ED 297 214

AU"' :'OR

INSTITUTIONSPONS AGENCY

REPORT NOPUB DATEGRANTNOTEAVAILABLE FROM

PUB TYPE

.EDRS PRICE

DESCRIPTORS

ABSTRACT

DOCUMENT RESUME

CG 020 988

Goldscheider, Calvin; Goldscheider, Frances K.Moving Out and Marriage: What Do Young AdultsExpect?Rand Corp., Santa Monica, Calif.National Inst. of Child Health and Human Development(NIH), Bethesda, Md.RAND/N-2634-NICHDJul 87P50-HD1263913p.

The Rand Corporation, Publications Department, 1700Main St., Santa Monica, CA 90406-2138 ($4.00; 25%discount on 25 or more copies).Reports - Research/Technical (143)

MF01 Plus Postage. PC Not Available from EDRS.Ethnic Status; *Expectation; High Schools; *HighSchool Seniors; Marriage; *Place of Residence;Population Trends; Racial Differences; ReligiousFactors; *Residential Patterns; Sex Differences;Socioeconomic Status

Living independently before marriage is part of abroader pattern of family and demographic change characterizingmodern societies since World War II. This study examined expectationsabout premarital residential independence among young adults. Datawere obtained from 28,240 high school seniors who participated in theHigh School and Beyond study of the class of 1980. Students wereasked at what age they expected to get married and at what age theyexpected to live in their own home or apartment. Those who gave anexpected age at residence younger than an expected age at marriagewere defined as expecting independent residence before marriage. Theresults revealed that approximately 70% of young adults surveyedexpected to move out of their parental home before marriage,suggesting that new norms are emerging which fit into patterns ofindependence in the transition to adulthood. There was substantialvariation in factors affecting expectations about premaritalresidential independence. Young men more than young women, those withmore parental resourcesl, those who expected to marry at older ages,and those who did not have ethnic and religious ties that linked themto their parental home until marriage expected tc live independently.Religious, racial, and ethnic differences interacted in complex wayswith gender and socioeconomic status to influence expectations aboutpremarital residential independence. (Author/NB)

3000000000000000000000000000000000000000000000000000 0000000000000E* Reproductions supplied by EDRS are thebest that can be made

from the original document. M#3000000000000000000000000000 300000000000000000000000000000000000000000E

4,7

'4#f..stspi#0!` by : cr .awatlet440, 10,0iiAtA t '61411, Health and fitiniatt Zeiieiopment ,tAiiria): '40'partiiierit of 'Oct

4020:Seriesip ,e;Repaf the:ptirciiatpubileition:Cogtien-##0(Elikii44100.1*.0 Q* 04e404460.44S405Pral i*irch results:

sTtie RAND No[e reports reiearclit'foirgei**17f*HAtio,,c;:otoo.ati-90,do,hoi:70cessarilyitOflect--ttie

, .

A RAND NOTE N-2634-NICHD

Moving Out and Marriage:What Do Young Adults Expect?

Calvin Goldscheider, Frances K. Goldscheider

July 1987

Prepared forThe National Institute of Child Healthand Human Development

RESEARCH NOTES

MOVING OUT AND MARRIAGE: WHAT DOYOUNG ADULTS EXPECT?*

CALVIN GOLDSCHEIDER PRANCES K. GOLDSCHEIDERBrown University and The Rand Corporation

Living independently before marriage is part of a broader pattern of family anddemographic change characterizing modern societies since World War 11. This paperexamines expectations about premarital residential independence among young adultsand is based on data from a survey of the high school class of 1980. That about 70percent of young adults surveyed expect to move out of the parental- home beforemarriage suggests that new norms are emerging which fit into patterns of independencein the transition; to adulthood. There is 'substantial variation in factors affectingexpectations about premarital residential independence. Young men more than youngwomen, thosewith more parental resources, tlsose who expect to many at older ages,and those who do not have ethnic and religious ties that link them to their parentalhome until marriage expect to live independently. Religious, racial, and ethnicdifferences interact in complex ways with gender and socioeconomic status to influenceerpectations about premarital residential independence.

INTRODUCTION

A new phenomenon has emerged in thetransition to adulthood: many young adults areestablishing an independent household prior tomarriage. Such a life cycle pattern has beenextremely rare, historically and comparatively(Hajnal 1982). Premarital residential indepen-dence is the most recent example of a broaderpattern of family change in modem societiessince World War IIthe tendency for adults tolive apart from other adult family members(Kuznets 1978). The tendency. was earlierevident for widows (Kobrin 1976), and thenappeared among younger adults as risingdivorce rates led to nonfamily living andone-parent families (Ross and Sawhill 1975). Itis now emerging at the beginning of adulthoodas children leave the parental home at youngerages (Goldscheider and LeBourdais 1986).Living in a family setting is now increasinglylimited to sharply defined life cycle stages,married adults, and minor children.

Direct all correspondence to Frances K. Gold-scheider, Depamnent of Sociology, Brown University,Providence, RI 02912.

The order of authors is alphabetical. The researchreported here is part of a project on Premarital ResidentialIndependence among American Ethnic Communitiessupported 14 NICHD Center Grant P50-HD12639 to TheRand Corporation. Allan F. Abraltamse and Joan W.Keeley helped in organizing the data, and JulieDeVanzo, Peter Morrison, and Linda Waite providedconstructive comments. An early version of this paperwas presented at the annual meetings' of the PopulationAssociation of America, San Francisco, April, 1986.

Most scholars consider these changes to bebenign, a continuation of the long-standingdecline in family extension as people are moreable to buy privacy and maintain the indepen-dence of adulthood (Michael et al. 1980).Possibly the normative definitions of who liveswith whom are becoming more restrictive,raising the concern that families are becomingless responsive to the economic needs ofmembers or potential members (Harever 1982).Moreover, young persons who live indepen-dently before marriage alter their family forma-tion patterns. This occurs partly because theychange their family orientations during theperiod they are away (Waite et al. 1986), butis also likely that the costs of maintaining aseparate residence delay new family formation.Young adults with independent residences mayalso suffer career consequences if they are lessable to draw on family resources than if theyshared living arrangements with their parents.

In this paper; we examine cxpectations aboutpremarital residential independence among youngpersons just on the threshold of adulthood.

THEORY AND HYPOTHESES

It is likely that many of the variables influencingnonfamily living at later stages of the life coursealso exert influence in early adulthood. Re-search has shown that income is positivelyrelated to being a household head (Cailiner1975). Some of the other,variables affecting theprobability of living independently are factorsthat make living alone possible such as healthand city size, and are of interest to policy

Reprinted from The American Sociological Review, 1987, Vol. 52 (Apri1:278-285),.© the American Socio-logical Association.

2

makers and planner3 (Soldo 1977; Chew ;983).However, other results suggest that there havebeen changes in the normative and preferencestructures underlying social life (Pampel 1983).

The importance of norms and preferences forthe extent of nonfamily living may be inferredfrom the large differences that appear betweenthe sexes and among ethnic groups. Women aremore oriented toward traditional, family-basedlife-styles than men, even when they havecomparable options (Scanzoni 1975). Similarly,traditional ethnic groups tend to be much morefamily oriented irr their "residence rules" andliving arrangements (Goody 1972; Kobrin andGoldscheider 1982) as well as in the timing andrate of marriage and divorce (Kobrin andGoldscheider 1978; Tienda and Angel 1982).

These patterns suggest that changes in gender-or ethnic-based family norms may have asubstantial influence on the growth of nonfamilyliving. The revolution in sex roles has increas-ingly allowed women independence from familyroles and responsibilities. This may increaseyoung women's expectations for independentliving before marriage. The American ethnicmosaic is also rapidly changing through theentry of new immigrant groups and generationalshifts in older ethnic communities; this, too,should be associated with changes in familynorms and relationships.

The analysis of variation in premaritalresidential independence therefore requires datathat measure the normative dimension underly-ing actual residence-related behavior. There areno data sets designed precisely to tap suchnorms. However, a study of the high schoolclass of 1980 (High School and Beyond) askedhigh school seniors about their expectations forthe timing of marriage and of living in a firsthome or apartment. To the extent that youngpeople talk about these individual-level plansand share them with each other, they representnorms for their generatkin.

This_measure alloWs us to exsmii-.c the extentto which actual patterns of residential inaepen-dence have been built into the expectations ofyoung adults. Collected late in the final year ofhigh school, these data reflect plans immedi-ately prior to the time many begin moving out ofthe parental home and before contingenciesemerge that affect actual residence and marriagechoices. They indicate more purely than actualbehavior the pressures that young people feel toleave home and move out on their own beforemarriage, even if they are not economically,ready.

We examine the relationships between, expec-tation's for premarital residential independenceand ethnicity, gender, and resources. There arelikely to be dimple); interactions among thesevariables. We expect the following:

AMERICAN SOCIOLOGICAL REVIEW

Boys are More likely than girls to expectpremarital residential independence. This differ-ence may be related to gender differences in thetiming of marriage or may reflect greaterindependence for males even before muriage.

Young adults from families with highersocioeconomic status are more likely to expectpremarital residential independence, since theyare rare ab'e to afford to maintain a separateresidence.

Young persons from ethnic groups with ahigh proportion first generation and those fromnioreAraditional religious backgrounds will beless likely to expect premarital residentialindependence. Immigrants have had less eitpo-sure to modern family values; more traditionalreligions tend to reinforce familism on a widevariety of issues.

Ethnic differences should decrease with in-creases in socioeconomic status, but religiousdifferences should not. Increasing, affluence.should indicate greater exposure to modernvalues for most immigrants, but this should notbe the case for those continuing to affiliate withmore traditional religious denominations.

Sex differences should be larger among thosefrom families with lower socioeconomic statusand those from more traditional ethnic groups.Convergence by gender as socioeconomic statusincreases should appear overall, and particularlyfor ,those in more traditional ethnic groups,where more resources are likely to imply greateracculturation to modern sex-role orientations.

Expectations for premarital residential inde-pendence should be closely related to expectedage at marriage. Those who expect to marryyoung have much less time for premaritalresidential independence than those who expectto marry later.

DATA AND MEASUREMENT

This piper draws on data collected from the28.240 seniors in the High 'Sehobl and Beyond(HS&B) study of the class of 1980. Informationwas obtained through questionnaires adminis-tered in ,high schools' and the sample isnationally representative, with some oversampl-ing of schools with large Hispanic enrollments.This survey provides the only data available thatallow us .to measure premarital residentialexpectations and to .classify respondents insufficient detail in terms of ritee, religion; andnational origin' that we can identify real ethniccommunities:''

The dependent variable was constructed from

' The data are slightly biased, in that those droppingout of School before completing, high school are notincluded. Further inforinition on this study is available inJones et al. (1982).

MOVING.OUT. AND MARRIAGE

two questions: (1) At'whainge do yOu expect toget married? and (2) At what ageo you expectto live in.your own hOme "-or ,apartment?2-71'hoiewhogaye tmeXpected.nge- at resideneeyoungerlhan,in expected age Sat.marriage were. definedas expecting independent residence beforemarriage. Thoie who gave- the same age formarriage and independent .residence,, well-as,thoiewhoexpected to marry,before eitabliShingan inckpendent 'residence, were classifiedmaking-4, "traditiOnal" respOnse - that. is not

- ". ,

inovingout befutemarnage:We measure ethnicity on the basis of

responses -to questions_; about race, mitiortit'-origin and- .religioni. ba:CkgrOund. This ispartictilarlrimportant, because the Meaning, ofethnic variation; is, obscured when measured forbrciad?categoriii . Hispanics) -or, only by: aSingle -factor :(e.g., Catholic). Historically,

nic group membership hat' 'arisen -fronici tbitiatiints, of .criteria, variously includingrats national taigin,,and religion. The richneasof the 'ethnie classifications alsk,allOtvs- us 'toexamine ethnic groups this are heterogeneous ona wide ifitietk of characteristics. We candiatinguish.thoSe thaiire mostly first generation

Asians)- from those: tiimitrily, of earlierinunigrant wives . (e.g. , .Jew ;).- or 'eVen longer--term -descent% (e.g.,. those froimNorthem andWestern Ent*); those.-,with distinctively loWsocioeconomic - status-(eg., blacka ancl-PuertoRicans)-froni the.few (Such as Asians and Jews)that have been particularly -successful. Simi-

2 The-Options included Single ages between eighteenand twenty-nitte..leisilian eighteen, thirty or more, havealready, doneio, and do riot t-expect to do' so,

3 Both the thedern'aitd traditional responses include avariety, ,orrelponse combinations. Among thole coded"modern" .arethe few:Cases-who were - already resklen-tially independent -but expected marriage later and thosewho do not 'cipect ,tri marry but expected independent-A-teeT1---"b-16e". whoexpect-to rennin .in the Parental, honie,tintil, some ageafter that chblen for marriage and those few who expectto marry do not expect to, be residentiallyindependent.

There is some ambiguity over the precise meaning of"own home or apartineni." `Did respondents repoitseiniauintintWous .living' ariangenients such as collegedorms -and militarybarineks,is their. "own home" (cf.

'CioldsCheidei-tinctDsVanw-1986)? To investigate this'.possibility: we tainilated'a'SeParateuestiOn on expectedliving- wrisigetnenti "next .year," which included asoptions -dorms and' barracks; by the sequence ofresidence-mirriage expectations (by age). The vastmajority of ,respondents ,planning to be in domes or'Winds (93'percent)' did not regard -them as' residential'indePende*.

Mete, were straightforwaid questions on "What isyour . . ?" kinetiNeiglit. percent responded 'to, thequestion about race; 93 . percent,, to religiOn: and 86percent, to origin.

3

laxly, -some groups have been characterized -byrapid relative upward mobility of women despitepatriarchal-cultural sources;- this appears to bethe case for Jews and-many Asians, and may be

:true for the black community as well. Othergroups;,,re thoUght. to be considerably morepatriarchal .and controlling of women, particu-

1arly Hispanics:To maximize this -heterogeneity, we divided'

the race-origin, categories into AsianzAinericans,Aniericarr Indians, Hispanics blacks and non -'Hispanic Whites. We subdivided Hispanicsftirther into _Cubans, MexiCaria, Puerto Ricans,nd "other Ifiapanics," and non - Hispanic whites

into those from more, traditionally, family-centered: origins ,(Southem Aiiipeiths). andothers. 'The .religiotis. denontinittioris Were Alsodaisy:kr:Lie maximize heterogeneity in'family-relited'.paiterns. .Protestants were subdividedintoInto the more traditionally fundainentalistdenom-inations, ineltiding Baptists and "other" Notes-nun denominations, and less traditionally funds-mentalist denominaticins7EPiseopaliansiPresbyterillis,Methodists, and Lutherans.

Other eitplanatory-Uiriables- included in theanalyaiS are gender and soCioeconrunic, status,measured, by a scale constructed by the surveygroup based on a combination of parentaleducation, occupation, and income.

ANALYSIS

Bivariate Patterns

Overall, a subitantial proportion (about seven-out of ten) of high school seniors expect to liveindependently in early adulthood before gettingmarried (Table 1): There is some variationamong_ ethnic communities, although few largedifferences emerge. There is no differencebetween blacks, and whites, while AmericanIndian's and Asians expect ,somewhat lowerlevels of premarital residential independence.The -four- Hispanic, subgroups- have-loWer,thinaverage expectations, with Cubans clearlyexceptional- even among Hispanics '(the onlygroup with less than 60 percent expectingpreniarital residential independence). -Variation

-among European-origin ethnics appears mini-mal. Those who are 'Jewish and those with noreligious background have theighest level ofexpected premarital' residential independence.The more traditional' Protestant denominations

,ha :T. the lowest.leVel among,religions grobps.A direct relationship emerges between paren-

tal socioeconomic status and Children's expecta-tions abOutpremarital-residential independence.Further, males in alleategOries are more, likelyto -expect premarital- residential- independence.Sharper,gender.differenees characterize Hispan-ics -and :particularly Cubans; -greater gendersimilarity-characterizes Asian-Americans, Jews,

4 AMERICAN SOCIOLOGICAL REVIEW

Table 1. 'Percentage Expecting Premarital ResidentialIndependence. by Sex and by Race-Origin Cate-gory, Socioeconomic Status. -and Religion

N foiTotal =Males Females Total

Total 69.8 75.0 65.2 25086Race -origin

White. non-Hispanic 69.9 75.0 65.4 15764-Southern European 69.8 73.S 64.5 1287Other 70.0 75.1 65.4 14477

Black 69.8 75.6 65.4 2648

Hispanic .64.1 72.0 '56.4 2781Mexican 63.8 72.1 55.4 1664Cuban 49.7 63.3 39.1 308Puerto Rican 65.7 76.5 57.6 266Other Hispanic 67.5 72.5 62.3 543

American Indian 66.7 72.6 61.9 533Asian-American 66.7 68.2 65.5 466

Socioeconomic scale

1.43'41 65.3 72.1 60.3 7506Middle 69.2 74.5 64.4 11833Upper 75.9 78.4 73.1 5747

ReligionTrad. Protestants 66.4 72.5 61.5 82%Other Protestants 72.6 77.2 68.4 3087Catholics 71.1 75.9 66.8 9310Other Christians 64.9 703 59.9 1494Jews 79.6 80.8 78.3 507Other religion 64.3 69.4 60.0 888No religion 79.7 80.9 78.3 1298

those with no religion, and those with highersocioeconomic status.

Multivariate Models

We used OLS !egression for the multivariateanalysis. because it provides a reasonably robust-approximation of logistic regression for outcomeprobabilities, such'-ai these; that do not deviatefar from -50-50. We present five equations,introducing' controls. and new variables as weproceed:(Table 2).5'

The-first equation-(Model 1) includes only therace and- national 'origin classifications, andshows significantly lower expectations of pre-marital residential independence among Cubansand, Mexicans, as well as among AmericanIndian and .Asian subpopulations.. than amongthe- reference 'category of' non4lispanic whitesnot reporting-southern European national origin.As- in .Table:1, the.Cubins are most distinctive.

s The equations that include more complex interactionterms Were also estimated,whh logistic regression.-Theserevealed no substantiVe differences in results. We presentthe OLS relults because, of their greater ease ofinterpretation:

Neither blacks nor Southern Europeans-emergeas significantly different.

In Model: 2- we introduce sex 'and socioeco-nomic status and aft,* that these are significant.Controlling-lor them eliminates American 'In-dian and Asian differences, suggesting thatsocioeconomic factors 'accounted 'for the biVari-ate patterns, Among- Mexican, and

doHispanics, these- factors weaken, but do notaccount for, their pirtieularly low- levels ofexpected: *Marital- residential. independence.However, blacks are significanth different from

.whites when secioeConotnic.itatua is introduced,with young black, adults .having somewhathigher levels Ofexpectation. Thus, the overallbiVariate 'similarity of -blaoki and Whites inModel'l is the.result ofe distortion' generntedby the tocioecOnomic'differences betweethesegroups; when socioeconoinieeVells Controlled;blacks are more likely to eipect premarital.residential independence.

'Model 3 consolidates- ethnic categories tohighlight groups differing significantly; mergingAmerican Indians, 'Asians, Puerto Ricans, OtherHiSpanics.. and Sotithern -Europeans. with theomitted grinip. 'We integrated :the latter- threewith the other religious subgroups by creating asribdivision.ainong Catholics 'to tav:traditicinal-ism, which includes- Hispanic' and SouthernEuropean Catholics. 'The religioui difference'sconstructed in this fishinn'are all significant.

We also introduce - interactions betWeen. Sex'and ethnicity in Model 3. Cubans and Mexicansshow signifiCintly greater sex differences inexpectations about premarital residential inde-pendence than other groups. There are alsogroups with more egalitarian ,gender ,patterns-than the reference group. Asians, Jews, andthose with no religion.(grouped-as-MINMALE)have almost no differences between young menand women -on this issue.

"The anticipated -interaction' between 'sex andsocioeconomic igii i Nan t Whileexpectations for premarital residential 'indepen-dence are generally -tied- to- socioeconomicresources, males require "fewer parental re-sources than females to buy similar levels ofpremarital residential independence. This effectwas a general. one, with -no significant differ-ences.-for-either 'traditional ethnic or religiousgroups (data not:presented). Further;-there,',wasno evidence either of rapid=nceidturatiOn among'higher-status members of any of' the ethnicgrotips, or' of resistance to more egalitarianexpectations -among' thoie at higher socioeco-nomic levels who are traditionally fundaMental-iit

On the other hand, an unanticipated findingrelating socioeconomic Slants with ethnicityemerged,-.arriong _String- blacks. The positiverelationship between- status ,and expectations for

8

MOVING OUT AND, MARRIAGE, 5

Table '2. :Increinental OLS Models 'Predicting Preitiarital Residential Independence (Unstandardized RegressionCoefficients) . . .

Variable

Modek1-(Race &

Ethnicity)

'Model 2(Gender.& SES)

Model 4' Model 5'MOdel 3 (Age at -Marriage).

(Religion) I 2

.

IniFrcq* .700" .652" 663" - .656" =.665"INDIAN -.033* - .015-ASIANMEXICAN

- .033*- .062"

- .028,- .030*' -..039* - .056" - .058" .

CUBAN - .203" - .181" - :227" - .203" - .207"PUERTO RICAN - .042 '.011. - -OTHER HISPANIC - .025

-BLACK - .002 .033" .034" .075" ,- .075"SOUTH EUROPEAN, -.011 - .012

MALE .092" .093" - .003 , .008-CUBAN MALE, .142' .114 .116-MEX MALE 056' .079" .081"'MINMALEEARLY'. MALE

- .065" - .083"- - :9086::

SES SCALE .054" .068," .010" - .017"MALESES - .033" - .010 ,-..008BLACKSES. -.043" -.004 .....mg

',MEXMALESES .014 .054* :051*EARLY 'SES' - - - - .04 r,LATE.SES - :043"TRAD PROT' -.044.i -:017" - .017*'OTHER REL - .055" -.Ws. - .043"JE*S. .090" .034* .048.,*EARLY .JEW 7 - " .243"NO RELIGION .069" .071"MODCATH, .:012223"**" - .000 -4.001TRADCATH -.021" -.052" -.052"LN:MAR. AGE - .594" .606"ADJ. R2 .0017 .0195 .0271- .2180 .2194

N 25499;1.96 > t >1.30,

so t > 1.96.

premarital residential independence does notappear 'in-the black ,community, since theirinteriCtien coefficient offsets almost entirely theeffect characterizing.the sinioe as a whole. Thisresult,. coupled With the positive Ovenill.coeffi-;cleat, suggests that.it islhelowest-status blaCksWho are,:particulailY more likely than whites toexpect- independent living ;Prior to marriage,with differences diminishing with increases inio,cioeennonuc status This .-distinctive black.pattern requires further reiearcli. linking it to,other family-related eventi;, such. as-,marriageand childbearing, specific tO,the'blaCk,ecitnti-

.nity..We-nOted earlier that theie.4ho,,e.)ipected-to

tnOve.Olii.at nirriage.eiketed-to.nitoy.earlieithan those who expected premarital residential.independence : Since ethnic groups ;jag intistatt-tially in ige,it Marriage, as do men and women,,Mcider'kintrndtiCeitheeipeatedage at marriage(theiltired athe natural "rig):

This change: prOditeed? se,VerillPOvierful ef-:fects. It does not -change . the. significance

either the Mexican but reversesThe sign among ;blacks; suggesting that thepattern of higher levels of, premaritaliesidentialindependence-among-blacks- does not-indicate a-"Modern:" fainily orientation 'but ',rather anunusual, marriage, pattern :, dontrolling:;for. theirconsiderably later age at marriage moves blackssignificantly-helow the norm. This interpretation-is reinforced by the.elimination of the previouslyobseiyedinteraCtion:betweeti race andsoCiaeco-noinia status when expected' marriage ageintiodueed.

The effeets',of. eipeetect "age at. Marriage on-religiOus, differences. are also not uniform.'Traditional Protestants have -16w levels ofexpecptions for'prthaiitil rest rata in epen-dence in part .becriusc- they .expeci-.tci Maryearly; 'their- coefficient decreases, substantiallywhen expected age at Mit:nage, is inn-educed:This also characterizes those of Other Religions.Analogou sly, Jews', and. those iih.- noare: more likely to expect-to move Out :of 'the

.9r 1"

_,

'6'

parental Miiirteforelheymarry in,part becausethey Pec..tó marry late.

The higher: se" -peCiationi,, of nOn-6outhein.,Eakei*-0. Catholics for premarital residential:independenee*..Iie 'eliminated when eiPeCted.PiairiagOise is,inuodueed. Thus; dieseCadio,licsfexpect ic leave the ,parintal home -befoietheir marriage because they expect trinianyline.In . contrast, : none of the -differ:Mai .betweentraditional Catholics and . others is explained byany tendency early. Fie *association of *Moving out 'getting married isacMally, stMngtherkl: .00rdr011ing...for, ex-.pected' timing. of marriage, ingse*ing that theeffects of -.expectiig ô in early' were:obscuring a, very strong avoidance of ribirdindlyliving These clair.. point to the differentialimportance of the timing Ot`nnirriage, for various

.Catholic sUbcoininiinities.:For die.ritoremodeassector, residential independence befOre marriagereflects the structural of tire timing of

'-marriage, for more .tradiiiintal. 'sector,,residential independeiri; before marriage is Vet);unlikely ixInatterwh.61)*riAte***uri!

Expectations about age at ii*giatealio'haYesubstantial effects on.iiviireenamieAndwider _

patterrit..,OVeiall :Sender. differentia *arc el im iPeed Thus, the original pattern of 'hi*expectations for premarital residential' indepen-dence -*Ong Males it:i.refleetiOn of their laterexpected age at initiriage. This is not the case,howeyerf,either, forthOsi groups with unusually

large ..,Seir differences.' or thOie' with .rnineegalitarian gender, Patterns: Cuban and

:males remain more likely to expect premaritalresidential' independence than ;females, evencoiltraling;for.expeCtertlaie- at *isirjaie.: Thereverse pattein'cliarseterizes.egalitarian4rOuPs,..such that Asian and Jewish women are molt-likely thin.ComPirable men to expeCt premaritaliiidChtial independence.

Theeffectofsôcioconomicstatus.if greatlyreduced' when, eipected :Marriage .age 1a cOrcz

This . suggest's. Mit. *while =parentalsocioeconomic ,_ .resources :affect expectationsabout 'premarital 'residential independence,: theirinfluence, Operatet-,priMarilY,thrciugh. delays ofMarriage,. =probably itliough the relationshipbetween !di:dation:int age-at-Marriage.

In ietteril; . the data in'Moilel 4 reaffirinthat-gencler differences n, :preMarital. residentialindependence diminish with ;higher. Meioeco-

i4i4;althotiSh this effect is no longersignifieant. .However,. When the interaction ofgender andaocioeeithoinie.4.101i.Yicitioieitipositive, significant :coefficient emerges forMeitiOant. This pat em hàracteiizes no otherracial or ethnic grool, (results not presented)This suggests that more egalitarian sex i roles

. emerge for most groups among the middle. andupper ;ell-1*i; the 'MeitiCans. le. -able to

AMERICAN. SOCIOLOGICAL REVIEW

withstand this effect.and Prot e4.1 their traditionalsex differentiation when they *quire more.resources. Tlietlarge, Size of dr, gronp, and, iiiSeogiaiihic.,,coneentration Mar-account for this.pattern, whik;smalkrormóre dispersixls:impsmay be Under grager.assimilatOry*pressure.

The- differential inpotiance of the .expeeted'drains of .marriage is further ,eliborated IAModel 5, where 'introduce* "early" andlate" marriage. is tinteractions,' which have aSignificant. effeet.-in:enmbinalion. with ,gender,socioeconomic status, and one religious .group.

;Males are Mare like)rihtut,fernaleslo'expectici*aye out of the 'Parental home before Marriageas a. result- 'Of' their' 'Wei expected age ..it

riage,:hni Only among those:Who marry:law;for those who -expect...16'.Many young,. :sexdifferences. reinain *mete:lido ("early.male").

The positive effect of socioeconomic status onpremarital inaldentisl indePers-denee .eharicter-izcs' only those who "expect early and latemarriages Among those expecting marriagemodal ages,,theerfeCcof socioeconomic statusOver from a-, positive to a negative (thoughrelàtivçly ininbstaitial) effect. This suggests-parental "resoureet-,have: Mneh-greater effect. onekPectitiOnt-foi:Preinaritil residential indepen-dence at the Margins,. when fewer expect to

-Marry:-Most Jews,eXpeet to rttairYiate-and.thereforo

'expect to move.out of the parentalhorne beforeMarriage, Bilt tholeJeWs 'who expectIO ratio

tearly.areNerynnlikely. to eripect to move out ofthe parental home except at rintiriage;This Mayreflect the very 'high- proportion of Jews who

Ile archexpect lo -on ,to-co'college. reg ess ofexpected age at: Marriage' Or- econOmic 611Ck-

lroU04 GOldscliekler 19.86iy Thole .whowant An:Many early. may be .trading off' dietesOurces.thit mould ncirtnifly:suPpOrt.premari-61 residential independence foredtication..

CObig.1/51ONS'ANI? DISCUSSION

the analysis" Suggests diatithere'are SubstantialVariations, in the: fact*. affeeting .preMarital.residential independence and that these relatiOn-Shipa :raise important ,questions 'for- theories ofgender; .ethilic,. And :family, :Change: .e,erall,rno,reloung,rnen.ditia young women expect :to''Hi/cf.:may, froth the .ptireirtal hOulehold beforethey many, as 40 OP* with more parentalrc*Suic.es-i9.13ted.ume*finOtlie"dt0)4?se. Y410-P?iir4P3 niartrat an older age andthose who do not have ties to traditional:ethnic-sad religiOtikgrotips.

,Early ,n defined; as, age overly, or .younger for'.fensiles'and age twenty-one or younger for males,'"late"as age twenty-(our" or older for females and age,iwenty-fiye or older (or inaks.

,

:*0

MOVING OUT AND MARRIAGE

However, these: relationships are complexlyinterwoven. The differences between Catholicsand Protestants would not haiie emerged without-disaggregating both groups into their morereligiously. and cultUrally traditional' compo-nents: Theimportance of distinguishing those of"Other" religions, horn anise claiMing "noreligion" is particularly clear, since the."othets"appear very 'family- oriented-on this' measure,similar "to. both traditional 'Protestants andCathO lici, while.the"no-religioni" present themost extreme likelihood of early. "apartness."The Iewi most closely4tsemble tttose-With no.,religion; "but, more ;of their -high level ofexpeetatiOns for premarital residential ndepen-dence, teems to ,result -from,,:their very late'.expected ages-at.Marriage.

Tlie higher levels of black expectations: orprernmitalie.sidential independence. appear to beprimarilY.iinked to the very unusual patterns of"marriage ineresisingly. characterizing' the blaek-compuinitY paitieululy among, thine Irphr thelowest-socioeconomic status-,families. In Con-

tlisiiinics, and PatticularlyMeXieans andCubans, are less likely'to expect to move out ofthe, iiarental. home before inarriage, no mailerhow .late :they .many what their socideco-

noniic particular patterns CharacterizeAsian .Americans or American Indians:

Greater' affluence:may be leading to. 'earlierresidential separation, between generations; evenainoneethnic and religious ginupS with a strongemphasis .on''fainily solidarity. This. May reducefamily 'control over other. aSpects of hildren'slives earlY.in adulthood, xi. young, people' makethe ,critical. work and family -decisions socharacteristic of that.life cycle stage. 'The effects

._ of such sncialmObility_are..Particularly.inarkedfor daughters, since the gender differenceslinking females more elosely'io.family roles are'also =reduced with ,increases in' socioeconomic'-status: was the case forall groups with onlyone interesting excepticin..HiSpanics'aMexiCanOrigin. Cletrithe traditional diAsions between.the seaes.iii being eiduced by the:procesSes ofsocialchangein this aret,of.faniily structure, as

'in others.Finallyancfmost-difectly, these data point to

ite ubiquity of .eXpectitions for premaritalresidential independence. iniong young adults.Most..yoting:.people. whatever ,their Socioeco-nontic _snit :ethnic baClcground, expect to.independently of the parental household "before'they..mairy.'ThOugli:the structural separation ofresidences is. relatively recent: in a behavioral'sense,, it- iVickntiy has heCome internaliied.nOrmitively by.yOang adults-There Will undoubt;ergy be thoSe.youngladults,Whowill not Actuallycarry out'.' their expectations: ,Veverthelesi,premarital, residential independence is evidentlyanew normative requirement;:,one that. young

people may try hard to realize, and perhaps atconsiderable cost. To do so may require tradingoff on other expenses, such as those formarriage and for education.

REFERENCES.

Carliner. Geoffrey: 1975. "Determinants of HouseholdHeadship." Journal of ,Marriage and the Family37:28=38.

CheW. Ken. 1983. "Metropolitan Differences in the.Level' of .Nonfarnily Households: A Cross`-SectionalCook at the,U.S. in '1970." Paper presented at theAnnual MeetingS 'of the population. Assiiciatiow ofAmetica.

Goldscheider, Calvin. 1986. Jewish 'contitSuity- andChangi:Emerging-Patieras i!Sinierica. Bloomington:Indiana .rsity Press

G,oldscheider. Frances and Julie DaVartio. 1986,"Semi-aut-ononiy-and Leaving Home -in Adulthood."Social Forces 61:187:-20t.

,Goldscheider, Francei; and Celine Lebourdais. 1986."The Falling Age at :Leasint HoMe. 1920-1979."Sociology and Social Research 70:9102.

Goody, Jack. 1972: "The Evo:ution, of the Family." InHosiselsold and Family. in Past Time, edited by PeterLasleo and Richard.Wall. LondoOLCambridie Univer-sity Press.

HajnaL John. 1982. "Household Formation- Patterns in:Comparative. Perspective;" Population and Develop.',Sent Review 8:44994.

Haieven. Tamara. 1982. "Ametican Familles in Transi-tion: Historical Perspectives on Change." P. 446 -66.in NorMal Family PrOcesses. cdited.by Froma Walsh.New lork: Guilford 'Press.

'Jones. Calvin: Miriam Clarke, Geraldine Mobney.'Harold MCWilliains.,Joanna Criwford..Biuce Stephen-son.. and, Roger- tdurangeat. 1982. High:School andBeyond: 19130 Senior Cohort, and First Followup:Datg File Users Milord, Washington: D.C,: NationalCenter for Education Statistics.

Kobrin Frances E.-1976, "The Primary Individual-andthe Changes- in' Living Arrangements Since1940;" Journal of Marriage and the Famity.38;233-39.

Kohrin'. Frances E. and Calvin Goldscheider. 1978: TheEthnic ,Faitor itt Family Structuri. and Mobility.Cambridge.- MA: Ballinger.

'1982. FamilY Extension or NonfamilyLife Cycle..Ecianomie. and Ethnic Factors," WesternSociological Revie13:103-18.

'Kuiriets. Simon. 0978. "Size and Age Struaure of-Faniily Hztsehrlds:'Explorstory Comparisons." Popelation and 1:eiielopris'entlevieW 4:187-233.

Michael:. Robert° T.. Victor. R. Fuchs. and Sharon R.Scott. 1980: "Changes. in -the Propensity to LiveAlone:)95071.976."'Demography 17:39-53.

Pampel: Fred C 1983. "Changes-in the Propensity toLive Atone: Evidence from Consecutive Cross-Seetional Surveis., 1960,1976." Demography

,20:43347;Rois:.Heather L. and Isabel V, Sahill. 1975. Time of

Traniition: The GroWth of 'FaMilies Headed: byWomeri. Washington, D.C.: The'Urbin Institute.

Scanzoni. John. 1975. Sex Riges. Life Styles- andChildbraritig..New-york: Free Press.

8

So 'do, Beth. 1977. "Demographic Composition andSource of Temporal Variation in the Living Arrange-ments Among the Elderly." Paper preserted at theAnnual- Meeting of the Population Association ofAmeba.

Tienda. Musa and Ronald Angel. 1932. "Headship and

AMERICAN SOCIOLOGICAL REVIEW

Household Composition among Black; Hismics andOther Whites." Social Forces 6::5011-53.

Waite Linda, Frances Crldscheider, and ChristineWitsberger. 1936. "The Development of Individral,ism: Nonfamily Living and the Ptns of Young Mereand Women." American Sociological Review 51:541-44.

12

4457.' "IT't

4

7 444 4

4

.

.41

!s1p0pUVATI.

014,1,EsEARCJI. CENTER- rt101r4

radiiMI013,

MAD41)48

-

r. ,' "7-

7";.'"r.