordination of individual in te r e s ts to those of the
family group, indexes of familism have tended to include
both nuclear familism (which Rogers and Sebald conceptual
ize as family in te g ra tio n ) and extended familism (which
they denote as kinship o r ie n ta tio n ). Such indexes are am
biguous and have questionable v a lid ity .
The d if fe re n t group stru c tu re of the two sys
tems demands d if f e r e n t f a m ilis tic a t titu d e s . Sub
sequent em pirical data support the statem ent th a t
one can hardly serve each of two reference groups
to the extent that he could serve one. As a person
surrenders individualism to the extended fam ily,
th i s makes i t impossible at times to focus on nu
clear family goals.
Several w rite rs have presented evidence to support
E verett M. Rogers and Hans Sebald, M A D istin c tio n
between Familism, Family In teg ra tio n , and Kinship O rien ta
tio n ," Marriage and Family Living, XXIV (February, 1962),
26.
154
the hypothesis of no a sso c ia tio n between fam ily in te g ra tio n
28 29
and kin sh ip o r ie n ta tio n . Both Wilkening and Cleland
found no c o rre la tio n between general familism measures and
a s p e c if ic family in te g ra tio n scale. Rogers and Sebald, who
developed scales of both fam ily in te g ra tio n and kinship o r i-
30
e n ta tio n , found no r e la tio n s h ip between them.
I t should be noted th a t the sc a le s used by these
w rite rs were based on behavioral d a ta . A ttitu d in a l measures
of fam ilism , on the other hand, have tended to yield a p o s i
tiv e asso c ia tio n between nuclear and kinship in d ic a to rs , as
was the case in the present research . Litwakt s extended
fam ily -o rien ted respondents tended to be nuclear fam ily-
31
o rien te d as w ell. When B ard is’s familism sc a le , composed
of a t t i t u d i n a l item s, was subdivided into the two subcate
g o ries of nuclear and extended familism, a strong c o rre la -
32
tio n was found between the two.
In v estig atin g the a s so c ia tio n between a t t i t u d i n a l
(B ardis) and behavioral (Sebald) measures of family
28
Eugene A. Wilkening, "Change in Farm Technology
as Related to Familism, Family Decision Making, and Family
In te g ra tio n ," American S ociological Review, XIX (February,
1954), 29-37.
29
Courtney B. Cleland, "Familism in Rural Saskatche
wan," Rural Sociology, XX (September, 1955), 249-257.
30
Rogers and Sebald, op. c i t . , 25-30.
31
Litwak, op. c i t .
32
Rogers and Sebald, op. c i t .
155
in te g ra tio n , Rogers and Sebald found no s ig n ific a n t c o rre la
tio n between them. They concluded th a t fam ily in te g ra tio n
a ttitu d e s and behavior may be separate dimensions, and p re
sented suggestions for the improvement of measures of fa -
... . 33
milism.
The findings of the present research co n trib u te only
in d ir e c tly to the complex question of the r e la tio n s h ip be
tween nuclear and extended familism. Both a t t i t u d i n a l and
behavioral measures of extended familism were employed, and
the finding th a t they were not associated with each other is
c o n siste n t with Rogers and Sebald’s findings noted above.
For nuclear familism only an a ttitu d in a l measure was used,
and the fin d in g s have opened i t s v a lid ity to question.
Several w r i t e r s ^ have, however, used l i f e cycle
stages and/or d iffe re n ce s in actual or desired f e r t i l i t y as
in d ic ato rs of nuclear familism, regarding the younger and
larger households as the more f a m il is ti c . In d ic a to rs of age
of head of household and size of household, defined as an
tic ip a te d s tru c tu ra l c o rre la te s of familism, were included
as v a ria b le s in the present research. So, a lso , was area of
residence; and suburban residence has been described by Bell
33Ib id .
3^Bell, op. c i t . ; Reuben H ill with E lise Boulding,
Families under S tre ss: Adjustment to the C rises of War Sep
aration and Reunion (.New York: Harper B ros., 1^49;, pp.
124-126; E. Gartley Jaco and Ivan Belknap, " Is a New Family
Form Emerging in the Urban Fringe?" American Sociological
Review, XVIII (October, 1953), 551 - 5 ^ T .
156
and Jaco and Belknap, among o th ers, as a c o rre la te of fam il-
35
ism.
All th ree of these measures were, in the p rese n t r e
search, neg ativ ely associated with extended fam ily o rie n ta
tio n . To the degree th at these measures can be used appro
p r ia te ly as in d ic ato rs of nuclear familism, as other stu d ies
have done, our fin d in g s may be regarded as providing f u r
ther confirm ation of a lack of asso cia tio n between nuclear
and extended familism.
Evaluation of the Theory of Urban Familism
This research has attempted to develop a functional
theory of urban familism based on an axiomatic model. Sub
urban residence and lower r e s id e n tia l m obility were con
ceived as functions of f a m il is ti c o rie n ta tio n s and behav-
3 6
i o r , and additional s tru c tu ra l c h a r a c te r is tic s were hy
pothesized as fu rth e r c o rre la te s of familism and residence
behavi o r .
B ell, op. c i t . ; Jaco and Belknap, op. c i t .
O i l
For the sake of c l a r ity in d isc u ssio n , measures of
familism have been re fe rre d to in the course of th is r e
search as the independent v a ria b le s, and r e s id e n tia l lo c a
tio n and s t a b i l i t y as the c r ite r io n v a ria b le s. As o r ig in a l
ly sta te d (Chapter I, page 12} however, each was ac tu a lly
regarded as a p o te n tia l function of the other, with no nec
essary d ire c tio n or causation im plied. The axiomatic model,
which allows lo g ic a lly for the r e v e r s ib i lity of hypotheses,
makes th is assumption c le a r.
157
The theory in which these v a ria b le s were form ally
in te r r e la te d f a ile d , as a whole, to survive em pirical t e s t
ing. I t was unsuccessful p r in c ip a lly in (1) supporting the
assumption of a re la tio n s h ip between extended and nuclear
family o rie n ta tio n and (2) p red ic tin g r e s id e n tia l s t a b i l i t y
on the b a sis of measures of familism. Use of the axiomatic
model did, however, f a c i l i t a t e lo c atio n of the sources of
the f a ilu r e of hypotheses to meet the em pirical t e s t , and
o ffer a guide to d ire c tio n s for fu rth e r in v e s tig a tio n .
I t was p o ssib le to id e n tify , through analysis of
findings and some re v isio n of measurement, an em p irically
supported sub-theory of the o rig in a l which stands upon i t s
own. This revised theory, which s ta te s the re la tio n s h ip s
between extended family o rie n ta tio n , size of household and
age of head of household, and area of residence, in te g ra te s
and summarizes previous research fin d in g s. The inclusion
of extended family o rie n ta tio n in th is theory i s not, how
ever, necessary to explain observed d iffe re n ce s in the data;
the intro d u ctio n of co n tro ls for age and size la rg e ly elim
inated the efficacy of extended family o rie n ta tio n in p re
d ic tin g area of residence. There i s , then, no clear e v i
dence of a functional re la tio n s h ip between extended fam ily
o rie n ta tio n and non-suburban residence.
On the other hand, the analysis id e n tifie d sectio n s
of the theory in which the axiomatic model is not ap p lic a
ble. Notable among these is the in te rr e la tio n s h ip among
158
age of head of household, siz e of household, area of r e s i
dence, and r e s i d e n t i a l s t a b i l i t y . The a s s o c ia tio n s between
age and siz e and age and area are ap p a re n tly independent of
the a s s o c ia tio n between age and r e s id e n tia l s t a b i l i t y ;
th e re fo re , these r e la tio n s h ip s cannot be lo g ic a l ly i n t e
grated in an axiom atic theory.
I t may be f u r th e r suggested th a t the r e la tio n s h ip
between extended and nuclear fam ily o r ie n ta tio n , and be
tween each of these v a ria b le s and other v a r ia b le s , also
37
f a i l s to f i t the axiom atic model. The adequacy of the
measure of n u clear fam ilism used in t h i s research is sub
je c t to q u estio n . N ev erth eless, i t i s , on i t s fa c e , very
s im ila r to th a t employed by Litwak, whose fin d in g s were not
38
e n t i r e l y c o n s is te n t w ith those of the p re se n t re se a rch .
The a t t i t u d i n a l measures of fam ily o r ie n ta tio n were a s s o c i
ated w ith each oth er; n e v e rth e le s s , th e ir a s s o c ia tio n w ith
other v a ria b le s was not in the same d ir e c tio n . This prob
lem was not d i r e c t l y incorporated in to the theory as a hy
p o th e s is , and fu rth e r in v e s tig a tio n is needed. Thus f a r ,
the e f f o r t made in th is and other s tu d ie s to in te g ra te the
two l i n e s of s o c io lo g ic a l concern with fam ilism , extended
kin r e la tio n s h ip s and nuclear fam ily o r ie n ta tio n (or
37
Cf. Rogers and S ebald's claim th a t extended and
nuclear familism are u n re la te d to each o th er; op. c i t .
38_ . . . .
Litwak, op. c i t .
159
39
in te g ra tio n ) , has not been f r u i t f u l .
The analysis did not support the hypothesis of a
functional r e la tio n s h ip between familism and r e s id e n tia l
s t a b i l i t y , thus confirming Litwak*s findings th a t extended
family re la tio n s h ip s are maintained independent of resid en
t i a l moves.40 At the same time, however, r e s id e n tia l s t a
b i l i t y did appear, to some e x te n t, to be a function of the
presence or absence of kin in the m etropolitan area. The
l a t t e r was defined as an extended-fam ily s tr u c tu r a l charac
t e r i s t i c , rath er than as a measure of familism. There may
be some b a sis, however, for regarding i t as a behavioral
measure of familism which supplements the measure of i n t e r
action with kin.
Because th ere was no asso cia tio n between age of
head of household or size of household and residence of kin,
i t was not thought necessary to introduce co n tro ls for
these variables in measuring the asso ciatio n between r e s i
dence of kin and r e s id e n tia l s t a b i l i t y . Unless the i n tr o
duction of fu rth e r co n tro ls, notably th a t for socioeconomic
s t a t u s , 4* r e s u lts in the elim ination of th is re la tio n s h ip ,
i t cannot be said th a t re s id e n tia l m obility is alto g eth er
■*°For a discussion of other stu d ies in th is area,
re f e r to Chapter I, pages 7-11.
4®Litwak, op. c i t .
4*See discussion below, pages 162-166.
160
independent of influence from the extended kin network.
On the whole, th is research has not provided much
support for the conception of a functional r e la tio n s h ip be
tween the extended family network and other so cial sys-
4?
terns, with the exception of the fin d in g , only p a r t i a l l y
explained, of a re la tio n s h ip between kin residence and r e s
id e n tia l s t a b i l i t y . Nor has i t c l a r if i e d the re la tio n s h ip
between nuclear familism and other v a ria b le s.
Unresolved Questions
Further research on the questions ra ise d by th is
study should include a tte n tio n to problems in the measure
ment of familism, to the e lu c id a tio n of the asso cia tio n be
tween residence of kin and r e s id e n tia l s t a b i l i t y , and to
the possible e f f e c ts of socioeconomic fa c to rs on the fin d
ings reported in this research . D irectio n s fo r research
w ill be suggested below, following which a b r ie f summary
w ill conclude the paper.
Measurement of Familism
Problems in the measurement of familism have been
discussed at length above (see Chapter V, pages 109-121,
and Chapter VI, pages 152-156). The present research con
firmed previous re p o rts of a lack of asso cia tio n between an
42
Cf. Sussman and Burchinal, op. c i t . , and d isc u s
sion above, Chapter I , pages 3-4.
161
a t t i t u d i n a l measure of extended fam ilism and r e s i d e n t i a l mo
b i l i t y , but a p o ssib le a s s o c ia tio n between extended family
behavior (as r e f le c te d in resid en ce of kin) and m o b ility was
re p o rte d . Findings concerning the r e la tio n s h ip between ex
tended and n u clear f a m i l i s t i c a t t i t u d e s were p a r t i a l and in
c o n s is te n t. Because nuclear f a m i l i s t i c behavior was not
measured, no in v e s tig a tio n of i t s r e la tio n s h i p to extended
f a m i l i s t i c behavior was p o s s ib le .
The a t t i t u d i n a l measures of fam ilism used in th is
43
study appear very s im ila r to those used by Litwak and
yielded sim ila r r e s u l t s . They were s e le c te d , on the b a sis
of th e ir face v a l i d i t y , from among the items included in a
fam ilism scale in the survey in te rv iew . The extended-fam ily
item se le c te d was the only such measure included in the
sc a le . The sc ale contained two l o g ic a l ly p o ssib le measures
of nuclear fam ilism , but t r i a l an a ly sis of both in d icated
th a t th e sin g le item se le c te d d isc rim in a ted alone about as
Litwak, op. c i t . , 388. Litwak re lie d on respond
ent agreement with four sTatements—the f i r s t two measuring
n u c le a r, the l a s t two extended, fam ily o r ie n ta tio n . How
ever, he a c tu a lly had only one e f f e c tiv e measure of extended
fam ily o r ie n ta tio n , since the second e l i c i t e d very l i t t l e
agreement. The statem ents were as follow s:
1. G enerally I lik e the whole fam ily to spend evenings
to g e th e r.
2. I want a house where fam ily members can spend time
to g e th e r .
3. I want a lo c a tio n which would make i t easy fo r r e l
a tiv e s to g et to g e th e r.
4. I want a house with enough room for our p aren ts to
f e e l f re e to move in .
162
e ffe c tiv e ly as the two in combination. The design of the
research, which involved a d is tin c tio n between nuclear and
extended family o rie n ta tio n , required the lo g ic a l se le c tio n
of appropriate items rather than the use of the scale as a
whole.
Since the d is tin c tio n between the two types of fam
i l y o rie n ta tio n as measured proved u n sa tisfa c to ry , fu rth e r
analysis u til iz i n g the e n tire scale as a measure of familism
may be undertaken. A rec e n tly completed la te n t class analy
s is of the scale makes possible the use of la te n t classe s as
in d ic a to rs of the presence or absence of familism, here de
fined as including both nuclear and extended-family compon
ents. However, past c ritic is m s of th is type of composite
44
scale cast some doubt upon the v a lid ity of i t s use.
Future studies in th is area should give considera
tion to the inclusion of separate a ttitu d in a l and behavioral
measures of both extended and nuclear familism. Further
c la r if i c a t io n both of the in te rre la tio n s h ip s among these
v ariab les and of th e ir p red ictiv e value with respect to oth
er family c h a ra c te ris tic s is needed.
Kin Residence and R esidential S ta b ility
This research has found a p o sitiv e association be
tween residence of kin and r e s id e n tia l s t a b i l i t y ;
44
See discussion above, pages 152-156.
163
respondents with r e l a t i v e s in the m etropolitan area were
le s s mobile than respondents without kin p re se n t. When con
t r o l s were introduced for the d ire c tio n of moves, however,
i t was apparent that the great m ajority of a l l moves took
place w ithin the county, and th a t out-of-country moves were
no more c h a r a c te r is tic of those with r e la tiv e s absent than
of those with r e la tiv e s p re se n t. Since i t had been a n t i c i
pated th a t m obility r e la te d to residence of kin would serve
to re lo c ate movers closer to th e ir kin, th is finding with
regard to d ire c tio n of moves was unexpected, and no explan
a tio n has yet been o ffered .
I t is p o ssib le th at Litwak’s hypothesis of "geo-
graphical coalescence" among kin may provide d ire c tio n in
the search for an explanation. Litwak has suggested th at
while the extended fam ily may encourage moves of the young
nuclear family as these are required for occupational
achievement, fa m ilie s which have reached a career plateau
w ill tend to s e t t l e near kin or enable kin to jo in them.
This hypothesis pertained p rim arily to the m iddle-class,
45
b u re a u c ra tic a lly employed family head. I t is lo g ic a lly
fe a s ib le th a t co n tra stin g p a tte rn s of m obility ex istin g be
tween younger and older households, and/or between middle-
c la s s bu reau cratic heads of households and those not in th is
category, could cancel each other out in the data a n a ly sis.
^ L itw a k , op. c i t . , 387.
164
Therefore, i t would be ap p ro p riate to introduce co n tro ls fo r
age of head of household and for socioeconomic c h a ra c te ris
tic s into the an a ly sis of the a s so c ia tio n between residence
Afi
of kin and r e s id e n tia l s t a b i l i t y .
I t should be noted th a t those d a ta in the present
research which might r e l a t e to the question of geographical
coalescence lend no support, as analyzed thus f a r , to the
hypothesis. For the most p a rt th ere was no as so c ia tio n be
tween residence of kin and age of head of household or size
of household, except fo r some lim ited evidence in support of
Reiss*s report th a t d ista n ce of kin residence tends to in-
47
crease with age of the head of the household. In our sam
ple, households in which the head was over 60 were le s s
l ik e ly than others to have kin in the immediate neighborhood.
One- and two-person households were also le s s li k e l y to have
kin in the neighborhood, though only one-person households
were also le s s lik e ly to have kin in the m etropolitan area.
Control for Socioeconomic D ifferences
Socioeconomic le v e l and occupational type were not
46For the in tro d u ctio n of c o n tro ls to be p r a c t ic a l,
a la rg e r sample than our 979, or the use of a re tro s p e c tiv e
measure of m o b ility , might be necessary. The actu al number
of mobile respondents (226) in th is one-year panel design
was too small to permit extensive use of co n tro l v aria b le s
when d iffe re n ce s w ithin the mobile group alone were of p r in
cip al in te re s t.
47
R eiss, op. c i t .
165
introduced as v a ria b le s in t h i s study. I t was deemed d e s i r
able to b u ild , if p o s s ib le , a generalized theory in which
measures of fam ilism and of r e s id e n tia l lo c a tio n and mobil
it y could be in te r r e l a te d without lim ita tio n by so c ia l c la ss
c r i t e r i a . Since the generalized theory was not achieved,
however, fu rth e r a n a ly sis should u t i l i z e so cial c la s s meas
ures as c o n tro ls, since there i s considerable reason to be
lie v e th a t the family c h a r a c te r i s tic s measured d i f f e r by
socioeconomic le v e l.
Gans, for example, has claimed th a t so cial c la s s to
gether with fam ily l i f e cycle is cru c ia l in determ ining
areas of r e s i d e n c e . ^ The present research has v e rifie d
the influence of l i f e cycle stages; adding co n sid eratio n of
socioeconomic c h a r a c te r i s tic s would serve to round out the
a n a ly s is . Mowrer’ s an a ly sis of suburban d ata has coupled
occupational d iffe re n c e s with the ’’suburban c y c le ’’ in ac
counting for d iffe re n c e s in familism as well as in other
49
c h a r a c t e r i s t i c s . Litwak’s theory of fam ily o r ie n ta tio n
and geographic m o b ility is described as ap p licab le prim arily
48
Herbert J . Gans, ’’Urbanism and Suburbanism as
Ways of L ife: A R e-evaluation of D e f in iti o n s ,” Human Be
havior and Social P rocesses: An I n te r a c tio n is t Approach,
ed. Arnold M. Rose CBoston: Houghton M ifflin Co. , 19t>£),
pp. 625-648.
49
Mowrer, ’’Sequential and C lass V ariables of the
Family in the Suburban A rea,” lo c . c i t .
166
to those in bureaucratic m iddle-class occupations.^0 In tro
duction of so cial class in d ic a to rs of the analysis of our
data would permit more precise comparison with the findings
of such stu d ies.
Summary
This research has developed an axiomatic theory of
urban familism, including measures of familism, an ticipated
s tru c tu ra l c o rre la te s of familism, and c rite rio n v ariab les
of area of residence and r e s id e n tia l s t a b i l i t y , and su b je c t
ed i t to empirical t e s t . The theory as a whole fa ile d to
conform to the em pirical d ata; consequently, a more in te n
sive analysis was undertaken in an e ff o rt to c l a r if y r e l a
tionships among the v ariab les. P a rtic u la r consideration
was given in th is analysis to problems in the measurement
of familism and to the formulation of an em pirically sup
ported sub-theory of the o rig in a l theory.
The most important findings of the research are as
follow s:
1. The greater the extended family o rie n ta tio n , the
older the head of the household, the smaller the household,
and the older the area of residence. The in te rre la tio n sh ip s
among these v ariab les are consistent with the axiomatic
model.
50 . . . . ,
Litwak, op. c i t .
167
2. The e ffe c tiv e n e s s of the a t t i t u d i n a l measure, ex
tended fam ily o r ie n ta tio n , in p re d ic tin g area of residence
appears to be elim inated by the in tro d u c tio n of co n tro ls
fo r age of head and siz e of household. S tru c tu r a l v ariab les
can probably be s u b s titu te d fo r the so c ial-p sy c h o lo g ica l
v a ria b le in p re d ic tin g area of resid en ce.
3. N either extended fam ily o rie n ta tio n nor in te ra c tio n
with kin liv in g in the m etropolitan area is p re d ic tiv e of
r e s id e n tia l s t a b i l i t y .
4. Residence of k in , which is asso ciated with frequen
cy of in te ra c tio n with kin, is p re d ic tiv e of r e s id e n tia l
s t a b i l i t y . Households with kin p rese n t in the m etropolitan
area are le s s mobile than households without kin p re se n t.
5. The g reater p a r t of the m obility p red icted by r e s i
dence of kin took place w ithin the county. D iffe re n ces in
d ir e c tio n of move (w ithin or out of the county) were not
p red icted by residence of k in . The a s so c ia tio n between
residence of kin and decreased in tram e tro p o litan m obility
was not a n tic ip a te d and is not re a d ily explained.
6. The newer the area of residence, the la rg e r the
household and the g reater the r e s id e n tia l s t a b i l i t y .
7. The newer the area of residence and the la rg e r the
household, the younger the head of the household. These
a s so c ia tio n s are, however, apparently independent of th a t
between age and r e s id e n tia l s t a b i l i t y , which is not p re
d ic ted by an axiomatic model. The younger the head of the
168
household, the l e s s the r e s id e n tia l s t a b i l i t y .
8 . The la rg e m ajority of households in the m etropoli
tan area p a r tic ip a t e a c tiv e ly in extended kin r e la tio n s h ip s .
9. Extended kin re la tio n s h ip s ty p ic a lly extend across
two or more g en e ra tio n s, with the most in ten siv e in te r a c
tio n th a t between parents and th e ir grown ch ild re n . Of
these two g en eratio n s, the older is more a c tiv e ly involved
in kin re la tio n s h ip s than the younger.
10. Nuclear familism has not been shown to be a sso c i
ated with suburban residence or with r e s id e n tia l s t a b i l i t y .
Doubts were r a is e d , however, as to the adequacy of the at-
tit u d in a l measure used.
11. The conception of a fu n ctio n a l re la tio n s h ip be
tween the extended fam ily network and other so cial systems
is not stro n g ly supported by th is research .
12. The axiomatic model was f r u i t f u l in e s ta b lis h in g
the nature of p re d ic te d asso c ia tio n s and in c la rify in g the
nature of a sso c ia tio n s which fa ile d to f i t the model.
LITERATURE CITED
Axelrod, M orris. ’’Urban S tru c tu re and S ocial P a r t i c i p a
t i o n , ” American S ociological Review, XXI (February,
1956), T S - H T .------------- -------------------
B ardis, Panos D. ’’A ttitu d es toward the Family among College
Students and th e ir P a r e n ts ,” Sociology and S ocial
R esearch, XLIII (May-June, 1959), 332-358.
B ell, Wendell. "Social Choice, Life S ty le s , and Suburban
R esidence,”' The Suburban Community, ed. William M.
Dobriner (New York: G. ! p . 'Putnam's Sons, 1958), pp.
225-247.
_________, and Boat, Marion D. "Urban Neighborhoods and In
formal Social R e la tio n s," American Journal of S oci
ology, LXII (January, 1957TJ 391-398.
Cleland, Courtney B. "Familism in Rural Saskatchewan,"
Rural Sociology, XX (September, 1955), 249-257.
Costner, Herbert L. " C r ite r ia for Measures of A sso c ia tio n ,”
American S ociological Review, XXX (June, 1965),
-------------------------
Duncan, Beverly; Sabagh, Georges; and Van Arsdol, Maurice D.,
J r . "P attern s of C ity Growth," American Journal of
Sociology, LXVII (January, 1962)” 418-429.
Duncan, O tis Dudley, and R eiss, Albert J . , J r . Social Char
a c t e r i s t i c s of Urban and Rural Communities, 1950.
New York: John Wiley and Sons, 1956.
Fava, Sylvia F le is . "C ontrasts in Neighboring: New York
City and a Suburban County," The Suburban Community,
ed. William M. Dobriner (New York: G. £. Putnam's
Sons, 1958), pp. 122-131.
Foote, Nelson N. "Family Living as P la y ," Marriage and Fam
i l y L iving, XVII (November, 1955), 296-361.
169
170
Freedman, Ronald; Whelpton, P ascal K .; and Campbell, Arthur
A. Family P lanning, S t e r i l i t y , and Population
Growth. New York: McGraw-Hill Book Co., 1959.
Gans, Herbert J . "Urbanism and Suburbanism as Ways of L ife:
A R e-evaluation of D e f in iti o n s ,’1 Human Behavior and
Social P ro cesses: An I n te r a c tio n is t Approach, ed.
Arnold M. Rose ( Boston: Houghton M ifflin Co.,
1962), pp. 625-648.
Glick, Paul C. American F am ilies. New York: John Wiley
and Sons^ 19 57.
_________. ’’The Life Cycle of the Fam ily,” Marriage and
Family L iv in g , XVII (February, 1955), 3-9.
Gold, David. ’’Comment on fA C ritiq u e of T ests of S i g n i f i
c a n c e ,” ’ American S ociological Review, XXIII (Feb
ruary, 1958 ) , 85i-§6.
Greer, S c o tt, and Kube, E lla . Urban Worlds: A Comparative
Study of Four Los Angeles A reas. Los Angeles: Oc
c id e n ta l College, 1955.
H a lle r, A. O. "The Urban Fam ily," American Journal of So
ciology, LXVI (May, 1961), 621-622.
H ill, Reuben, with Boulding, E lise . Families under S tr e ss:
Adjustment to the Crises of W ar Separation and Re
union. New York: Harper B ros., 1949.
Jaco, E. G artley , and Belknap, Ivan. "Is a New Family Form
Emerging in the Urban Fringe?" American S o cio lo g i
cal Review, XVIII (October, 1953), 351-557.
Kish, L e slie . " S t a t i s t i c a l Problems in Research Design,"
American S ociological Review, XXIV (June, 1959),
'323-137:------------ -------------------------
Litwak, Eugene. "Geographic M obility and Extended Family
Cohesion," American S ociological Review, XXV (June,
1960), 385-19TI
_________. "Occupational M obility and Extended Family Cohe
sio n ," American S ociological Review, XXV (February,
1960), 9 - 2 l .
171
Locke, Harvey C. "A Model for Small-Group Theory: General
iz a tio n s Derived from Empirical R esearch,” U n iv ersi
ty of Southern C a lifo rn ia , mimeo. (Paper read be
fore the F if t y - th i r d Annual Meeting of the American
S ociological S ociety, at S e a t tle , Wash., 1958.)
Meeker, Marchia, with H a rris, Joan R. Background for P la n
ning . Research Report No. 17, Research Department,
Welfare Planning Council, Los Angeles Region, 1964.
Mowrer, Ernest R. "The Family in Suburbia," The Suburban
Community, ed. William M. Dobriner (New York: 57 P.
Putnam1s Sons, 1958), pp. 147-164.
"Sequential and Class V ariables of the Family in
the Suburban A rea," Social Forces, XL (December,
1961), 107-112.
Parsons, T a lc o tt. "Revised A nalytical Approach to the The
ory of S ocial S t r a t i f i c a t i o n , " C lass, S ta tu s, and
Power: A Reader in Social S t r a t i f i c a t i o n , ed.
Reinhard Bendix and Seymour M. L ipsei CGlencoe,
111.: The Free P re ss, 1953), pp. 116ff.
"The Social S tru c tu re of the Fam ily," The Fam ily:
'I t s Function and D estiny, ed. Ruth N. Anshen (.'New
York: Harper and B ros., 1949), pp. 241-274.
R eiss, Paul J . "The Extended Kinship System: C o rrela tes
of and A ttitu d e s on Frequency of I n te r a c tio n ," Mar
riage and Family Living, XXIV (November, 1962),
333-339.
Robins, Lee N., and Tomanec, Miroda. "Closeness to Blood
R e lativ es Outside the Immediate Fam ily," Marriage
and Family L iv in g , XXIV (November, 1962), 340-346.
Rogers, E verett M., and Sebald, Hans. "A D is tin c tio n be
tween Familism, Family In te g ra tio n , and Kinship Or
ie n ta t io n ," Marriage and Family L iv in g , XXIV (Feb
ruary, 1962)"J 33-30.
Rossi, P eter H. Why Fam ilies Move: A Study in the Social
r ' ^ ± a l M o b ility . Glencoe,
Sabagh, Georges, and Van Arsdol, Maurice D., J r . , e t a l .
The R estle ss M etropolis: R e sid e n tia l M obility in
Los Angeles. Los Angeles: U niversity of Southern
C a lifo rn ia , 1965, mimeo.
172
Schmidt, John Frank, and Rohrer, Wayne C. ’’The R e latio n
ship of Family Type to Social P a r t i c i p a t i o n ,” Mar
riage and Family Living, XVIII (August, 1956),
224-230.
Schrag, Clarence. ’’Some Notes on Crim inological T heory,”
Conference on Research Planning on Crime and De
linquency, ed. William R. Larson (.Los Angeles:
Youth Studies Center, U niversity of Southern C a li
fo rn ia , 1962), pp. 1-16.
Schwirian, Kent P ., and Prehn, John W . "An Axiomatic Theo
ry of U rb a n iza tio n ,” American S ociological Review,
XXVII (December, 1962), 812'-825.
Selvin, Hanan C. ”A C ritiq u e of T ests of S ignificance in
Survey R esearch,” American S ociological Review,
XXII (October, 1957^7"STS
Sharp, Harry, and Axelrod, M orris. "Mutual Aid among Rela
tiv e s in an Urban P o p u la tio n ,” P rin c ip le s of Soci
ology , ed. Ronald Freedman e_t aT"I (New York:
Henry Holt and Co., 1956), pp.T 3 3 -4 3 9 .
S iegel, Sidney. Nonparametric S t a t i s t i c s for the Behavior-
al Sciences. New York: McGraw-hill Book Co., In c .,
--------------
Sussman, Marvin B. "Family C ontinuity: S ele ctiv e Factors
which Affect R elationships between Families at Gen
e ra tio n a l L evels," Marriage and Family Living, XVI
(May, 1954), 1 1 2 - 1 2 W .
_________. "The Iso lated Nuclear Family: Fact or F ic tio n ,"
Selected Studies in Marriage and the Family, ed .
Robert F. Winch, Robert McGinnis, and Herbert R.
Barringer (New York: H olt, Rinehart and Winston,
1962), pp. 49-57. Adapted from Social Problems, VI
(1959), 333-340.
_________, and Burchinal, Lee. "Kin Family Network: Unher
alded S tru c tu re in Current C onceptualizations of
Family Functioning," Marriage and Family L iv in g ,
XXIV (August, 1962), 231-240.
Wilkening, Eugene A. "Change in Farm Technology as Related
to Familism, Family Decision Making and Family In
te g r a tio n ," American Sociolo g ic a l Review, XIX (Feb
ruary, 1954)~ 2^-57.
173
Winch, Robert F. The Modern Fam ily. New York: H olt, Rine
h a rt and Winston, 1963.
Z e tte rb e rg , Hans. "On Axiomatic Theories in S o c io lo g y ,”
The Language of S ocial Research: A Reader in the
Methodology of S ocial k e se arc h , ed. P aul I f. L azars-
fe ld and Morris Rosenberg (Glencoe, 1 1 1 .: The Free
P re ss, 1955), pp. 533-540. Reprinted from On Theo-
ry and V e r if ic a tion in Sociology (The T re s le r P re s s ,
1954), pp. 16-28.
Zimmerman, Carle C. Family and C i v i l i z a t i o n . New York:
Harper and B ro s ., 1^47.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

An Empirical Examination Of The Relationship Of Vertical Occupational Mobility And Horizontal Residential Mobility

PDF

Consensus Of Role Perceptions In A Welfare Planning Council

PDF

Social Components Of Housing Cost In The Western Metropolis

PDF

A Study Of Relationships Between Occupational And Marital Roles And Marital Adjustment

PDF

Complementarity, Homogeneity, Heterogeneity And Marital Stability

PDF

Conflict And Generalized Conflict In Verbal And Motor Responses

PDF

Class Consciousness And Social Mobility In A Mexican-American Community

PDF

Reward Expectancy Strength As Related To The Magnitude Of Frustration In Children

PDF

Perception Of The Power Structure By Social Class In A California Community

PDF

Ecology Of Negro Communities In Los Angeles County: 1940-1959

PDF

Differential Fertility Behavior And Values In Rural And Semi-Urban Costa Rica

PDF

The Construction And Empirical Test Of A Theory Based On Selected Variables In Small-Group Interaction

PDF

Social Class Membership And Ethnic Prejudice In Cedar City

PDF

Economic Differentiation And Social Organization Of Standard Metropolitanareas In The United States: 1950

PDF

Some Social Factors Affecting The Power Structure And Status Of A Professional Association In Reference To Social Work

PDF

Group Factors And Individual Internalization Of A Value

PDF

Determinants Of Intercounty Migration: California, 1970-1973

PDF

On Becoming A Parent: Attitude And Feeling Changes

PDF

Referential Dissociation And Response To Stress

PDF

The Major League Professional Baseball Player: A Sociological Study

Asset Metadata

Creator Jones, Elizabeth Jensen (author)

Core Title Familism, Suburbanization, And Residential Mobility In A Metropolis

Degree Doctor of Philosophy

Degree Program Sociology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,sociology, individual and family studies

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Van Arsdol, Maurice D., Jr. (committee chair), Longstreth, Langdon E. (committee member), McDonagh, Edward C. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-124237

Unique identifier UC11360022

Identifier 6706502.pdf (filename),usctheses-c18-124237 (legacy record id)

Legacy Identifier 6706502.pdf

Dmrecord 124237

Document Type Dissertation

Rights Jones, Elizabeth Jensen

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA