You are on page 1of 40

Social Science History Association

Family Structure in Los Angeles, California: 1850-1900


Author(s): Barbara Laslett and Katherine Nash
Source: Social Science History, Vol. 20, No. 1 (Spring, 1996), pp. 1-39
Published by: Social Science History Association
Stable URL: http://www.jstor.org/stable/1171502
Accessed: 07-04-2015 22:32 UTC

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.

Social Science History Association is collaborating with JSTOR to digitize, preserve and extend access to Social Science
History.

http://www.jstor.org

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
BarbaraLaslettandKatherine
Nash

FamilyStructure
in Los Angeles,
California
1850-1900

In an overviewof recentresearchon the historyof the family, Tamara


Hareven(1991)pointsout thatthisfieldof studytookits inspiration from
developments in historicaldemography and from the "new social history"
ofthe1960s.Familyhistorians, likeothersocialhistorians,
had "a commit-
mentto reconstructing thelifepatternsofordinary people,to viewingthem
as actorsas wellas subjectsin theprocessofchange"(ibid.:95). The flower-
ing of researchin thisfieldhas providedus witha moredetailedunder-
SocialScience 20:1(spring
History C 1996bytheSocialScienceHistory
1996).Copyright Association.

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
2 SocialScience
History

standingoftherelationship betweensocialchangeand familylifethanwas


previously available.
We have learned,amongotherthings, thatratherthana
singletrajectory ofchangefromextendedfamilylifebeforeindustrialization
to thenuclearfamilyafterward, changesin familyorganization haverarely
beeninvariant, linear,orunidirectional.We havealso learnedthatchangesin
family organization over timereflectedthe agencyoffamily membersas they
triedto meetthe demandsof social reproduction and thattheirsuccesses
and failureswereshapedbyhistorically specific conditionsand resources.
The conceptoffamilystrategies is one waythatagencyamongfamilies
has been theorized.(For a reviewof thisconceptsee Moen and Wething-
ton 1992.)It reflects an assumptionthatpeople'slivesare not determined
by social structure but ratherthatindividuals,families,and othersocial
groupsmakechoicesbetweenalternative coursesof action;throughthese
choicesand thebehaviorsthatfollowfromthem,humanagentsconstruct
and changesocialinstitutions. Questionsneed to be raised,however, about
how familystrategiesare decided on. Do all familymembersaffectthe
choicesto an equal degree?How do individualsocialcharacteristics, suchas
race/ethnicity, class, maritalstatus,and gender(each of which mayreflect
variationin resources, power,meanings, and values)affectfamilystrategies
and contestation overthem?'
Certainly,individualcharacteristics alonedo notshapethechoicesthat
individualsand familiesmake because the social and historicalcontexts
withinwhichtheyare madevaries.The socialorganization ofwelfare, cul-
turalbeliefsand practices,the availability of nonfamilialresourcesto care
fordependents, and howaccess to theseresourcesis sociallydistributed--
whatLaslettandBrennerhavecalledtheorganization ofsocialreproduction
(Brennerand Laslett1986;Laslettand Brenner1989)--allplayimportant
rolesin familystrategies. Thus, herewe are also interested in howchanges
in socialinstitutions overhistorical
time--through urbanizationand in the
organization of the economy, thedistribution and sourcesof power,gender,
racerelations, and culture- affect familylife.
Our empiricalfindings showthatchangesin bothpopulationcharac-
teristicsand structural relationsaffected the historyof familystructure in
Los
nineteenth-century Angeles.2Assuming that familyliving was valued
by mostnineteenth-century Californians, changingrelationships to there-
sources(material,cultural,and institutional) associatedwithlivingin one

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
Structure
Family inLosAngeles3

kindoffamily oranothercanalsohelpaccountforchangesin thedistribution


ofLos Angelesfamiliesbetween1850and 1900.Changingpopulationchar-
especiallythefloodofEuropeanAmericanmigrants
acteristics, and thevir-
tualdisappearanceofHispanics, also to a
contributed distribution offamily
typesdifferent at the end of thenineteenth centurythanin themiddleof
it. Although the relationsobserved in the analysiswe presenthereare sta-
tisticalonly,theyallowus to theorizefamilylifeat one timeand place and
to considerhowbothsocialstructure and humanagencyshapedchangesin
familyorganization in thenineteenth-century UnitedStates.3
Many different historical
currents and social processesat workin Los
Angelesduringthelasthalfofthenineteenth centuryaffected familyorga-
nization.The expansionof U.S. territory to includethe Americansouth-
west;the subsequenttransformation of the economyfromhometo local,
to nationaland thento international marketproduction;and the highrate
ofEuropeanAmericanmigration intoCalifornia - all had significant conse-
quences forhow and whyfamily structure changed as it did. But we need to
ask questionsotherthaneconomicand demographic ones.In particular, our
findings show that and
gender ethnic/race relations were also consequential
forfamilystructure in thistimeand place and,it seemsreasonableto sug-
gest,in othersas well.
U.S. women'shistorians and sociologists havedemonstrated thatmajor
transformations in genderrelationsoccurredin thenineteenth century.4As
productiveactivitiesmovedawayfromthedomesticsphereto themarket-
place,men'sworkwas increasingly definedin termsof occupationswithin
a developingcapitalistlaborforce,and theexpectations forwomen'swork,
especiallyfor middle-class urbanwomen, were increasingly limitedtofamily
responsibilitieswithinthe domesticsphere- particularly maintaining the
homeand childcare. Alongwiththeseeconomicchanges,a new ideology
aboutgenderrelationswithinthefamilyand withinthesocietyas a whole
also developed:theideologyofthenaturaland necessary"separatespheres"
forEuropeanAmericanwomenin the home and forEuropeanAmerican
menin themarketplace. As theseideologicaland organizational changesoc-
curredoverthecourseof thenineteenth century, therewerealso counter-
trends.New occupationsin thewagedsectorof theeconomy- factory and
white-collar to
work- becameavailable women they as were to men,and
centuryonward,womenbecamein-
earlyin thenineteenth
fromrelatively

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
4 SocialScience
History

creasinglyactivein publicinstitutions:in thechurch(Ryan1981),in educa-


tion(Clifford1991;Rury1991; Strober and Lanford1986),in culturemore
generally(Douglas 1977;Bennion1990),and,bytheend ofthecentury, in
politics(Skocpol1992; Gordon 1991b,1992; Brenner and Laslett1991).
Whilethisaccountofchangingfamily and genderrelationsin thenine-
teenthcenturyhas been importantforunderstanding the experiencesof
bothworking- and middle-classEuropeanAmericanwomenandmenin the
UnitedStates,forothers,particularly thoseindifferent racial/ethnic
groups,
changesin familyand genderrelationswereaffected by somewhatdiffer-
enttemporalities,cultures,politics,and economics.Clearly,theabolitionof
slaverywas theimportant watershed fornineteenth-century AfricanAmeri-
cans. Discrimination,limitson the availability of adequatelywagedlabor,
differential
mortality(especiallyformen),and AfricanAmerican(compared
to EuropeanAmerican)women'shigherparticipation in wagedworkafter
1865all affectedAfricanAmericanfamilylife(Goldin1977;Giddings1984;
Jones1985).
ForAsianimmigrants to theUnitedStates,especiallyChinesemigrants
to California,Hawaii,and thePhilippines,thehistoryof legal restrictions
leftan unfavorable legacyforfamilyformation. On the mainland,migra-
tionpoliciesand practicesfirstencouragedcheap male laborto enterthe
country.Women'sentrance,however,was limited,and thenChinesemi-
gration was prohibitedaltogether (Glenn 1983; Hirata 1979;Takaki1989:
chap.2). For immigrant womenin general,themigration processand,per-
haps, the Americanization programsthey confrontedin manypartsof the
UnitedStateswereimportant as well(Sanchez1990).AndforNativeAmeri-
cans,differingby time and place,thetakeovers oflandbyEuropeansettlers
and thepoliticalchangesassociatedwiththemundermined theirindigenous
formsofcommunity organization, and
familylife, gender relations
(O'Brien
1996;Phillips1980).
The historiesofHispanicwomenand familiesin thesouthwestern part
oftheUnitedStatesalso variedbylocationand timing(Alvarez1987;Gris-
wolddel Castillo1984;Garcia1980;Sanchez1990;Deutsch1987;Castanada
1990).In New Mexicanvillagesin thelatenineteenth forinstance,
century,
migrationaffected genderrelations;whilemenwereawayestablishing new
settlementselsewhere,women had more in
authority the and
family in the

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
Structure
Family inLosAngeles5

community. But women'spositionbecamea sourceofcontention withmen


whenthesefamiliessettledin Colorado(Deutsch1987).Hispanicmigration
fromBaja to Alta Californiawas shapedby the "miningcircuit"and the
familyand kinshiptiesestablished by migrants alongit intosouthernCali-
fornia(Alvarez1987).Of criticalimportance forunderstanding thesouth-
westernHispanicand NativeAmericanexperience was thehistory ofpoliti-
cal relationsbetweenMexico and the UnitedStates,especiallythe ceding
of territoryfromMexico to theUnitedStatesin theTreatyof Guadalupe
Hidalgoin 1848(Griswolddel Castillo1990;Padilla1988).
Giventhesehistorical sourcesofvariation, itis important toexaminedi-
versityin relationto familylifeand to developa theoretical perspectivethat
can accountforit. In thepresentation to follow, our goal is twofold:With
individualleveldatafromthefederalmanuscript censusschedulesfor1850,
1870, 1880, and 1900,we describe familyorganization in Los Angeles,Cali-
fornia,duringthelastfifty yearsof thenineteenth century.We also pursue
whatLouise Tilly(1989)hascalledanalytic questions relevantto understand-
ing how race/ethnicity, class, lifecourse,and genderrelationsrelatedto
historical in
changes familyorganization. We willthenconsiderourempiri-
cal findings in lightof the theoretical perspectivethatAbrams(1981)sug-
gests:We ask how theintersection ofsocial structure and humanagencyin
Los Angeles,California, in thesecondhalfofthenineteenth centuryshaped
in
changes familyorganization in it.

Los Angeles in the Nineteenth Century

Priorto 1850,social lifein Los Angelescenteredaroundthelargeranchos,


whose productionwas gearedprimarily to theirown consumptionneeds.
The majoreconomicactivities werecattleranching andwineproduction, and
tradewas limitedlargelyto thebarterof tallowand hides forluxurygoods.
Giventhenaturaland socialconditions -a mildclimateand the"domesti-
cation"oftheNativeAmericanswhohad livedin southernCalifornia prior
to itscolonization
bySpain - domesticunits were able to achievea consider-
able amountof economicand socialautonomy(see Fogelson1967;Cleland
1969;McWilliams1946).This situationwas notto lastlongaftertheAmeri-
can takeover(Pitt 1970; Griswolddel Castillo1979,1984,1990). Between

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
6 SocialScience
History

1850and 1870,Los Angelesdevelopedfroman isolated,self-sustaining, and


largelybartereconomyto a trade-dependent, moneyeconomyas European
Americanlife,customs,and institutions becameincreasingly dominant.
Centralto thesestructural changesweresomeof theparticularities of
theCalifornia case- thelegalefforts, largelysuccessful,to wrest land in and
aroundLos AngelesfromCalifornios, whichwas achievedin partthrough
the workof the Land Commissionestablishedin 1851.(Native California
Hispano-Mexicanswereknownas Californios and Californias.)The plung-
ingpricesforcattle,thedroughtsofthe1860s,and thelavishlifestyles that
reducedthe fortunes of thoseCaliforniolandownerswho retainedtitleto
theirproperty, and a patternofmigration amongfamiliesofMexicanheri-
tageto locationsthatoffered bettereconomic opportunities and thesupport
of kin (Alvarez1987) help accountforthe changesthatoccurredas well.
Race/ethnic relationswerealso affected bytheparticular politicalresponses
oftheAngloin-migrants to theHispanicpresencein AmericanLos Ange-
les: the "race war" and ethnictensionsfromthe 1850s to the 1890s that
progressively excludedtheCaliforniopopulationfrompoliticalactivity and
fostered ethnichostilities.5
From 1880to 1900Los Angeleswas transformed froma smallcityto
a majormetropolis. In 1880thecityservedprimarily as a servicecenterfor
thesurrounding richagricultural land;economicdevelopment was limited
bothbyinadequatepopulationandaccessprimarily to localmarkets. All this
changedradicallyin themid-1880swiththeconstruction of twodirectrail
linksto theeasternpartofthecountry anda pricewarbetweenthem,result-
ingin a rapidinfluxofpopulation(Dumke 1944).In 1880theLos Angeles
populationwasapproximately 11thousand;in 1900itwasover100thousand
(Fogelson1967).6Althoughneitherheavyindustry normanufacturing came
to Los Angelesuntilwellintothe twentieth century, considerable diversi-
ficationoftheeconomicbase occurredbetween1880and 1900.Agriculture
was increasingly orientedtowarda nationalmarket, and occupational oppor-
tunitiesin theserviceand commercial sectorsoftheeconomyexpanded.
Althoughwomencould inheritproperty in theirownnamebothprior
to and afterCaliforniawas ceded fromMexicoin 1848,alienationfromthe
land suffered by the Californiopopulationin generalafterthe Treatyof
GuadalupeHidalgoaffected women'smaterialfortunes in severalways.As
productionincreasingly tookplace withina marketratherthana domestic

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
inLosAngeles7
Structure
Family

economy, thevalueofwomen'sworkin homeproductionwas undermined.


At thesametime,women'saccessto marketemployment was severelylim-
ited; forthemostpart,in Los Angelesas elsewhere,paid employment for
womenwasprimarily as householddomesticandserviceworkers (see Laslett
1990; Sobek 1991;Locke 1982;but see also Folbreand Abel 1989).Never-
theless,newformsofwageworkdid becomeavailableto womenovertime.
White-collar occupationsopenedup in commerce, whiletheriseand spread
ofeducationalinstitutions createda demandforeducatedlaborthatwomen
could fill(Laslett 1990).Our analysisalso demonstrates the importance of
maritalstatus-one centraldeterminant of genderrelations-forwomen's
domesticarrangements.
In termsof ethnicity and the nationaloriginsof its population,Los
Angeleswas markedly different fromotherpartsoftheUnitedStateswhere
family lifehas been studied. Most ofthepeoplewholivedthereat thetime
Californiawas ceded to the UnitedStateswereof Hispanic descent,pri-
marilyMexican. But the Hispanic demographicpresencein Los Angeles
declinedsteadilyoverthefifty yearsofthisstudy- frommorethan70% of
the populationin 1850to less thanan estimated3% in 1900.7It is unclear
how muchof thisdeclinewas due to out-migration oftheHispanicpopu-
lationor to otherdemographicprocesses.8 What is clear,however,is that
theeconomic,political,and culturalHispanicpresencewas severelyunder-
minedbythedelugeoftheincomingEuropeanAmericanpopulation.
The in-migration to Californiaof EuropeanAmericans, especiallybe-
tween1880and 1900,was heavilyofpersonsfromelsewherein theUnited
States. In 1900,62.8% of all individualssurveyed(N = 6,839) reported
havingbeen born elsewherein the United States.9These were not the
poor "huddled masses" featuredin much of the historicalliteratureon
nineteenth-century migration to theUnitedStatesfromEurope,norwere
theythesingleadventurers thatcame to Californiain searchof gold.They
were,rather,relatively prosperousfamiliesprimarily fromthe midwestin
searchof a betterqualityof life--thebrightsunshineand vividlycolored
flowers-thatwereso muchstressedin theLos Angelescityfathers'pub-
lic relationsliterature(Marsh 1990;Fogelson1967;Fishman1987;McWil-
liams1946).
These massivetransformations - froma small,largelyMexicano/aand
Chicano/apopulation to a largelyEuropean American one; froma largely

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
8 SocialScience
History

indigenousto a largelymigrantpopulation;froma largelyself-sufficient


to a largelymarket-dependent community - raisequestionsaboutcompari-

sons. In whatwaywas Los Angelesthe same place in 1900 thatit was in


1850?Since thepopulationchangedso dramatically overtheperiodofthis
study,how can we separatehistoricalchangesin Los Angeles'seconomic
and socialorganization fromchangesin itsdemographic composition?If we
findchangesin familyorganization, howwillwe knowifthesechangesre-
sultfromstructural and culturalchangesor demographic ones?
This is morethana technicalproblem,however;itis characteristic ofall
studiesthatattemptto understand socialchangesoverhistoricaltime.Such
changesusuallyincludebothpopulationand structural but
transformations,
themagnitudeofchangeswhoseeffects are beingassesseddependson the
dateat whicha studybegins.If,in thiscase,datafrom1850werenotavail-
able,theextentofchangewouldbe muchless marked.But 1850is a crucial
timepointforthisanalysissinceit incorporates a Los Angelesthatwas still
morelikethesocietyit had been thanliketheone it was to become.What
is unique in thisresearchon Los Angelesis our capacityto examinemas-
sive changeovera fifty-year periodoftimewitha single,relatively uniform
source.1'While thisadvantage of our data does not eliminatethe analytic,
and interpretive
statistical, problems thesedataalsopose,we believeour
that
resultsenrichhow we understandlarge-scalehistoricaltransformations--
boththosespecificto Los Angelesfamiliesin the secondhalfof thenine-
teenthcentury and moregenerally.

Data and Methods

The analysisto be reportedhereis basedon datafromtheindividualmanu-


scriptfederalcensusschedulesfor1850,1870,1880,and 1900.Information
on theentirepopulationofthecityofLos Angelesin 1850(N = 1,610)was
putintomachine-readable form,as wasa 50% sampleofhouseholds(andall
the familiesand individualscontainedwithinthem)in 1870and 1880(N =
3,932 and 5,083individuals, and all individualscontainedin a
respectively)
10% randomsampleofhouseholdsin 1900(N = 10,103)."In linewithearlier
analysesof thesedata,we focusprimarily on characteristicsof the family
head.Sincetheprimary goal hereis to understand in
changes familylifeover

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
inLosAngeles9
Structure
Family

timeand in orderto maximizetemporalcomparisons, we use as indepen-


dentvariablesonlythoseitemscommonto all fourcensusyears.12
Table 1 presentsdescriptive statistics of the
on selectedcharacteristics
family heads. Changes and continuities in these characteristics
are impor-
tantto our arguments, sincewe are interested in whetherthe proportions
ofdifferent typesoffamiliesin nineteenth-century Los Angelescan be ex-
plainedbyshiftsinpopulationcharacteristics orshiftsinsocialrelationships.
Certaincautionsneed to be voicedabout thiskindof argument.First,as
alreadyexplained,we arenottrying to arguethatpopulationchangesrather
thanchangesin social relationships accountforthe changingdistribution
of familytypes;we expectbothto occur.Changingpopulationcharacteris-
tics are linkedto changesin social organization, and changesin the latter
are shapedby demographictransformations. We do not thinkone can be
adequately understood apartfrom the other. Second,we wantto notethat
relatively stablepopulationcharacteristics do notnecessarilyimplyrelatively
stablesocialrelations.For example,the firstrowof Table 1 showsthatthe
percentage offemaleheadsremainedrelatively stablefrom1850to 1900,but
thisdoes notnecessarily meanthatthesocialcharacteristics ofwomenheads
werethesame in all thosefifty years.Indeed,our analysissuggeststhatin
1850womenfamilyheadsmayhavebeenlargelywidowedlandowners while
in 1900theymayhavebeenlargelywageearningand nevermarried.
While Table 1 reportsthat the percentageof femaleheads did not
changemarkedly overtime,thatis not trueforthe ethniccompositionof
family heads. As we have alreadyindicated,the proportionof European
Americanheadsincreasedsharplyoverthefifty yearsconsideredherewhile
Hispanic heads virtuallydisappeared.By 1900 the percentageofbothAsian
and AfricanAmericanfamilyheadshad also increased,buttheywerea very
smallpercentage ofthetotal,evenat theend ofthecentury.13
Table 1 also indicatesan agingoffamilyheads:theproportion ofyoung
heads declinedand thepercentageof old (54+) headsincreased,and these
changeshelpexplainthechangingdistribution offamilytypes.14The occu-
pational distributionof familyheads also underwent transformations.First,
as footnotee of Table 1 shows, the of
proportion family heads who were
farmersdeclinedsharplybetween1850 and 1900 whileotherwhite-collar
occupations - thosewe havecalledbourgeois and thoselabeled"mid-level"

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
10 SocialScienceHistory

Table 1 Frequencydistributions (%) of familyheads by selected


Los Angeles,California:1850,1870,1880,and 1900
characteristics,
Familyheads 1850 1870 1880 1900

Gender
Female 19.8 14.3 16.7 17.3
Male 80.2 85.7 83.3 82.7

Ethnicitya
African
American 0.4 0.9 0.8 2.4
Asian 0.0 1.0 4.6 3.1
EuropeanAmerican 28.3 61.0 73.6 94.5
Hispanic 70.9 37.1 20.3 []b
NativeAmerican 0.4 []c 0.7 []d

Age
Young(18-33) 32.5 33.4 29.1 19.8
Middle(34-53) 52.6 56.5 54.9 54.2
Old (54 andolder) 14.9 10.1 16.0 26.0

Occupatione
Bourgeois 31.3 29.8 23.1 28.0
Middle-level 7.5 9.4 23.8 16.0
Skilled 11.2 12.4 20.8 16.7
Semi-unskilled 23.1 34.2 18.8 23.6
NA/Other 26.9 14.2 13.5 15.7

Birthplace
California 39.6 16.4 13.4 6.3
Mexico 25.7 19.2 9.0 0.8
OtherwithinU.S. 18.3 29.3 36.6 62.5
Otherforeign 16.4 35.1 41.0 30.4
N 268 920 1,206 2,542
aSee note5 formeasurement ofrace/ethnicity.
bSee note5 forwhythisinformation is notavailablefor1900.
CIn1870,Indian-headedfamilieswerelistedon a separatecensusscheduleand weretherefore notsampled.
dNativeAmericanswerenotpartofour 1900sample.
eOccupationalcategoriesused areas follows:Bourgeois: professionals, farmers,
agriculturalists, proprietors,
managersand government and semiprofessionals.
officials, The proportions of familyheads in thiscate-
gorythatwerefarmers, ranchersand agriculturalists areas follows:1850-19.3%,1870-19.8%,1880-9.6%,
1900-3.4%. Mid-leveloccupations:clericaland sales,pettyproprietors, lowerlevelmanagers,self-employed
artisans,and minorgovernment officials.
Skilled: skilledworkersonly; thereis no information about
whetheror not theyare self-employed. Semi-unskilled: Service,semiskilled,and unskilledworkers.NA:
All cases withmissingvaluesor listedas out ofthelaborforce.

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
inLosAngeles11
Structure
Family

- increased.Still,althoughthereis somevariationin thepercentage in what


we havecalledtheskilled,semi-skilled, and unskilledoccupations, thereis
no markedincreaseor decrease.1isThe proportion ofthosewithoutoccupa-
tionsdeclinedsharplybetween1850and 1870,butwe suspectthatis due to
thelackofinformation on womenin theearlieryear.
Changes in the birthplacesofheadsmirror whatwe foundforethnicity.
The percentageof California-born people (nonmigrants)declinedsharply
overtime;thosewhohad migratedfromMexicovirtually disappeared,and
therewas a sharpincreasein familyheads who werebornin the United
States(but not in California)and in foreigncountriesotherthanMexico.
of disentangling
The difficulties birthplace, and migration-and
ethnicity,
the different of to
types migration nineteenth-century Los Angeles--will
be discussedfurther as we presentfindings fromour multivariateanalysis,
whichincludeseach oftheindividualcharacteristics presented in Table 1.

FamilyStructure

Familystructure hasbeenoperationalized herein termsofmaritalorparent-


childlinksto thefamilyhead amongpersonswho shareda givenresidence
on the date of censusenumeration. It excludespersonswho maybe part
of the same kinshipgroupbut who live in otherfamilies;it also excludes
personswho are relatedto each otherthroughotherkinshipties.Family
structure was constructed
as a four-categorynominalvariable:16
1. Simple (or nuclear) families-i.e., domesticunits that included a
husband-wife, husband-wife-child(ren) units--andsingle-parent-child
families.Simplefamiliesalonedid notcontaineitherextendedkin(kin
beyondimmediateparent-child/husband-wife ties)or non-kin.
2. Complexfamilies-i.e.,simplefamiliesextendedwithkin;thesefami-
lies mayalso haveincludednon-kin,'7 and simplefamilyunitsextended
only with non-kin;theseunits did notcontainextendedkin.
3. Non-kinfamilies--i.e.,domesticunitsin whichpersonswere living
alone,withunrelatedothers,or withkin wheretherewas neithera
parent-child nora husband-wife relationshipto theheadoffamily(sib-
for
lings, instance).18
Unlikesubsequentcensuses,the enumerations in 1850 and 1870 pro-
videdno explicitinformation on familyrelationships; however,mostsuch

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
12 SocialScienceHistory

Table 2 (%) of familytypesin Los Angeles,


Frequencydistributions
1850,1870,1880,and 1900
California:
Inferred
relationships Observedrelationships

Familytype 1850 1870 1880 1900

Simplefamilies 28.3 45.3 51.8 57.8


Husband-wife 2.9 8.2 11.0 13.7
Parent-child 8.2 8.5 9.9 8.7
Husband-wife-child 17.2 28.7 30.9 35.3

Complexfamilies 43.7 23.8 27.1 25.1


Simplepluskina 14.2 6.4 9.8 14.2
Simpleplusnon-kin 29.5 17.4 17.3 10.9

Non-kin 28.0 30.9 21.1 17.1


N 268 920 1,206 2,542
aSome of these familiesalso includenon-kin.In 1850,20 of the 38 simpleplus kin families(or 47%)
includednon-kin.In 1870,17 ofthe59 simpleplus kinfamilies(or 29%) includednon-kin.In 1880,34 of
the118simpleplus kinfamilies(or 29%) includednon-kin.In 1900,76 ofthe286 simpleplus kinfamilies
(or 21%) includednon-kin.

relationships can be inferredfrominformation on surname,age, sex, se-


quenceoflistingwithinthefamily, race,and birthplace (Laslett1975,1977).
Such inferred have
relationships particular flawsrelevant to thisanalysis:
Kin withdifferent surnamesfromthefamilyhead- suchas mothers-in-law
or marrieddaughters--usually cannotbe identified as kin.19
Relativeto later
years,statisticsbasedon the1850and 1870censusyearsarelikelyto under-
estimateresidencewithkinandoverstate residencewithnon-kin. Therefore,
comparisonsof familycompositionbetween1850-70and 1880-1900must
be madewithcaution.Comparisons withinthesetwotimeperiods,however,
are less likelyto be affected
bydifference in thewayinformation on family
relationships, and thus familystructure,was obtained.20
The distributions offamiliesovertimeforthethreemajorfamilystruc-
turecategories and withinthesimpleand complexfamily arepre-
categories
sentedin Table 2. There was a sharpincreaseovertimein thepercentage
of householdsorganizedalongsimplefamilylines,from28.3% in 1850to
57.8% in 1900.Althoughtheincreasecan be observedthroughout thetime
period,it was mostmarkedbetween1850and 1870.Earlydebatesaboutthe

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
in Los Angeles 13
FamilyStructure

history ofthefamilysuggestedthatthetransformation fromagricultural to


industrial societiesproducedchangeofthiskind(Hareven1991;Goode 1963;
P. Laslett1965;P. Laslettand Wall1972).YetLos Angelescertainly was not
"industrial" by 1870: As Table the of
1 shows, proportion family heads who
werefarmers was19.3% in 1850and 19.8% in 1870.AndalthoughLos Ange-
les did changeoverthis20-yearperiod,it did so muchless thanit would
between1880and 1900,whenthe proportion of simplefamiliesincreased
byonly 7%. The questionis, how much of the changingdistributions, and
theirtiming,are due to populationcharacteristics, such as ethnicity,or to
structural changesin, for or
instance,race,class, gender relations? We will
considerthisquestionin thecontextofthemultivariate analysisto follow.
The breakdowns withinthe simpleand complexfamilycategoriesare
also presentedin Table 2. They can providea firststeptowardunderstand-
ing the changingdistributions. Amongsimplefamilies,the increaseover
timewas partlydue to thehigherpercentage ofmarriedcoupleslivingalone
and partlyto thegreaterpercentage ofhusband-wife-child residentialunits.
Althoughtheproportion ofparent-child unitsvariedoverthe50-yearperiod
coveredbythesedata,in general,and incontrast to theothertypesofsimple
families, thesingle-parent residentialunitswereneverlargepercentages of
thetotals,nordid theychangeovertimeto anymarkeddegree.
In contrastto the increasein simplefamiliesin Los Angelesbetween
1850 and 1900,therewas a markeddeclinein complexfamiliesoverthe
sameperiodoftime,althoughwithsomeinteresting patternswhenwe look
withinthe complexfamilycategory.Complex familiesdeclinedbetween
1850and 1870(from14.2% to 6.4%) but increased(from9.8% to 14.2%)
between1880and 1900.It is likelythatthishad to do withthegreaterkin
availabilityof EuropeanAmericanin-migrants or withthechangingsocial
relationships that made some kin more dependentthantheymighthave
been earlier.(These possibilities willbe consideredin interpreting there-
sultsofthemultivariate analysis.)Decline in the percentage Hispanicsin
of
thepopulationmayalso partlyexplainthedeclinein complexfamiliesover
time.Griswolddel Castillo(1979) reportsthatwhenHispanicwomenmar-
ried,thenewcouplemovedin withthebride'sparents.
Table 3 presentsa cross-tabulation of familytypeby ethnicity forall
fourcensusyears,withfamiliesheadedby EuropeanAmericansbeingdis-
tinguished fromthoseheadedbynon-European Americans.21Between1850

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
14 SocialScienceHistory

Table 3 Familytype(in %) byethnicity


(non-European Americans/
EuropeanAmericans),amongfamilyheads,Los Angeles,
California:
1850,1870,1880,1900a
Inferred
relationships Observedrelationships

Familytype 1850 1870 1880 1900

American
Non-European heads
Simplefamilies 33.9 50.7 48.6 29.3
Complexfamilies 45.9 23.7 21.9 8.0
Simplepluskin 16.2 9.8 9.4 7.1
Simpleplusnon-kin 29.7 13.9 12.5 2.9
Non-kin 20.3 25.6 29.5 60.7
N 192 359 319 140

EuropeanAmerican heads
Simplefamilies 14.5 41.9 53.0 59.4
families
Complex 38.2 23.9 29.0 25.9
Simplepluskin 9.2 4.3 9.9 14.5
Simpleplusnon-kin 29.0 19.6 19.1 11.4
Non-kin 47.4 34.2 18.0 14.6
N 76 561 887 2,402
aSee note5 formeasurement
ofrace/ethnicity.

and 1880thelattercategoryincludedprimarily Hispanics;in 1900thiswas


no longerthecase-by then5.4% ofthepopulationwaseitherAsianorAfri-
can American.Table 3 showsthatin 1850non-EuropeanAmericanfamily
heads,primarily Hispanics,weremorelikelyto liveincomplexunits--espe-
ciallyfamiliesthatincludedkin(comparedto non-kin)--than werefamilies
headedby EuropeanAmericans.Althoughthe overallpercentageof com-
plex familiesdeclinedbetween1850 and 1870 forfamilyunitsheadedby
EuropeanAmericansand bynon-EuropeanAmericans, a largerpercentage
ofcomplexfamiliesthatincludedkinwereheadedbynon-EuropeanAmeri-
cans-primarilypeopleofHispanicdescent--thanEuropeanAmericans. A
different betweenethnicity
relationship and family typeis seenforthe1880-
1900period.In both1880and 1900a smallerproportion offamiliesheaded
bynon-EuropeanAmericansincludedextendedkinthanwas trueforfami-
lies headedbyEuropean-Americans, althoughthedifference is minimalfor

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
inLosAngeles15
Structure
Family

1880.In 1900,theresultsfornon-EuropeanAmericanfamilyheadsreferto
AfricanAmericansand Asians.Whateveritscomposition, however, smaller
proportions of familyunitsheaded by non-EuropeanAmericanswere in
eithertypeofcomplexfamily. Since theywerealso a smallerpercentageof
thesimplefamilies, whathad becomeofthem?Table 3 showsthata much
largerpercentage familyunitsheadedby non-EuropeanAmericanswas
of
notbased on kinshiptiesas definedin thisanalysis.Non-EuropeanAmeri-
can familyheads,it wouldappear,weremuchmoredisadvantaged in terms
oflivingwithkinbytheend ofthecenturythantheyhad beenearlier.
The possibleexplanationsforthe changingdistributions reportedin
2
Table are varied.For one, the changefroman agricultural subsistence
to
economy wagea labor,market economymay have allowed adult children
greatereconomicindependencefromtheirparents,thusreducingcoresi-
dence aftermarriageand increasingthe proportion of husband-wife only
residentialunitsamongboththeyounger andtheolderfamily headunits.For
another,inheritancepatternsmaybe partoftheexplanation. Dahlin (1980)
pointsout thatcoresidencebetween and
parents childrenwas closelylinked
to inheritancepatternsand would occur morefrequently in ruralareasthan
in urbanareasbecauseparentswouldpass farmson to theirchildren.The
urbanization ofLos Angelesmaythusbe implicatedin thechangingdistri-
butions, mightthechanging
as age distributions
offamily heads.(See Table 1
fortheincreaseovertimein thepercentage offamilyheadsin theoldestage
category.)
But if the growthof a marketeconomyhelps explainthe increasein
simplefamilyunits,as varioussociologistshave suggested(Parsons1959;
Burgess1960), whydid the proportionof single-parent residentialunits
not increasealong with the othersimplefamilytypes(see Table 2)? Under
marketconditions, it mighthavebeen harderforsingleparent-childunits
to sustainthemselves -
especiallysincean increasinglylargeproportion of
thesefamiliesovertimewerefemale-headed (59% in 1850,72% in 1870,
80% in 1880,and 82% in 1900).Giventheirabilityto inheritland in their
own names,women,even unmarriedwomen,may have had moreof the
socialauthorityand materialresourcesnecessaryto becomefamilyheadsin
a subsistencethanin a marketeconomy. Womenwithoutsuchresourcesand
authority,however, would be less likelyto be familyheads--and,indeed,

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
16 SocialScience
History

thepossibility of greaterdependenceamongwomenwhowerenotmarried
mightexplain increasebetween1880and 1900in thepercentage
the ofcom-
plexfamiliesthatincludedkin.
Changesin the ethniccompositionof the population,withEuropean
Americansbecomingthemostnumerouspresencefrom1870onward,may
also explainthechangingdistribution of familytypesreportedin Table 2.
As alreadysuggested, thedifferent norms,practices, and experiencesofHis-
panic families--the of
centrality kinship ties in organizing theirsocial re-
lations;thedeclinein theproportion of thepopulationthatwas Hispanic;
and the discriminatory social regimethatnon-EuropeanAmericansfaced
in Californiaonce it was partof the UnitedStates--allmaybe partof an
explanation forthechangingpatterns offamilyrelationsin Los Angelesbe-
tween1850and 1900.Single-parent, female-headed Hispanicfamiliesmay
havebeenespeciallyresponsive to theseconditions, migrating outofthearea
to a moresupportive kinenvironment.
At thesametime,as Table2 also reports, therewas a markeddeclinein
theproportion offamiliesthatwerenotorganizedalongkinshiplinesat all
and in theproportion of householdsextendedwithnon-kinonly.The per-
centageofhouseholdheadslivingin non-kinfamiliesdeclinedsteadilyfrom
28% in 1850to 17% in 1900,butpartofthischangemaybe an artifact ofthe
data sincebefore1880some kincannotbe distinguished fromnon-kin.In
1880and 1900,wheremaritalstatusinformation is available,theproportion
ofunmarried femaleheadsis similarforwomenat thetwotimepoints(79%
in 1880 and 82% in 1900).The percentageof men who wereunmarried,
however, had changed:19% in 1880and 5% in 1900,whichmaypartlyac-
countforthedeclinein thepercentage ofnon-kinfamilies.
For the complexfamiliesthatincludednon-kinonly,thereappearto
be two temporalinflections: a sharpfall(from29.5% to 17.4%) between
1850and 1870,a relatively steadystateoverthenextdecade(whichis when
the shiftfrominferred to observedfamilyrelationships occurs),and then
anotherdecline(from17.3% to 10.9%) between1880and 1900.Whiledif-
ferentprocedureswere used to establishfamilyrelationships and family
structure in 1850and 1870comparedto 1880and 1900,thebiggesttransfor-
mationsoccurredwithin "codingregimes,"notbetweenthem.Normsabout
familyprivacy(Modell and Hareven1973;Laslett1973)mayhelp explain
thesechanges,as mightchangein thepopulationofin-migrants fromsingle

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
inLosAngeles17
Structure
Family

personsto families.The declinein theproportion oftheHispanicpopulation


mayalso havebeen important. In general,then,thereare severalpossible
explanations fortheobservedchangesin family organization: changingeco-
nomic,gender, and race relations;changingpopulation characteristicsand
politicalregimes;lifecoursetransitions; and familynorms.
Next we examinethepossibleexplanations forchangesin nineteenth-
centuryLos Angelesfamilystructurethroughmultivariable analyses;the
resultsof theseanalysesare presentedin the following threetables.Each
category ofthefamilystructure variablehasbeentransformed intoa dummy
variablewiththefamilytypeofinterest compared to all others.Because of
its dichotomousform,we use a logisticregressionstatistical procedurefor
ouranalyses.22
Logisticregressioncoefficients indicatethe log-oddsthata particular
headlivesin a particular of
type family. Since it is cumbersome to interpret
thedatain theseterms,thecoefficients havebeenexponentiated so thatone
maysimply talkabout the odds rather than the log-odds. Coefficientsless
thanI are interpreted as reducingtheodds of a head livingin a particular
familytype, while coefficientsgreaterthan1 increasetheodds.The expo-
nentiatedlogisticregression equationis multiplicative ratherthanadditive.
a
Hence, coefficient of 1 means thatthe independent variable has no effectat
all on thedependentvariable.
The entriesin thetablesweregeneratedby single-year logisticregres-
sionequations.The resultsforeach familytypehavebeen aggregatedinto
singletablesforease of readingresultsacrosstime;thecodingdifferences
betweenthefirstand lasttwosetsofyearsareindicatedin each table.While
we recognizethelimitation thatdataavailability imposeson examining tem-
poral change,we nevertheless thinkit is usefulto comparethe resultsof
theseregression analyses.We discuss theeffects oftheindependent variables
in each equationforeach different yeariftheyreachthe.05 levelofsignifi-
cance.Sincewe haveperformed no testforwhetherthedifferences between
yearsare statistically significant--unwise, in our opinion, in view of the
differences betweenthe measurement of the dependentvariablesbetween
1850-1870and 1880-1900-we makeno claimaboutsignificant differences
betweenyears.Instead,we are attentive to whethera givenstatistical rela-
in
tionshipis significant anygivenequation foreach year.

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
18 SocialScience
History

Simple Families
As we saw in Table 2, thepercentageof simplefamiliesin Los Angelesin-
creasedsteadilybetween1850and 1900.Table4 reportstheeffects offamily
heads'socialcharacteristicson theoddsofheadinga simplefamily(versusall
othertypes).Genderis significantly relatedto headinga simplefamilyunit
(net ofthe other in
variables theequations)onlyin 1900,whenmenweresig-
nificantlymorelikelythanwomentoheadsuchunits.Womenheadedsimple
familiesprimarily whentheyweresingleparents--that is, whentheywere
widowed,nevermarried,or divorced.Is thefinding forgenderreportedin
Table4 reallytheresultofmaritalstatus?23Anddid maritalstatusmakeless
of a difference to womenwhentherewas morehome-basedthanmarket-
basedproduction? Whenmaritalstatusis enteredintoboththe1880and 1900
equations(theonlytwoyearswhenthisinformation is available;thefullre-
sultsarenotpresented here),theeffectofgenderdisappears:themarriedare
significantlymorelikelythanthenonmarried to headsimplefamilies. How-
in
ever, 1880, when marital statusis added to the equation, male heads are
significantlylesslikelythanfemaleheadsto livein simplefamilies;in 1900
thesignificance ofgenderdisappears.It shouldbe keptin mindthatwomen
headsofsimplefamilieswereoverwhelmingly singleparents, and theywere
able to sustainfamilyheadshipin 1880butnotin 1900.Whatevertheexpla-
nationforthehigherprobability thatwomenratherthanmenwouldhead
simple familiesin 1880,net ofthe othervariablesenteredintotheequation,
it no longeroperatedin thesamewayby 1900.In thekindofmarketecon-
omythatexistedin late-nineteenth-century Los Angeles,womenheadsof
single-parent familiesmayhavehad fewermeansthanmento maintainau-
tonomousfamiliesthanhad beenthecase earlier.But,as we shallsee,these
findings do nothold acrossall typesof families.In general,maritalstatus
may have beenmoreconsequential forwomenthanformenas familyheads,
especiallyundermarketconditions, butsincewe currently haveinformation
on it for1880and 1900only,we cannotpursuethisidea further here.
Earlierwe suggestedthatthe expansionof the EuropeanAmerican
populationin nineteenth-century Los Angeleshelpsaccountfortheincrease
in the percentageof simplefamiliesovertime.Table 4 providesevidence
forthis conclusion;people of EuropeanAmericanextractionare signifi-
cantlymorelikelyto head such unitsin 1880and 1900thanin theearlier
two decades.And sincewe knowthatEuropeanAmericanswerethe ma-

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
inLos Angeles 19
FamilyStructure

Table 4 on headinga
Effectsof heads' social backgroundcharacteristics
"simple" family(versusall othertypes)in Los Angeles,California:1850,
1870, 1880,and 1900 (logisticregression)a
Inferred
relationships Observedrelationships
variablesb
Independent 1850 1870 1880 1900

Intercept .10* 1.33 .56 .17**


Male [Female] 1.12 .60 1.06 2.45***
EuropeanAmerican
ethnicity
[Non-European
American] 1.68 .89 2.04*** 4.31***
Bourgeoisoccupation
[Semi-unskilled] .64 .60** .49*** .68***
Middle-level
occupation
[Semi-unskilled] .34 .47** .47*** .74*
Skilledoccupation
[Semi-unskilled] .48 .95 .87 1.25
Out ofthelaborforce
[Semi-unskilled] .40 .43 .59 1.04
Middleage (34-53)
[Young18-33] 2.05* 1.88*** 1.64*** 1.36**
Old age (54 andup)
[Young18-33] 1.67 1.37 1.37 .85
California-born
[OtherU.S.] 5.10 .92 1.86* 1.00
Mexican-born
[OtherU.S.] 5.19 1.37 2.67** .38*
Foreign-born
[OtherU.S.] 1.08 .85 1.09 .94
N 268 920 1,206 2,542
aAdjustedodds ratiosare reportedhere.
bThe missingcategoryis indicatedin [squarebrackets].
*p < .05; **p < .01; ***p < .001.

jorityoftheLos Angelespopulation--morethan60% in 1870,almost80%


on familystructure
in 1880,and over90% in 1900--theimpactofethnicity
is veryimportant.The overwhelming presence EuropeanAmericansin
of
be a majordetermi-
Los Angeleswould,inevitably,
late-nineteenth-century
nantoffamilystructureoverall.

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
20 SocialScience
History

Anotherelementof the associationbetweenethnicity and headinga


simplefamilyunit,however, should also be considered--onethathas to do
withchangingrelationships betweenethnicity and simplefamilyliving.In
1850and 1870theethnicity ofthefamilyhead is notsignificantly relatedto
headingsimplefamilies;in 1880and 1900,in contrast, ethnicity is signifi-
cantly related:heads of European American descent are significantly more
likelyto headsimplefamiliesthanarethoseofnon-EuropeanAmericande-
scent.Howeverdifferent the particular circumstances thatpeople of color
facedinlate-nineteenth-century Los Angeles- thedeclining populationand
opportunities of Mexican the
Americans, legal and historical restrictions on
AfricanAmericans and Asians--their experiences did notfacilitate livingin
simple families.24
In 1850,occupationdoes not appearto have had a significant impact
on headinga simplefamily, but by 1870thiswas no longerthecase.While
theseresultsmaybe an artifact of thesmallersamplesize in 1850thanin
any of the following years,a substantive interpretation is also possible.In a
self-sufficient
economy, thevariousresourcesassociatedwithdifferent occu-
pationsmay be less for
important simplefamilyliving than in an economy
organized more along market lines. In 1870, when the occupation of the
head does makea significant difference, net of the effects of the other vari-
ables includedin theequations,familyheads in midleveland professional
occupationsaresignificantly lesslikelythanheadsin thesemi-and unskilled
categoryto head simplefamilies.In nineteenth-century Los Angeles,the
of
practice living in nuclear families may have emerged firstin the lower
statusoccupationalgroups.Familyheads in skilledoccupationswerenot
significantlymoreor less likelyto livein simplefamiliesthanwerethosein
semi-orunskilledoccupations, whilefamily headsin thehigherstatusoccu-
pations were less
significantly likely to do so. As wewillsee inTable 5, heads
in theseoccupationsaresignificantly morelikelytoheadcomplexfamilies -

quite a differentresultthan the Parsonian model would lead us to expect.


Earlierdebatesabout historicalchangesin familystructure have sug-
gestedthatstagein thefamilylifecycle(as measuredbytheage ofthefamily
head) is relevantto theirdistribution (Berkner1972).Our resultssuggest
thisto be the case fornineteenth-century Los Angelesas well,withthose
headsin themiddleages (34-53) beingsignificantly morelikelythanolder
or youngerheadsto livein simplefamilies.In 1880and 1900,however, this

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
FamilyStructure
in Los Angeles 21

Table 5 Effects ofheads'socialbackground characteristics


on headinga
complexfamily (versusall othertypes)in Los Angeles,
California:
1850,
1870,1880,and 1900(logisticregression)a
Inferred
relationships Observedrelationships
Independent
variablesb 1850 1870 1880 1900

Intercept .58 .06** .07*** .08**


Male [Female] 1.36 1.53 2.21** 1.29
EuropeanAmerican
Ethnicity
[Non-European
American] .60 1.37 1.60 2.60"**
Bourgeois
[Semi-unskilled] 1.90 3.73*** 2.70*** 1.95"**
Middle-level
occupation
[Semi-unskilled] 1.87 2.00* 1.90"* 1.61**
Skilledoccupation
[Semi-unskilled] 1.38 .94 1.17 1.14
Outofthelaborforce
[Semi-unskilled] 1.70 1.35 2.80** 1.35
Middleage (34-53)
[Young18-33] 1.20 1.37 1.56** 1.01
Oldage(54andup)
[Young18-33] 1.34 1.52 1.47 1.18
California-born
[OtherU.S.] .78 2.63* 1.32 1.01
Mexican-born
[OtherU.S.] .52 1.76 1.00 1.01
Foreign-born
[OtherU.S.] .48 .96 .56*** .72**
N 268 920 1,206 2,542
aAdjustedodds ratiosare reportedhere.
bThe missingcategoryis indicatedin [squarebrackets].
*P < .05;**p< .01;***p< .001.

relationshipdisappearswhenmaritalstatusis enteredintotheequation(not
shown);thustheapparenteffects ofage aredue to thehead'smaritalstatus.
(It is unfortunate
thatwe do not havea measureof maritalstatusfor1850
and 1870,whichwouldmakeit possibleto examinechangeovera period
of moredramaticsocialchanges.)Althoughthelife-cyclepatternoffamily

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
22 SocialScience
History

headshipincreasesthe odds of headinga simplefamily(versusanyother


type)whenmaritalstatusis notin theequationin each ofthecensusyears,
itcannotaccountforthechangein theincreasedpercentage ofsimplefami-
lies between1850and 1900sincetheproportion ofmiddle-ageheadsin the
overtime.
sampledid notincreasesubstantially

Complex Families
Table 5 allowsus to lookat therelationship ofeachoftheindependent vari-
ablesto headinga complexfamilyunit.In 1850onlyone oftheindependent
variablesis significantly relatedto headinga complexfamily:beingin the
highest statusoccupationalgroup.When real estateownershipis entered
intothe equation,however,it becomesclearthatproperty ownership, not
occupationaltitle, is the relevant determinant (notshown).By 1870 the re-
lationshipbetweena head's property ownershipand complexfamilyliving
is no longersignificant and it appearsthatthesystem ofeconomicrelations
ratherthansimplyone'splacein thedivisionoflaboris theimportant factor.
In a self-sufficient,
agricultural economy, it is landownership thatcounts!
Men headsweresignificantly morelikelyto livein complexfamiliesin
1880thanwerewomenheads,butthisresultdisappearswhenmaritalstatus
is enteredintotheequation.It is not"gender"perse thatmakesa difference,
but genderrelationsas theywereshapedby maritalstatus.25 Genderdoes
notappearto havehad an impacton headinga complexfamily, netofother
head characteristics in 1900,but home owners(in comparisonto renters)
weresignificantly morelikelyto headsuchfamiliesin thatyear(notshown).
This suggeststhathomeownershipin 1900mayhavehad the same func-
tionalrelationship to headinga complexfamilyin thelatertimeperiodthat
landownership did in 1850- thatis,bothreflected wealth,and bothsignifi-
cantlyincreasedtheodds ofheadingsimplefamiliesthatincludedextended
kinand/ornon-kin.
Debates in historical familyresearchareperhapsmostdevelopedin re-
lationto the questionof extendedfamilylivingin past times.Anderson
(1971)seesextendedfamily livingas a resultofneed,whereasRuggles(1987)
sees it as the resultof affluence. To the extentthatour findings can con-
tributeto thatdebate,it appearsthattheseoutcomesare notcontradictory,
at leastas theyare touchedby genderrelations:Womenheadsmayindeed

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
Structure
Family inLosAngeles23

needextendedkinto be able to maintainautonomoushouseholds,and men


heads,who are likelyto be moreaffluent, maywantto have extendedkin
livingwiththem.These genderdifferences in familyrelationships mayhelp
explain what at firstglancewould appear to be contradictory results.
Our findings fortheeffects ofoccupationon headingextendedfamilies
do supportRuggles's(1987)positionthathigherstatus,and thewealththat
is presumably associatedwithit,is associatedwithfamilyextension. Table 5
showsthat,at leastfrom1870through 1900, heads in all white-collar
occupa-
tionsweresignificantly morelikelythanheadsin semi-and unskilledoccu-
pationsto live in such families,and the finding holdswhenmaritalstatus
and property ownershipare enteredintotherelevantequations.The risein
the proportion of heads in theseoccupations,especiallythosein midlevel
white-collar ones,mayalso helpexplainthechangingdistribution offamily
in
types general. We also findthat in 1900 those of European American de-
scentare significantly morelikelyto head complexfamilies,net of all the
othervariablesin theequations,and thesefindings holdwhenmaritalstatus
is controlled.It wouldappearthatbytheendofthecentury, whatever advan-
tagemay have remained,especially to Californio heads, untilthattime (see
Griswolddel Castillo1979),withthe creationof an Anglo-American city,
familyheads of colorweresignificantly disadvantaged in relationto living
in complexfamilies.Migrationis also relatedto headingcomplexfamilies,
althoughnotconsistently acrosstime.Foreign-born headsweresignificantly
less likelythanheadsbornin theUnitedStatesoutsideofCalifornia to head
complex families in 1880 and 1900, and the relationships remain significant
whenmaritalstatusis enteredintothe two laterequations.This provides
further supportfortheideathatforeign migration haddifferent implications
forkinavailability thandid migration fromwithintheUnitedStates.

Non-kinFamilies

Table 6 showsthat,netof theothervariablesin the equation,men family


lesslikelytoheadnon-kinfamiliesthanwerewomen
headsweresignificantly
familyheadsin 1880and 1900:thereis no significant between
relationship
and
gender heading thiskind offamilyin thetwoearlier
census years.Marital
statusappearstoaccountforthisdifference:Whenenteredintothe1880and
1900equations,it is highlysignificant
in itsown right(withnever-married,

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
24 SocialScienceHistory

Table 6 Effectsof heads' social backgroundcharacteristics


on headinga
non-kinfamily (versusall othertypes)in Los Angeles,
structure California:
1850,1870,1880,and 1900(logistic
regression)a
Inferred
relationships Observedrelationships
b
variables
Independent 1850 1870 1880 1900

Intercept 1.84 .78 2.34* 9.20***


Male [Female] .45 1.30 .41** .19"**
EuropeanAmerican
ethnicity
[Non-European
American] 1.17 1.01 .29"** .09"**
Bourgeoisoccupation
[Semi-unskilled] .81 .48*** .81 .73
Middle-level
occupation
[Semi-unskilled] 1.38 1.33 1.51 .88
Skilledoccupation
[Semi-unskilled] 1.51 1.06 1.04 .49**
Outofthelaborforce
[Semi-unskilled] 1.45 2.24 .77 .66"
Middleage(34-53)
[Young18-33] .40** .38*** .32*** .53***
Oldage(54andup)
[Young18-33] .42 .50** .48*** 1.03
California-born
[OtherU.S.] .24* .50 .43** 1.01
Mexican-born
[Other
U.S.] .50 .44* .34** 3.67**
Foreign-born
[Other
U.S.] 1.90 1.21 1.86*** 1.84"**
N 268 920 1,206 2,542
aAdjustedodds ratiosare reportedhere.
bThe missingcategoryis indicatedin [squarebrackets].
*p < .05; **p < .01; ***p < .001.

widowed,and divorcedheadsmorelikelyto headnon-kinfamilies;datanot


presented)and changestherelationship betweengenderand headingnon-
kinfamilies.
There are no significant differences
betweenmale and female
familyheadsin relationto livingin non-kinfamilies.
Here again,itis unfor-
tunatethatwe do nothaveinformation on maritalstatusfor1850and 1870.

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
in Los Angeles 25
FamilyStructure

In addition,in non-kinfamiliesas in simplefamilies, changesin thesys-


temof productionand theexpandedmarketdemandforfemalelabormay
haveincreasedtheabilityof(nonmarried) womenheadsoffamiliesto livein
separate residences more than men. Whereas womenand menfamilyheads
had been significantly different fromeach otherin 1880,thiswas no longer
the case in 1900. (This finding shows up onlywhenmaritalstatusis addedto
theequations.)In 1900,itwouldappear,womenhad no lessaccessthanmen
to resourcesthatallowedthemto head non-kinfamilies.Of course,census
data cannottell us whetherlivingoutsideof maritaland parent-child kin
tieswas an advantageor a disadvantage forwomencomparedto men,but
thelackofsignificant difference betweenwomenand menheadsin 1900but
not 1880in relationto non-kinfamiliessuggeststhatgenderrelationshad
changedconsiderably overthe last20 yearsof the nineteenth centuryand
are relevantto explaining in
changes family structure.
Ethnicity is as significantlyrelatedto headingnon-kinfamiliesin 1880
and 1900 as it was to headingsimplefamilies:inbothyearsEuropeanAmeri-
cans weresignificantly less likelyto head theseunitsthanwerethe His-
panics,Asians, and African Americanswho made up the "non-European
American"group.(This remainstruewhenmaritalstatusis enteredintothe
equation.)If theEuropeanAmericanswereable to achievelivingwithkin
linkedbymaritaland parent-child ties,as ourresultsin Tables4 and 5 sug-
gest, then the greater likelihood ofheadingnon-kinfamiliesbyotherethnic
groupsmaybe seen as a consequenceof theirdisadvantages. Again,it ap-
pears that all peoples of color shared these disadvantagessince the results
are the same when Hispanicswere the largestproportionof the "other"
ethnicgroup(in 1880)or whenthe"other"ethnicgroupwas composedof
AsiansandAfricanAmericans(in 1900).It also makesevenmoreremarkable
theaccomplishments ofindividualslikeBiddyMason,an AfricanAmerican
matriarch in nineteenth-century Los Angeles,whowasableto sustaina vital
life this
family throughout period(Hayden1989).
The findings foroccupationare considerably less regularin relationto
non-kinthanto simplefamilies;itdoesnotappeartohavebeenoccupational
activitythatwas particularly predictiveof headingthiskind of domestic
unit.It maybe thatheadinga non-kinfamilywas a "strategy"forhigher
andlowerstatusheads,althoughfordifferent reasons.For groupswithlower
resources,heading a non-kin familymay have been a negative"choice"--

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
26 SocialScience
History

one thatcircumstances mandated - whereashigherstatusfamily headsmay


havechosento livein non-kinfamiliesformorepositivereasons,perhapsas
a meansofaccumulating resources(Newmark1970).There also appearsto
havebeen a significant and stablelife-cycle
patternbetween1850and 1900,
withheadsin themiddleage groupsignificantly less likelythanheadsin the
youngestgroup to live in non-kin families.
When maritalstatusis included
in theequationsfor1880and 1900,however, only1880remainssignificant.
Table 6 also showsthatheadsbornin Californiaweresignificantly less
likelyto livein non-kinfamiliesin 1850and 1880.26 In addition,migration,
especiallymigration from foreigncountries,when comparedto migration
fromelsewhereintheUnitedStates,wasstrongly relatedtoheadingnon-kin
familiesat theend ofthecentury, whichwouldbe expectedifthe internal
migrants camein familieswhiletheforeign migrants did not.The negative
and significanteffect ofbeingMexican-born in 1870and 1880is perhapsdue
to thepresenceofotherMexicansin thecity.But as theproportion ofHis-
panics declined (by1900), thisresourceceased to be availableto them.Given
thesharpdecreasein theMexican-born Los Angelespopulation, thisdemo-
in
graphicchangemay partexplain change the in the proportion ofnon-kin
familiesin thecityas a whole.
Taken together, thesefindings suggestthatbothchangingpopulation
characteristicsand changingsocial relationships help explainthe decline
in the proportionof non-kinfamiliesin Los Angelesbetween1850 and
1900.Race and ethnicrelationsand therisein theproportion ofEuropean-
Americanfamily headsbothseemto helpaccountforthedeclinein thepro-
portionofnon-kinfamiliesin latenineteenth-century Los Angeles.Migra-
tionalso has significant effectsbut,unfortunately, notknowwhether
we do
the migration patternsconsistedof familiesor individuals,sincethemea-
sureis basedon birthplace.27

Summaryand Conclusion
Discussionofhistoricalchangesin Los Angelesfamilystructurein thesec-
ond halfof the nineteenth centurycan be separatedintotwo parts-one
(How
descriptive did thedistribution
offamilytypeschangeovertime?),the
otheranalytic(Whydid family changeas itdid?).We havetried
organization
to addressbothquestionsin theanalysespresentedhere,and we thinkit is

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
inLos Angeles27
Structure
Family

important to pointto thecomplexity ofthefindings we havereported.This


is truenotonlyforfindings aboutthedistribution offamilytypesovertime
but also in our multivariate analyses.In them,all of theindependent vari-
ablesthatreportcharacteristics ofthefamilyhead- gender,ethnicity/race,
occupation,age, and birthplace(and,whereavailable,property ownership
and maritalstatus)- had significant on
impacts family structure
of one type
oranother, in one timeperiodoranother.Designingtheanalysisas we have,
we could be attentive to historical
detail,particularly to theexperiencesof
social groupswho haveall too oftenbeen "hiddenfromhistory." The re-
sultsofdoingthisare gratifying. It is clearthatin Los Angeles,California,
overthelastfifty yearsofthenineteenth century, theexperiencesofdiffer-
entrace,class,and gendergroupswerequitevaried.The challengenowis to
formulate a theoretical explanationthatcan accountforthisdiversity. While
evidenceforeach ofourinterpretations is limitedbythedataused here,our
analysesdo provideplausibleexplanations and areaswherefurther research
is warranted.
In general,we thinkthatchangesin theorganization oftheeconomy-
fromdomesticto marketproduction--had an effecton therelationshipbe-
tweengenderand familystructure. When land was the basis of domestic
production, both women and men who werelandowners couldsupportfami-
lies. As wage laborand marketproductionexpanded,domesticproduction
declined,and somewomenwereadvantagedand someweredisadvantaged,
withmaritalstatusbeingof greaterconsequenceforthe typeof familya
woman(comparedto a man) headed.When land was centralto self-and
familymaintenance, Los Angeleswomenwho had access to it could both
commandthe loyaltyand supportof familymembers,or theycould hire
non-kinto workwiththemon theland. In a wage-labormarketeconomy,
marriedwomenweremoredisadvantaged: Littlewageworkwas availableto
them,so theyweremoreeconomically dependenton husbands,sons,and
brothers. Undertheseconditions, being singleparentwas morelikelyto be
a
associatedwitheconomicneed thanit wouldhavebeen earlier.Unmarried
women(divorced,widow,and never-married), on theotherhand,had more
waged workavailableto them,as well as a culturethatlegitimated their
greater involvement in the publicsphere. As a consequence,theymayhave
beenbetterable to sustainthemselves moreindependent offamilytiesthan
been thecase. In a studyofdivorceand women'sstatusbe-
had previously

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
28 SocialScience
History

tween1850and 1890in California, Griswold(1980)arguesthatwomenwere


notnearlyas bereftofresourceswhentheyweredivorcedas theireconomic
statusalonewouldlead one to suppose.Our findings suggestthatthesere-
sourceschangedovertime,althoughwithdifferent implicationsforwomen
familyheadsdependingon maritalstatus.Widowswhoweresingleparents
maywellhavebeenin a moredifficult positionat theendofthecentury than
theyhad been earlier.For divorcedwomenwhodid nothead single-parent
families, however, thepatternofsomeindependencethatGriswoldreports
is quiteconsistent withthefindings presentedhere,especiallyforthelater
decadesofthecentury.
In termsof race/ethnicity, we also see a patternof historicalvariation
in the significant findings. multivariate
The analysisshowsan increasein
thesignificance ofbeinga non-EuropeanAmericanforheadingall residen-
tialgroups--kinand non-kin--over time.But whereastherewas,arguably,
someadvantageto womenwholivedoutsideoffamilyunitsin relationto a
decreasein patriarchal controlto whichtheyweresubjected,itis difficult to
see anysuch advantageforthenon-Europeanfamilyheads.The race rela-
tionsthatdisadvantaged familyheadsof color--including thosethatled to
thedeclinein theMexican-heritage presence in Los Angeles- wereincreas-
ingly institutionalizedand applied to Asian and AfricanAmericanas wellas
Hispanicfamily heads. a
Being non-European head,quitesimply, disadvan-
taged such persons from establishing in
kin-basedfamilies Los Angelesatall.
Thus, it appearsthateconomicchangeaffected genderand race rela-
tions, but in somewhat differentways. In relation to gender,beingmarried
mayhaveimposedgreaterconstraints on womenthanhad previously been
the case; it decreasedthe capacityof womento head kin-basedfamilies.
At the same time,theirpossibility of headingnon-kinfamiliesincreased,
allowing them and
greaterautonomy freedomfromthepatriarchal relations
thatso shapedwomen'sand men'slivesduringthistimeperiod.Women's
increasinginvolvement overtimein publicactivitieswouldalso havesup-
porteda cultureof women'semployment, even whenwagedworkwas not
requiredby economic necessity (see Laslett1990).In contrast, thedisadvan-
tagesforfamilyheadsof colorwereunrelenting overtheperioddiscussed
here.Althoughtheremaybe different explanationsforwhythiswas true
forHispanic,Asian,and AfricanAmericanfamilyheads,the last 50 years
of thenineteenth centurywerenotpropitiousfortheirfamilylives.Living

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
inLosAngeles29
Structure
Family

as headsof non-kinfamilyunitswas notnecessarily a matterof choiceand


increasing as
opportunities, mayit have been for some EuropeanAmerican
women,but ratherreflected thelimitsthatracismimposedon them,both
and
institutionally personally.
For all thelimitations of ourone constantmeasureofeconomicstand-
ing,head's occupation, resultsfortheeffects
the on familystructure areex-
tremelyinteresting. In general,we see a "class" patternof familystructure
emerging almost immediately aftertheAmericantakeoverofCalifornia.(It
is important to recallthattheresultsto whichwe are referring hereare net
oftheeffects oftheothervariablesin theequations.)Familyheadsin white-
collaroccupations -whether of highor low status- weresignificantly less
likely to head simple families and more
significantly likely to head complex
families. Ruggles(1987)was rightthatitwas onlythewealthywhocouldaf-
fordextendedfamilyliving.But it also appearsthatthe "private"nuclear
familywas the emergingpatternforfamilylivingin nineteenth-century
California, althoughthosein lowerratherthanhigherstatusoccupationsap-
pearedto havefirstadoptedthisfamilyform.
Race,class,and genderrelations, then,areall partofthestoryofchanges
in familystructurein nineteenth-century Los Angeles.In contrast,the
effectsof lifecoursevariationon familystructure wererelativelyinvariant
overtime.Headshipofa residential unitbeganin non-kinfamilies, regard-
less ofclass,although,we suggestedearlier,it mayhavebeenmore"strate-
gic" amonghigherstatusfamilyheadsand theresultofgreaterpoverty and
controloverimmigration among lower resource In
groups. contrast, simple
familyliving was characteristic of all thoseheads in the middle age group,
whereasbeingin theoldestage groupdid nothavemucheffecton thetype
offamilyheaded.
The impactofbirthplace, ormigration, is,as we havealreadydiscussed,
difficult to disentangle fromtheeffects ofethnicity,and in part,thesome-
whatrandompatternofresultsmaybe due tothisfact.To theextentthatany
consistent configuration can be observed,it relatesmoreto non-kinfamily
living than to kin-based families.Heads withgreaterkinavailability were
morelikelyto headthelattertypeoffamily;thosewithlesserkinavailability
weremorelikelyto head theformer type.In comparisonto migrants from
otherpartsoftheUnitedStates,headswhohad migrated fromEuropewere
less likelyto have kinavailablewithwhomto formcomplexfamilies.For

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
30 SocialScience
History

theMexican-born group,thepotentialforkinavailability declinedthrough-


out theperiodexaminedherebut especiallyin 1900.Kin availability, then,
appearsto helpexplainfamilystructure, and themigration histories ofdif-
ferent socialgroupsstructured thatavailability.
In general,the findings we have reportedheresuggesttheneed to be
attentiveto thehistoricallyspecificcircumstances oftimeand placeas they
affectfamilyforms.But it is notsufficient to stopwitha call formorehis-
in historical
toricalspecificity studiesofthefamily. The theoreticalapproach
we haveused to interpret our findings-onethatis attentive to bothsocial
structure and humanagency-suggests a broaderconclusion:We thinkthat
theemphasison historical specificity is requiredformoregeneraltheorizing.
Social structuresetstheparameters withinwhichhumanactorsmaketheir
choices,although,as we arguedearlier,it does notdetermine them.But all
socialactorsarenotequal in termsofeitherchoicesorresources. Hereagain,
historicalspecificity
is ofgreatimportance. The Hispanic,Asian,and Afri-
can Americanresidentsof post-1850Los Angeleswerenotequallyplaced
to makethesamechoicesthattheincomingEuropeanAmericanswereable
to make,particularly thoseEuropeanAmericanswho broughtwiththem
resourcesof skill,money,and kin thatfacilitated participationin the ex-
pandingeconomic,social,and culturallifeofthecity.Considering changes
in Los Angelesfamilystructure in thesecondhalfofthenineteenth century
as intersections
ofsocialstructure and humanagencyhas beenuseful.It is
fromthisperspective thatthefindings reportedin thisarticlehavea theo-
reticalcontributionto maketo historical sociologyand socialhistory, as well
as an empiricalcontribution to thestudyoffamilylifein pasttimes.

Notes

Weacknowledge withthanks support fromtheNationalScienceFoundation (NSF/SES-


8910522)and theGraduateSchoolat theUniversity ofMinnesota. We also thankKarl
Krohnforhisinvaluable assistance throughouttheanalyticstagesofourworkandJohn
ModellandM. J.Maynesforveryhelpfulreadings ofan earlierdraftofthisarticle.
1 Whilewe are interested in thesebroadtheoreticalquestions,we are awareofthe
ofanyone typeofdatain relation
limitations to them.Censusdata,on whichthe
empiricalanalysisto followis based,areusefultodescribesomematerial resources
and populationcharacteristics.These descriptions--for
instance,offamily organi-
zation-can alsobe viewedas theoutcomesofchoices,butthedatado notprovide

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
FamilyStructure
in Los Angeles 31

information aboutintentions andmeanings; autobiographies, oralhistories,andper-


sonalnarratives are bettersuitedto thatpurpose.(See, forinstance, Weber1989;
Padilla1993;PersonalNarratives Group1989;Laslett1991;Deutsch1994;Maynes
1995.)Nevertheless, censusdataare usefulforreconstructing theorganization of
Americanfamilylifeand forsuggesting questionsaboutmeaningand intentions
thatneedtobe answered withothersources.
2 Laslett(1980)discussesthedistinction betweenchanging socialrelationships and
changesin population. This logicis alsoappliedhere.It shouldbe noted,however,
thatwhilethisdistinction canbe madequiteclearlyin statistical analysis,inactual
historicalsituations, changing population distributions andchanging socialrelation-
shipsarelikelytobe closelyconnected.
3 Abrams(1981)makesthecaseforviewing historically specificoutcomes as intersec-
tionsofsocialstructure andhumanagency. We followthisperspective ininterpret-
ingtheresultsofthisanalysis.
4 It is impossibleto reviewfullythisliterature here,evenforjusttheUnitedStates.
For someimportant examplesofworkin thisfield,see Welter1978;Bloch 1978;
Cott1977;Laslett1973;Modelland Hareven1973;Yans-McLaughlin 1977;Ryan
1981;Smith-Rosenberg 1975,1985;Degler1980;Hareven1982;MorgenandGolden
1987.Foran analytic overview ofthisliterature see BrennerandLaslett1986;Las-
lettandBrenner 1989;fora criticaloverview see Kerber1988.
5 Pitt(1970)contrasts thispatternin California witha quitedifferent one in New
Mexico,wheretheHispanicpopulationwas able to retainconsiderable political
power.He alsomakesclearhowmuchclassdifferences withintheCalifornio popu-
lationrelatedtotheseprocesses.
6 The citypopulation wasjustover1,600in 1850andjustunder6,000in 1870.
7 Although raceis reported on theindividual federal censusschedules, beingofHis-
panicdescentwas notone oftherecognized racialcategories (White,Black,Chi-
nese,Japanese, Indian,and Mulatto).SincetheSpanishsurnamepopulation is so
important forunderstanding Californiahistory, surname, placeofbirth,and,from
1870onward,placeofparents'birthwereusedto code a personas Hispanic.For
1850through 1880,codingwasdonebyhandanddataentry wasa separatemanual
process.The dataentryprocesswasdifferent for1900,however. Information from
thecensusscheduleswas directly enteredintomachine-readable formwithout an
intermediate codingprocedure, and theracialcategories used bythecensuswere
theonlyonesentered. Thus,information on theHispanicpopulation is notavail-
ableforthatyear,a seriousdisadvantage giventheimportance ofrace/ethnicity for
thestorybeingtoldhere.Butevenifonetreatsall Mexican-born Los Angelesresi-
dentsas Hispanic(11.9%ofthesampleofindividuals, 0.8% ofthesampleoffamily
heads)and 20% oftheCalifornia-born population as Hispanic(approximately the
samepercentage as in 1880),theestimated proportion ofHispanicsfor1900would
stillbe lessthan3% oftheentirepopulation. It is importanttoconsider thesenum-
berswithcaution.In theearlyyears,especially, thereis goodreasonto expectthat

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
32 SocialScienceHistory

theHispanicpopulation wasundercounted; censusenumerators oftendidnotspeak


SpanishanddidnotvisittheMexicanbarrios(see Griswold delCastillo1979:178).
Ifthatwasthecase,however, itwouldmakethedeclinein theHispanicpopulation
evenmorestriking.
8 Migration is measured usingbirthplace information foreachpersonlisted,butthis
technique undercounts peoplewhomayhavemigrated twoormoretimesafter birth.
9 The percentage figures forindividualsbornelsewhere in theUnitedStates,outside
ofCalifornia, were15.7%in 1850,27.3% in 1870,and35% in 1880.
10 As willbe discussedlaterin thetext,differences in thesedataovertimealso affect
ouranalyses.
11 Fire destroyed theindividualcensusschedulesfor1890.Although datafor1860
werecollected, preliminary analysissuggested theywereunreliable, so weexcluded
themfromour analysis.Samplesweredrawnfromthecensus-defined dwelling
units;thesedatawerealsothebasisforthestudyreported inLaslett1975and1977.
Griswold del Castillo(1979:183-84)provides an independent testofthereliability
oftheseanalyses. In thisarticle,theunitis thecensus-defined family. (See U.S. De-
partment ofCommerce, BureauoftheCensus,1978,fortheoperative definitionsof
dwelling unitsandfamilies.) Although thewording variessomewhat acrossyears, in
general,a dwelling unitwasdefined as a houseorabodestanding aloneorseparated
bywallsandsharing a commonroof.A family wasdefined as a unitlivingtogether
andprovided forin common.Morethanonefamily couldlivein a singledwelling
unit,andmorethanonedwelling unitcouldexistata singleaddress.In 1900samples
weredrawnfromtheareawithinthecitylimitsandfroma carefully chosengroup
of contiguous suburbs.Although locationwas enteredintoeach of the"unique"
equations (seenote12)for1900,itshowednostatistically significant effectsinanyof
them,so wecombined thetwosamples --cityandsuburbs--for ourfinalanalysis.
Therehasbeendebatein theliterature on theappropriate unitofanalysisfor
historicalresearchon families: Shouldit be family members treatedas individu-
als,or shoulditbe thecollective unit?Familystructure, thesubjectofthispaper,
is,obviously, a collective family and we followthepracticeofusing
characteristic,
family headinformation tocharacterize theunitas a whole.Criticisms ofthisprac-
ticehavebeen lodgedfroma variety of perspectives--see, forinstance, Berkner
1972;Ruggles1987;Alter1988.In theanalysispresented here,we use family head
datasinceourinterest is morefocusedon thefamily as a socialinstitution thanon
theindividual lifecourseor experience. Nevertheless, usinga family-level unitof
analysisandheadcharacteristics todescribe wholefamilies haslimitations thatneed
tobe keptinmind.
12 We havealsoexamined "unique"equationsthatincludevariables availableforpar-
ticularbutnotall censusyears:in 1850,thevalueoftherealestateownedandour
confidence in thefamily structurevariable(see Laslett1975,1977);in 1870,these
twovariables plus the value of thepersonalproperty owned;in 1880,marital status;
in 1900,maritalstatus,whether thehomewas ownedor rented,and whether the

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
in Los Angeles 33
FamilyStructure

family livedwithinthecitylimitsor beyondthem.Space limitations prohibitour


presenting fullytheresultsofeachuniqueanalysis; wewillrefer tothemas theyare
relevant tounderstanding theresultsonwhichwe do focus.
13 The samplesofindividuals livingin Los Angeles(tablenotpresented) showthat
Hispanicsdeclinedfrom71% in 1850to 40.6% in 1870andto 21.5% in 1880.For
Asians,thefigures are0.3% in 1850,2.4% in 1870,8.6% in 1880,and3.4% in 1900.
For African Americans, thepercentages are0.7% in 1850,1.0% in 1870,0.8% in
1880,and2.0% in 1900.
14 Forpossibleerrors in agereporting, seeMasonandCope 1987.
15 In viewofthemassivechangesin theLos Angeleseconomy between1850and1900,
therelative stability intheoccupations reported The absenceofin-
is quitestriking.
formation on themeanings oftheseoccupational instance,
titles--for whether the
respondent wasself-employed or salaried-makesit difficult to interpretthefind-
ingsassociated withoccupational changes.Although thesecondary literature
makes
clearthatduringtheperiodcoveredbythisstudytheclassstructure inLos Angeles
changedsubstantially, theambiguity ofmeaning oftheavailableoccupational titles
affectsourcapacity totalkinmuchempirical detailaboutthesechanges.
16 The family structure variable usesa modified version ofP. LaslettandWall's(1972)
classification.
17 The percentage ofcomplexfamilies withextended kinthatalsoincludednon-kin is
presented inTable2, footnote a.
18 The percentages ofnon-kinfamilies thatincludedsuchkinwere17.3% in 1850,
13.7%in 1870,10.2%in 1880,and17.2%in 1900.
19 Thesepossibilities weretakenintoaccountin codingfamily relationshipsanda re-
code-that is,agreement
liability betweentwoindependent coders--wasattached
to eachinferred relationship to thefamily head(see Laslett1975,1977).The infer-
enceprocedure usedto code family relationships for1850and 1870wasbasedon
a setofuniform codingrulesthatusedtheinstructions issuedto censusenumera-
tors(see UnitedStatesDepartment of Commerce, Bureauof theCensus,1978).
Reliabilitycodeswereentered intothe"unique"multivariate equations,mostly un-
reported here,thatincludedinformation availableforsomebutnotall years.
20 On thewhole,Table1 showsthatmajorchanges, whentheyoccurred, werewithin
"codingregimes" rather thanbetweenthem.
21 We are uncomfortable withaggregating all households headedby peopleofcolor
intoa singlecategory since,as discussedin thetext,theirexperiences andlivingar-
rangements varied.Unfortunately, changing samplecharacteristics overtimemade
itnecessary todo so in thestatistical analyses.
22 Earlierresearchon thesenineteenth-century Los Angelescensusdataused vari-
of
ants OLS statisticalprocedures (see Laslett 1975,1977,1990;Masonetal. 1987).
In thisanalysis,however, we use a logisticregression model:a multivariate ana-
lytictechnique similar in interpretation to OLS regression but more appropriate
forcategoric dependentvariables(see Aldrichand Nelson1984).The decisionto

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
34 SocialScienceHistory

usethistechnique wasmadeafterhaving analyzedthedatawitha variety ofstatisti-


caltechniques, including simple OLS regression and LIMDEP- the lattera statistical
program that allows one to a
analyze multicategory nominal dependentvariable
in a singleequation(Greene1989).Sinceeachoftheseprocedures yieldedsimilar
results,we decided thata logistic approach had the most advantages fewest
and dis-
advantages for our purposes: it was more appropriate for a dichotomous nominal
dependent variable while,at thesametime,morefamiliar to mostaudiencesinter-
estedinthisanalysis.
23 Maritalstatuswas analyzedas a three-category dummyvariable:married, never
married, widowed/divorced. (The Ns for the divorced group were very small in
bothyearsthattheinformation wasavailable.) In theregression equation, "married"
wasthemissing category andnever-married andwidowed/divorced headsweresig-
less
nificantly likely than married heads to live in simple families.
24 Variability inthecomposition ofthenon-European American ethnicgroupin 1880
meansthatthereis notas closea correlation between and
ethnicity birthplace inthat
year as in the others. By 1900,however, all
virtually non-Anglos were also migrants.
In 1900AfricanAmericans and AsianAmericans comprisedthe non-European
ethnicgroup,whilein 188079% of thisgroupwas Hispanic.It shouldbe noted
thatthesignificance oftherelationship between ethnicity/race andheading a simple
family remains unchanged, even when marital status is added to the equation.
25 For1880and1900thereis an interesting finding regarding marital statusandhead-
ingcomplexfamilies (notshown).Marriedheads(in comparison tonever-married,
widowed, and divorced heads) were more likely to live in extended families in 1880,
as mightbe expected.But beingunmarried did notoperatein thesamewayfor
never-married, divorced,and widowedheads.Never-married headsweresignifi-
less
cantly likely to head complexfamilies, but thiswas not the case forthewidowed
anddivorced; therewasno significant difference betweentheonce-married andthe
presently married, and this finding holds for both 1880 and 1900, net of the other
variables entered intotheequations.
26 The significant relationship for1850disappears whenproperty ownership is entered
intotheequation.Those headswhoownedproperty weresignificantly less likely
to livein non-kin families thanthosewhodidnot.This is alsotruein 1870,butat
thattimepoint,thereis also a negative and significant relationship betweenbeing
California-born andlivingina non-kin family.
27 The effects ofmarital statusin1880and1900areverystrong: Itis thenever-married,
widowed, anddivorced family heads who live in non-kin families, notthemarried.
A head'sproperty ownership also has a and
significant negative effecton livingina
non-kin family, butwe haveno measurefor1880and themeaning oftheproperty
is different in eachoftheotheryears.In 1850information on realestateproperty is
available;in 1870itis realandpersonalproperty, andin 1900itis whether thehead
ownedorrentedhis/her ownhome.

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
in Los Angeles 35
FamilyStructure

References

Abrams, Philip(1981)Historical Sociology. Ithaca,NY: CornellUniversity Press.


Aldrich,John,and Forrest D. Nelson (1984) Linear Probability,Logit,and Probit Models.
Beverly Hills, CA: Sage.
Alter,George(1988)Familyand theFemaleLifeCourse.Madison:University ofWis-
consinPress.
Alvarez,RobertR.,Jr.(1987)Familia:Migration andAdaptation in Baja andAltaCali-
fornia,1800-1975. Berkeley: University ofCalifornia Press.
Anderson, Michael(1971)FamilyStructure in Nineteenth-Century Lancashire.Cam-
bridge:Cambridge University Press.
Bennion,Sherilyn Cox (1990)Equal totheOccasion:WomenEditorsoftheNineteenth-
century West.Reno:University ofNevadaPress.
Berkner,Lutz K. (1972) "The stem family and thedevelopmental cycleofthepeasant
household: Aneighteenth-century Austrian example." American Historical Review
77: 398-418.
Bloch,Ruth(1978)"Untangling therootsofmodernsexroles:A survey offourcenturies
ofchange."Signs4: 237-52.
Brenner,Johanna, and BarbaraLaslett(1986)"Social reproduction and thefamily," in
Ulf Himmelstrand (ed.) The Social of
Reproduction Organization and Culture.
London:Sage: 116-31.
- (1991)"Gender,socialreproduction andwomen'sself-organization in thedevel-
opment of the U.S. welfare state." Gender and Society 5: 311-33.
Burgess,Ernest(1960)AginginWestern Societies.Chicago:University ofChicagoPress.
Castanada,Antonia I. (1990)"Gender, race and culture: Spanish-Mexican womeninhis-
toriography of frontier California." Frontiers 11:8-20.
Cleland,RobertGlass (1969)The Cattleon a ThousandHills: SouthernCalifornia,
1850-1880.San Marino,CA: Huntington Library.
Clifford,GeraldineJoncich(1991)"'Daughtersintoteachers':Educationaland demo-
graphicinfluences on the transformation of teachinginto 'women'swork'in
America," in Alison Prenticeand Marjorie Theobald(eds.)WomenWhoTaught:
R.
Perspectives on the History ofWomen andTeaching. Toronto: University ofToronto
Press:115-35.
Cott,NancyE (1977)The BondsofWomanhood: "Woman'sSphere"in NewEngland,
1780-1835. NewHaven,CT: YaleUniversity Press.
Dahlin,Michel(1980)"Perspectives on thefamily lifeoftheelderlyin 1900."Geron-
20:
tologist 99-107.
Degler,Carl(1980)AtOdds:WomenandtheFamilyin AmericafromtheRevolution to
thePresent.NewYork:OxfordUniversity Press.
Deutsch,Sarah(1987)No SeparateRefuge:Culture,Class and Genderon an Anglo-
HispanicFrontier in theAmerican Southwest, 1880-1940.NewYork:OxfordUni-
Press.
versity

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
36 SocialScienceHistory

(1994)"Gender,Labor History, and Chicano/aEthnicIdentity." Frontiers14


(2): 1-22.
Douglas,Ann(1977)The Feminization ofAmerican Culture.NewYork:Avon.
Dumke, Glenn S. (1944)The Boom ofthe in
Eighties Southern California.San Marino,
CA: Huntington Library.
Fishman,Robert(1987)BourgeoisUtopias:The Riseand Fall ofSuburbia.NewYork:
Basic.
Fogelson, RobertM. (1967)The Fragmented Metropolis:Los Angeles, 1850-1930.Cam-
bridge, MA: Harvard University Press.
Folbre,Nancy,andMarjorieAbel(1989)"Women's workandwomen'shouseholds: Gen-
derbiasin theU.S. Census."SocialResearch56: 545-69.
Garcia,MarioT (1980)"The Chicanain American history:The MexicanwomenofEl
Paso,1880-1920-acasestudy." PacificHistorical
Review49: 315-37.
Giddings, Paula(1984)WhereandWhenI Enter:The ImpactofBlackWomenon Race
andSex in America. NewYork:Bantam.
Glenn,EvelynNakano(1983)"Splithousehold, smallproducer and dual wageearner:
An analysisof Chinese-American family strategies."
Journal ofMarriageand the
Family 45: 35-46.
Goldin,Claudia(1977)"Femalelaborforceparticipation: The origins ofblackandwhite
difference, 1870-1880."Journal of Economic History 37: 87-108.
Goode,William J.(1963)WorldRevolution andFamilyPatterns. NewYork:FreePress.
Gordon, Linda (ed.)(1991a)Women, and
theState Welfare. Madison:University ofWis-
consinPress.
(1991b)"Blackand whitevisionsofwelfare: Women'swelfare activism,1890-
1945."Journal ofAmerican History78: 559-90.
(1992)"Socialinsurance andpublicassistance: The influence ofgenderin wel-
farethought in theU.S., 1890-1935." American HistoricalReview97: 19-54.
Greene, WilliamH. (1989)LIMDEP (version5.1).NewYork:Econometric Software.
Griswold, RobertL. (1980)"Apartbutnotadrift: Wives,divorceand independence in
California, 1850-1890." Pacific Review
Historical 49: 265-83.
Griswold del Castillo,
Richard(1979)The Los AngelesBarrio,1850-1890: A SocialHis-
tory. Berkeley: University ofCaliforniaPress.
- (1984)La Familia:ChicanoFamiliesintheUrbanSouthwest, 1848tothePresent.
NotreDame,IN: University ofNotreDame Press.
- (1990)The TreatyofGuadalupeHidalgo:A LegacyofConflict. Norman:Uni-
versity of Oklahoma Press.
Hareven,TamaraK. (1982)FamilyTime and Industrial Time: The Relationship be-
tweentheFamilyandWorkin a New EnglandIndustrial Community. New York:
Cambridge University Press.
- (1991)"The history ofthefamily andthecomplexity ofsocialchange."Ameri-
canHistorical Review96: 95-124.

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
in Los Angeles 37
FamilyStructure

Hayden,Dolores(1989)"BiddyMason'sLos Angeles,1856-1891."California History


68: 86-99.
Hirata, Lucie Cheng (1979) "Free, indentured, enslaved:Chinese prostitutes in
nineteenth-century America."Signs5: 3-29.
Jones,Jacqueline (1985)LaborofLove:BlackWomen, WorkandtheFamilyfromSlavery
tothePresent.NewYork:Basic.
Kerber,LindaK. (1988)"Separatesphere,femaleworlds, women'splace:The rhetoric
ofwomen'shistory." Journal ofAmerican History75: 9-39.
Laslett,Barbara(1973)"The family as a publicandprivate Anhistorical
institution: per-
spective." Journal ofMarriageandtheFamily35: 480-92.
(1975)"Householdstructure on an American frontier:Los Angeles,California
in 1850."American Journal ofSociology81:109-28.
(1977) "Social changeand the family:Los Angeles,California, 1850-1870."
American Sociological Review42: 268-91.
(1980)"Beyondmethodology: The placeoftheory in quantitative
historicalre-
search."American Sociological Review45: 214-28.
(1990)"Women'sworkin late-nineteenth-century Los Angeles:Class,gender
andthecultureofnewwomanhood." Continuity andChange5: 417-41.
(1991)"Biography as historicalsociology."TheoryandSociety20: 511-38.
Laslett,Barbara,andJohanna Brenner (1989)"Genderandsocialreproduction: Histori-
cal perspectives." AnnualReviewofSociology15:381-404.
Laslett,Peter(1965)The WorldWeHave Lost.London:Methuen.
Laslett,Peter,andRichardWall(eds.)(1972)HouseholdandFamilyinPastTime.Cam-
bridge:Cambridge University Press.
Locke,MaryLou (1982)"'Like a machineor an animal':Working womenofthelate
nineteenth century urbanfarwest,in San Francisco, Portlandand Los Angeles."
Ph.D. diss.,University ofCalifornia at San Diego.
Marsh,Margaret (1990)SuburbanLives.NewBrunswick, NJ:Rutgers University Press.
Mason,KarenOppenheim, andLisa G. Cope (1987)"Sourcesofage and date-of-birth
misreporting in the1900U.S. Census."Demography 24: 563-73.
Mason,KarenOppenheim, MaxineWeinstein, andBarbaraLaslett(1987)"The decline
offertility in Los Angeles, California, 1880-1900."Population Studies41:483-99.
in
Maynes,M. J.(1995)TakingtheHardRoad:LifeCourse French GermanWorkers'and
Autobiographies in theEra ofIndustrialization. ChapelHill: University ofNorth
CarolinaPress.
McWilliams, Cary(1946) SouthernCalifornia Country. New York:Duell, Sloan and
Pearce.
Modell,John,and TamaraK. Hareven(1973)"Urbanization and themalleablehouse-
hold: An of and
examination boarding lodging in American families."Journalof
Marriage and the Family 35: 467-79.
Moen,Phillis,andElaineWethington (1992)"The conceptoffamily adaptivestrategies."
AnnualReviewofSociology18:233-51.

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
38 SocialScienceHistory

Morgen, Myfanwy, andHildaH. Golden(1987)"Immigrant inanindustrial


families city:
A studyofhouseholds inHolyoke, 1880."Journal ofFamilyHistory 59-68.
(spring):
Newmark, Harris(1970)SixtyYearsin SouthernCalifornia: 1853-1913. Los Angeles:
ZeitlinandVerBrugge.
O'Brien,Jean(1996) "DivorcedfromtheLand: Accommodation strategiesof Indian
womenin eighteenth-century New England,"in M. J.Maynes,AnnWaltner, Bir-
gitteSoland, and Ulrike Strasser (eds.) Gender,Kinship, Power: A Comparative
andInterdisciplinary History.NewYork:Routledge: 319-333.
Padilla,GenaroM. (1988)"The recovery ofChicanonineteenth-century autobiography."
American Quarterly 40: 286-306.
- (1993)My History, NotYours:The Formation ofMexicanAmerican Autobiog-
raphy. Madison: UniversityofWisconsin Press.
Parsons, Talcott(1959)"The socialstructure ofthefamily," in RuthAnshen(ed.) The
Family: Its Function and Destiny. New York: Harper Row:173-201.
and
PersonalNarratives Group (1989) Interpreting Women's Lives: FeministTheoryand
PersonalNarratives. Bloomington: Indiana UniversityPress.
Phillips,George Harwood (1980)"Indians in Los Angeles,1781-1875: Economicinte-
gration,social Pacific
disintegration." HistoricalReview49: 427-51.
Pitt,Leonard(1970)The DeclineoftheCalifornios: A SocialHistoryoftheSpanish-
Speaking Californians,1846-1890. Berkeley: UniversityofCaliforniaPress.
Ruggles,Steven(1987) ProlongedConnections: The Rise of the ExtendedFamily
in Nineteenth-Century Englandand America.Madison:University of Wiscon-
sinPress.
Rury, JohnL. (1991)EducationandWomen'sWork:FemaleSchoolingandtheDivision
ofLaborinUrbanAmerica, 1870-1930. Albany:StateUniversityofNewYorkPress.
Ryan,Mary(1981)Cradleof theMiddleClass: The Familyin Oneida County,New
York,1790-1865. NewYork:Cambridge UniversityPress.
Sanchez,GeorgeJ. (1990)"'Go afterthewomen':Americanization and theMexican
immigrant woman,1915-1929," inEllenCarolDuboisandVickiL. Ruiz(eds.)Un-
equal Sisters:A Multicultural Readerin U.S. Women'sHistory. NewYork:Rout-
ledge:250-46.
Skocpol,Theda (1992)Protecting SoldiersandMothers:The PoliticalOriginsofSocial
Policyin theUnitedStates.Cambridge, MA: HarvardUniversity Press.
Smith-Rosenberg, Carroll(1975)"The female worldofloveandritual:Relations between
womeninnineteenth-century America."Signs1(1):1-30.
- (1985)Disorderly Conduct:VisionsofGenderinVictorian America. NewYork:
OxfordUniversity Press.
Sobek,Matthew(1991)"Classanalysis andtheU.S. censuspublicusesamples."Histori-
cal Methods24: 171-81.
Strober, MyraH., and A. G. Lanford(1986)"The feminization ofpublicschoolteach-
ing:Cross-sectional 1850-1880."
analysis, Signs11:212-35.

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions
in Los Angeles 39
FamilyStructure

Takaki,Ronald(1989)Strangers froma DifferentShore:A HistoryofAsianAmericans.


NewYork:Penguin.
Tilly,Louise A. (1989) "Gender,women'shistory, and socialhistory."
Social Science
History13(winter): 439-62.
UnitedStatesDepartment ofCommerce, BureauoftheCensus(1978)Twenty Censuses:
Population andHousingQuestions, 1790-1980.Washington, DC: U.S. Government
Printing Office.
Weber,DevraA. (1989)"RaisFuerte:OralHistory andMexicanaFarmWorkers." Oral
History Review 17:47-62.
Welter, Barbara(1978)"The cultoftruewomanhood: 1820-1860."In MichaelGordon
(ed.) The American Familyin Social-Historical
Perspective, 2nded. NewYork:St.
Martin's:313-33.
Yans-McLaughlin, Virginia(1977)Familyand Community: ItalianImmigrantsin Buf-
falo, 1880-1930.
Ithaca,NY: CornellUniversityPress.

This content downloaded from 64.50.95.2 on Tue, 07 Apr 2015 22:32:30 UTC
All use subject to JSTOR Terms and Conditions

You might also like