romantic relationships among unmarried african americans and caribbean blacks: findings from the...

Romantic Relationships Among Unmarried AfricanAmericans and Caribbean Blacks: Findings From the

National Survey of American Life*

Karen D. Lincoln Robert Joseph Taylor James S. Jackson**

Abstract: This study investigated the correlates of relationship satisfaction, marriage expectations, and relationshiplongevity among unmarried African American and Black Caribbean (Caribbean Black) adults who are in a romanticrelationship. The study used data from the National Survey of American Life, a national representative sample ofAfrican Americans and Caribbean Blacks in the United States. The findings indicated that the correlates of relation-ship satisfaction, expectations of marriage, and relationship longevity were different for African Americans and BlackCaribbeans. For Black Caribbeans, indicators of socioeconomic status were particularly important correlates of rela-tionship satisfaction. For African Americans, indicators of parental status were important for relationship longevity.Policy and practice implications for nonmarital unions are discussed.

Key Words: African Americans, Afro-Caribbeans, cohabitation, marriage, romantic relationships, socioeconomicstatus, West Indians.

The number of unmarried adults in the UnitedStates has increased dramatically over the past fourdecades. The decline in marriage has been particu-larly pronounced among African Americans (AfricanAmerican Healthy Marriage Initiative, 2000). How-ever, declining marriage rates do not mean a declinein romantic relationships. The majority of investiga-tions on nonmarital unions focus primarily on tran-sitions to marriage. Although these studies areimportant for identifying barriers and facilitatingfactors for marriage entry, they provide less informa-tion about the nature of romantic unions amongunmarried adults that may or may not end in mar-riage. The focus of this investigation is to examinethe correlates of relationship satisfaction, expecta-tions of marriage, and relationship longevity amongunmarried African Americans and Caribbean Blacks.

Studies of nonmarital romantic relationshipsamong cohabiting and noncohabiting adults areimportant for at least two reasons. First, fewer cohab-iting unions are resulting in marriage, and this ismore likely the case for African Americans (Schoen& Owens, 1992) and Caribbean Blacks (Coppin,2000). Consequently, there is growing recognitionamong researchers that not all cohabitations are partof the process leading to marriage and are insteadalternative forms of union (Smock, 2000). Second,few studies account for whether unmarried respon-dents are coupled but not cohabiting.

Most studies examining social life among BlackAmericans have ignored the cultural and economicdiversity within the Black population in the UnitedStates. Blacks are usually treated as a monolith, with-out regard to ancestry or ethnicity. Yet, the Black

*The National Survey of American Life was funded by the National Institute of Mental Health (NIMH; U01-MH57716) with supplemental support from the Office

of Behavioral and Social Science Research at the National Institutes of Health and the University of Michigan. The preparation of this manuscript was supported

by grants from the NIMH (K01-MH69923) to Dr. Lincoln and the National Institute on Aging to Dr. Taylor (R01-AG18782) and Drs. Jackson and Taylor

(P30-AG15281). The authors would like to thank Dr. David H. Chae for his valuable assistance with the data analysis for this study.

**Karen D. Lincoln is an assistant professor of Social Work at the University of Southern California, 669 W 34th Street, MRF 327, Los Angeles, CA 90089-0411

([email protected]). Robert Joseph Taylor is Associate Dean for Research and Sheila Feld Collegiate Professor in the School of Social Work at the University of Michi-

gan, 1080 S. University Avenue, Ann Arbor, MI 48109-1106 ([email protected]). James S. Jackson is Professor of Psychology and Director of the Institute for Social

Research at the University of Michigan, 426 Thompson Street, Ann Arbor, MI 48106-1248 ([email protected]).

Family Relations, 57 (April 2008), 254–266. Blackwell Publishing.Copyright 2008 by the National Council on Family Relations.

population in the United States is becoming morediverse each year, fueled in large part by the immi-gration of Blacks of Caribbean descent. In 1990,Blacks in the United States totaled about 30 million,and by 2000, the total had increased to 36.2 million(U.S. Census Bureau, 2005a). Of the nearly 4% ofBlacks who were foreign born, 60% were from theCaribbean. Moreover, Caribbean Blacks make up25% of the Black American population in majorcities such as Boston, New York, Miami, and FortLauderdale (Logan & Deane, 2003, table 2). Accord-ingly, it is important to consider cultural, contextual,and demographic differences within the Black popu-lation that may be sources of variation in the natureof their romantic relationships.

Theories of Marital Behavior

A considerable body of research in sociology anddemography focuses on the role of economic anddemographic factors to explain changes in maritalbehavior among African Americans (e.g., Kielcolt& Fossett, 1995). Economic explanations generallyseek to explain the decline in marriage by connect-ing these changes to the economic circumstance ofmen and women. Theories of mate availabilityhighlight gender ratios and marriage market condi-tions as important demographic factors that effectthe decision to marry among African Americans.Mate availability is usually measured by the genderratio, that is, the number of men for every 100women. Gender ratios for African Americans tendto be low (i.e., below 100), although they varyacross metropolitan areas, counties, cities, andregions. Because of the chronic shortage of AfricanAmerican men relative to women, marriage marketconditions are considered to be a likely explana-tion for the low marriage rates among AfricanAmericans.

Studies of marriage patterns and family forma-tion among Caribbeans have primarily focused oneconomic explanations—namely, the impact of men(Coppin, 2000) and women’s (St. Bernard, 2003)labor force participation and educational attainmenton relationship formation. Caribbean women haveexperienced tremendous gains with regard to theirparticipation in the labor force. Between 1970 and2000, there were continual increases in female laborforce participation rates in Jamaica, Barbados, andTrinidad and Tobago. Such increases not only area function of several decades of women’s educational

achievements but also are often attributed to socialmovements that stimulated greater gender equity insocial relations. Increases in labor market participa-tion also revealed emergent trends in familial livingarrangements resulting in single mothers who soughtwork to support their families. Although some ear-lier studies reported an association between highersocioeconomic status and common-law marriages,other studies of marital behavior among Caribbeansgenerally supported the link between lower socio-economic status and nonmarital romantic unions(Henriques, 1953). Other research suggested thatthe prevalence of common-law marriages in theCaribbean reflected the lack of bargaining power ofthe women who were willing to use such unions toimprove their chances of marriage (Goode, 1960).Finally, work by Rodman (1971) suggested that al-though marriage was considered the ‘‘more respect-able institution’’ by lower-class Afro-Trinidadianrespondents, poverty was often cited as a factor thatprevented them from marrying. However, a studyby McKenzie (1993) reported a rising trend incommon-law unions among persons of highersocioeconomic status.

The review of the literature of marriage andromantic involvement among Black Caribbeanshighlights the diversity of this population and theimportance of acknowledging cultural differencesbetween African Americans and Black Caribbeans.Researchers of Black Caribbean life note that, con-sistent with many immigrant groups, Black Carib-beans hold strong cultural values that emphasizeeconomic, educational, and material success. Forexample, the social networks of Caribbean Blackshave been credited with facilitating their abilities tocreate niches in housing and labor markets, starttheir own businesses, financially shore up networkmembers who need assistance, and offer coresidenceto help other immigrants establish themselves imme-diately following migration (Waters, 1999). Thiskind of support is not only cultural but also hasimplications for increasing the level of socioeco-nomic status for Caribbean Blacks relative to U.S.-born African Americans. Consequently, the abilityto provide financial security in a relationship maynot only be an expectation for persons in a romanticrelationship but also may be more of a possibility forCaribbean Blacks relative to African Americans.

Overall, the empirical evidence suggests thatamong African Americans as well as Black Carib-beans, both economic and demographic factors have

Romantic Relationships, Unmarried African Americans, Caribbean Blacks � Lincoln et al. 255

contributed to recent changes in marital behavior. Asmall body of work suggests that economic factorsare integral to the formation and maintenance ofintimate unions among African Americans (e.g.,Ruggles, 1997) and may play a more prominentrole in the quality and stability of their marriages(Clark-Nicolas & Gray-Little, 1991). Economicfactors were found to be positively related to maritalsatisfaction and other dimensions of marital qualityfor African American men and women overall andwere particularly relevant for those with lowerincomes (Clark-Nicolas & Gray-Little).

Marriage and Romantic Relationships AmongAfrican Americans and Black Caribbeans

The decrease in the rates of marriage has been par-ticularly striking among African Americans, whoare less likely to be married than either Whites orHispanics (Harknett & McLanahan, 2004). Theyare also more likely than Whites to be separatedand divorced and less likely to remarry (Taylor,Tucker, Chatters, & Jayakody, 1997). Conse-quently, African Americans spend less time inmarriage than Whites. This decreased rate of mar-riage does not mean that they are not involved inromantic relationships. For instance, Tucker andTaylor (1997) found that 4 of 10 unmarried Afri-can Americans and about half of those aged 55years or younger were involved in romantic rela-tionships. Romantic involvements were importanteven among unmarried elderly African Americans,with 1 of 10 indicating that they were romanti-cally involved. Overall, one in three uninvolvedAfrican Americans expressed the desire to be ina relationship. In the same study, gender and agewere significantly correlated with the likelihood ofbeing involved in a romantic relationship, withmen and younger adults having a higher likelihoodof having a main romantic involvement.

Unmarried cohabitation has increased dramati-cally over the past three decades, climbing from500,000 couples in 1970 to nearly 5 million couplesin 2000 (U.S. Census Bureau, 2005b). Althoughcohabitation rates have increased more amongWhites (Bumpass & Lu, 2000), African Americanwomen have higher overall probabilities of cohabit-ing compared to White women (Schoen & Owens,1992). Moreover, although they hold similar expect-ations of marrying their cohabiting partners, AfricanAmerican women are less likely to make the

transition from cohabitation to marriage (Raley &Bumpass, 2003), leading some scholars to concludethat cohabitation has become an alternative to mar-riage (Smock, 2000) or being single (Jayakody &Cabrera, 2002) among African Americans. Althoughfew studies examine the determinants of cohabita-tion among African Americans, most racial compari-son studies attribute delays in marriage, declines inmarriage rates, and economic factors as contributingto the rising cohabitation rates among AfricanAmericans.

There is surprisingly little research on marriageand romantic involvement among Black Caribbeans.The literature on Black Caribbeans in the UnitedStates is mostly ethnographic and addresses issuessuch as immigration, discrimination, social mobility,and relationships with African Americans (Waters,1999). Available studies, however, indicate that mar-riage rates have been consistently low over the pastfew decades and that consensual unions (i.e., visitingand common-law unions) have persisted alongsideunions characterized by formal marriage for quitesome time. Studies in Trinidad and Jamaica reportthat legal marriage was more likely to prevail amongthose with greater economic stability and highersocial ranking (Henriques, 1953). Common-lawmarriages were more prevalent in rural than in urbanparts of Trinidad, often representing 80% – 90% ofall unions in rural districts (Matthews, 1952).Lower-class Afro-Guyanese viewed legal marriage asan upper-class symbol, its real significance being anact of conformity to ‘‘respectable’’ values (Smith,1956). However, more recent work indicates thatthe prevalence of common-law unions has increasedamong women from middle and higher socioeco-nomic levels (McKenzie, 1993). This trend has beenattributed to increasing tolerance of nonmaritalunions in Western societies as well as the home-grown example of lower income Caribbean groups.Research in Trinidad and Tobago has found thatthere were approximately 20 unmarried-couplehouseholds for every 100 married-couple households(Coppin, 2000).

In the present investigation, we seek to addresssome gaps in existing knowledge about romanticunions among Black Americans. Specifically, thecurrent study focuses on a wide range of outcomesthat assess the nature of nonmarital romantic unionsamong African Americans and Caribbean Blacks.This study is unique in that it uses a national proba-bility sample of African Americans and Black

Family Relations � Volume 57, Number 2 � April 2008256

Caribbeans in the United States, which allows forthe ability to examine both differences in romanticrelationships and distinctive demographic, socioeco-nomic status, and relationship profiles and widergeneralizability of study findings.

Methods

Data

The National Survey of American Life: Coping WithStress in the 21st Century (NSAL) was collected bythe Program for Research on Black Americans(PRBA) at the University of Michigan’s Institute forSocial Research (ISR). The fieldwork for the studywas completed by ISR’s Survey Research Center(SRC) in cooperation with PRBA. The NSAL samplehas a national multistage probability design. A totalof 6,082 face-to-face interviews were conducted in2001 – 2003 with persons aged 18 years or older,including 3,570 African Americans, 1,621 Blacks ofCaribbean descent, and 891 non-Hispanic Whites.The response rate is 71.3%. The NSAL sampleconsists of 64 primary sampling units; 56 overlapsubstantially with existing SRC National Sample pri-mary areas and the remaining 8 were chosen fromthe South in order to represent African Americans inproportion to their national distribution.

The African American sample is a national repre-sentative sample of households located in the 48coterminous states with at least one Black adult aged18 years or older who did not identify ancestral tiesin the Caribbean. The Black Caribbean sample wasselected from two area probability sampling frames—the core NSAL sampling frame and a probabilitysample of housing units from geographic areaswith a relatively high density of Caribbean descentpersons (more than 10% of the population). Ofthe total Black Caribbean respondents (N ¼ 1,621),265 respondents were selected from householdsin the core sample and 1,356 were selected fromhousing units in high Caribbean density areas (seeJackson et al., 2004).

Measures

Dependent variables. This analysis examinesthree dependent variables (relationship satisfaction,expectations of marriage, and relationship longev-ity). Relationship satisfaction was measured by the

following question: ‘‘Taking things all together,how satisfied are you with your current relation-ship?’’ Response categories ranged from 1 (very dis-satisfied) to 4 (very satisfied). We recognize thatthere may be qualitative differences between thedesire to marry and the expectations concerningmarriage. However, these terms have at times beenused interchangeably or as proxies of each other inthe literature (South, 1993). Because we wereinterested in understanding respondents’ assess-ments of the subjective expectation of marriage(rather than to predict a behavior such as mar-riage), we assessed respondents’ expectations of mar-riage with the question: ‘‘What do you think thelikelihood is that you will ever get married/remar-ried?’’ Response categories ranged from 1 (highlyunlikely) to 4 (highly likely). Relationship longevitywas a continuous variable assessing the number ofyears respondents have been in a romantic rela-tionship. The number of years in a romantic rela-tionship ranged from 1 to 60 years. The frequencydistribution of all the dependent variables used inthis analysis is presented in Table 1.

Independent variables. Relationship status wasmeasured by four categories—cohabiting/never mar-ried (referent), cohabiting/previously married, non-cohabiting/previously married, and noncohabiting/never married. It is important to note that relation-ship status measures the respondents’ current statusand that all of the categories are mutually exclusive.We also included two variables to assess parental sta-tus. We measure the parental status of the respondentby the following question: ‘‘How many children haveyou given birth to/fathered?’’ A dummy variable wascreated in order to compare those respondents withno children to those with children. The variable,partner has children from a previous relationship, wasassessed by the following question: ‘‘Does your cur-rent partner have any children from other relation-ships?’’ Respondent’s partner has no children froma previous relationship was the referent and was com-pared to those respondents whose partner has chil-dren from a previous relationship.

Four measures of socioeconomic status wereused: education (measured continuously), employ-ment status—which assessed whether respondentswere employed (referent), not working, or out ofthe labor force—income, and material hardship.Following other social science research, we haveassumed that income has a log-normal distribution.Income was, therefore, transformed by taking the


natural logarithm of income in order to achievea normal distribution. Material hardship was asummary score consisting of seven items assessingwhether or not respondents could meet basic

expenses, pay full rent or mortgage, pay full utili-ties, had utilities disconnected, had telephone dis-connected, were evicted for nonpayment of rent,and could not afford leisure activities in the past

Table 1. Characteristics of the Sample (N ¼ 1,358)a

Variable

African American

(N ¼ 951)

Caribbean Black

(N ¼ 407)

Variable

African American

(N ¼ 951)

Caribbean Black

(N ¼ 407)

Frequency % Frequency % Frequency % Frequency %

Gender Education** (years)

Male 304 40.04 158 49.35 0 – 8 45 4.57 12 6.00

Female 647 59.27 249 50.65 9 – 11 211 21.84 54 13.52

Age** (years) 12 372 39.78 132 31.06

18 – 24 188 23.52 108 30.07 13 – 15 218 22.69 122 29.95

25 – 34 269 27.25 137 35.53 16 64 7.22 52 8.56

35 – 44 231 23.14 80 19.35 171 41 3.90 35 10.92

45 – 54 163 17.00 59 9.95 Material hardship

55 – 64 63 5.63 14 3.31 None 539 57.50 251 61.18

65 – 74 28 2.67 6 0.48 1 142 14.32 61 10.66

751 9 0.79 3 1.31 2 105 11.46 30 13.38

Household income* 3 – 4 124 12.72 46 8.52

Less than $9,999 195 17.97 42 8.45 5 – 7 41 4.00 19 6.25

$10,000 – $19,999 244 21.44 74 20.64 Work status

$20,000 – $29,999 168 17.67 76 16.89 Working 681 71.37 313 74.16

$30,000 – $59,999 258 29.73 140 30.88 Not working 120 12.59 47 10.49

$60,0001 86 13.18 75 23.14 Out of the labor force 150 16.04 47 15.35

Relationship status*** Region***

Cohabiting/never married 166 21.19 97 31.39 Northeast 124 18.24 301 59.60

Cohabiting/previously married 93 11.93 33 10.64 Midwest 173 20.81 7 9.67

Previously married 271 23.62 87 14.54 South 576 50.22 96 23.76

Never married 421 43.26 190 43.44 West 78 10.73 3 6.97

Country of originb Immigration statusc

Spanish Caribbean — — 46 14.62 U.S. born 951 100.00 1,076 94.05

Haiti — — 67 13.43 0 – 10 years — — 83 2.60

Jamaica — — 129 29.88 11 – 20 years — — 86 1.35

Trinidad and Tobago — — 48 8.43 211 years — — 98 2.00

Other English — — 111 33.64 Parental status/partner

Years in relationship* Children 804 56.36 804 56.36

1 162 19.48 80 20.74 No children 554 43.64 554 43.64

2 151 15.81 65 20.61 Parental status/respondent

3 127 13.23 61 9.32 Children 753 75.39 282 67.05

4 – 5 145 15.59 71 15.28 No children 198 24.61 125 32.95

6 – 9 150 15.90 65 18.41 Expectations of marriage***

10 – 19 144 16.02 36 9.10 Highly likely 322 36.83 146 42.82

201 43 3.97 12 6.54 Somewhat likely 331 35.67 173 36.92

Relationship satisfaction Somewhat unlikely 111 10.77 39 9.77

Very satisfied 421 46.60 159 38.32 Highly unlikely 184 16.73 47 10.50

Somewhat satisfied 414 42.60 194 46.13

Somewhat dissatisfied 91 8.91 42 10.82

Very dissatisfied 24 1.89 11 4.73

Note. Significant differences between African Americans and Caribbean Blacks: *p , .05. **p , .01. ***p , .001.aWeighted percentages. bCaribbean-specific variable. cAll African Americans are U.S. born.


12 months. A higher score on this item indicatedgreater material hardship. In addition, two variablesof particular relevance to Black Caribbeans were alsoincluded—immigration status and country of origin.Immigration status has four categories: U.S. born (ref-erent), and those Caribbean Blacks who resided inthe Unites States from 0 to 10 years, 11 to 20 years,and more than 20 years. Black Caribbean respon-dents reported over 25 different countries of origin,which were recoded into five categories: Jamaica (ref-erent), Spanish-speaking country (e.g., DominicanRepublic, Cuba, Puerto Rico), Haiti, Trinidad andTobago, and other English-speaking country (e.g.,Barbados). Other demographic variables included inthis analysis were ethnicity (0 ¼ African American;1¼ Caribbean Black), gender (0¼ male; 1¼ female),region (0 ¼ South for African Americans; 0 ¼ South,North Central, and West for Black Caribbeans), andage (measured continuously). The present analysiswas based on complete data for 1,362 African Amer-ican and Black Caribbean respondents (181 years)who were not currently married but indicated thatthey had a ‘‘main romantic’’ relationship for at least1 year. The demographic characteristics of this sam-ple are provided in Table 1.

Analytic Strategy

Linear regression analyses were conducted usingSUDAAN 9.0. To obtain results that are generaliz-able to the African American and Black Caribbeanpopulation, all the analyses incorporated the sam-ple’s race-adjusted weights. Weights in the NSALdata account for unequal probabilities of selection,nonresponse, and poststratification such that re-spondents are weighted in accordance with theirnumbers and proportions in the full population. Inaddition, all analyses accounted for the complexmultistage clustered design of the NSAL sample(i.e., clustering and stratification) when computingstandard errors.

Results

Relationship Satisfaction

Overall, respondents indicated that they were satis-fied with their relationship, with about 46% of theentire sample indicating that they were very satisfiedand only 11% indicating that they were dissatisfied.

There was no significant difference in relationshipsatisfaction between African Americans and BlackCaribbeans (analysis not shown). Among AfricanAmericans, those reporting material hardship wereless satisfied with their relationship compared totheir counterparts (Table 2, Model 1). Among BlackCaribbeans (Table 2, Model 1), gender, education,material hardship, parental status, relationship sta-tus, and country of origin were significantly relatedto relationship satisfaction. Black Caribbean womenand those with higher levels of education were moresatisfied with their relationship than men and thosewith less education. Material hardship decreasedrelationship satisfaction. Those with children wereless satisfied with their relationships compared tothose who had no children. Never-married BlackCaribbeans were less satisfied with their relationshipthan their counterparts who were cohabiting/nevermarried. Black Caribbeans from other English-speaking countries had lower levels of relationshipsatisfaction than those from Jamaica.

Expectations of Marriage

Nearly 4 of 10 respondents (37.26%) reported that itis highly likely that they will ever marry; 27%believed it is unlikely that they would ever marry.There were no significant differences between AfricanAmericans and Caribbean Blacks on this measure(analysis not shown). Among African Americans,region, age, gender, and income were significantlyassociated with the expectations of marriage (Table 2,Model 2). Lower expectations were held by women,older respondents, those who resided in the West(compared to the South), and those with less income.Among Black Caribbeans, age, relationship status,and parental status were significantly associated withthe expectations of marriage (Table 2, Model 2), withlower expectations expressed by those who wereolder, previously married, and those whose partnershad no children from a previous relationship (com-pared to those with children from a previousrelationship).

Relationship Longevity

The number of years in a romantic relationship variedfrom 1 to 60 years. The majority of the relationshipswere for a short duration, but more than one third ofthe relationships (35.76%) were for six or more years(Table 1). African Americans and Caribbean Blacks


Table2.Standardized

Regression

Estim

atesoftheEffectsofC

ovariatesonRelationshipVariables:A

frican

Americansand

Caribbean

Blacks

Variable

Relationship

Satisfaction

(Model1)

Expectation

sof

Marriage(M

odel2)

Yearsin

Relationship

(Model3)

African

Americans

(N¼

932)

CaribbeanBlacks

(N¼

390)

African

Americans

(N¼

930)

CaribbeanBlacks

(N¼

389)

African

Americans

(N¼

931)

CaribbeanBlacks

(N¼

391)

BSE

BSE

BSE

BSE

BSE

BSE

Region

Northeast

20.05

0.03

20.07

0.06

20.03

0.03

20.07

0.07

0.48

0.27

0.41

0.40

North

Central

20.06

0.04

——

0.05

0.05

——

0.21

0.33

——

West

0.01

0.03

——

20.07*

0.03

——

20.15

0.23

——

Age

20.03

0.03

0.08

0.06

20.37***

0.04

20.20*

0.08

2.27***

0.30

2.90***

0.29

Gender

Fem

ale

20.04

0.03

0.11**

0.04

20.13**

0.04

20.06

0.05

0.32

0.25

0.35

0.36

Education

20.04

0.04

0.05*

0.02

0.06

0.04

0.01

0.02

0.02

0.28

20.41**

0.14

Materialhardship

20.14***

0.04

20.14***

0.03

0.00

0.04

20.03

0.04

0.04

0.26

0.20

0.13

Respon

denthaschildren

20.04

0.03

20.15***

0.04

0.07

0.04

20.07

0.06

0.58*

0.24

0.47*

0.17

Partnerhaschildren

0.02

0.02

0.01

0.06

20.02

0.04

0.12*

0.05

20.46**

0.17

20.26

0.27

Relationship

status

Cohabiting/previouslymarried

0.01

0.03

20.11

0.06

0.06

0.05

20.10

0.06

20.10

0.29

0.20

0.16

Previouslymarried

20.04

0.04

20.05

0.03

20.07

0.05

20.11**

0.04

20.94**

0.32

20.44*

0.16

Nevermarried

20.06

0.03

20.14***

0.03

20.01

0.04

20.05

0.05

0.17

0.26

0.61**

0.23

Employm

entstatus

Unem

ployed

0.05

0.03

20.02

0.04

20.05

0.05

20.06

0.04

0.08

0.21

20.02

0.09

Outof

labor

force

0.03

0.03

0.00

0.03

20.05

0.04

0.10

0.06

0.03

0.17

0.67

0.45

Income

20.00

0.04

20.04

0.05

0.12*

0.06

0.01

0.06

20.06

0.29

20.28

0.15

Countryof

origin

SpanishCaribbean

——

20.02

0.02

——

20.03

0.03

——

20.17

0.21

Haiti

——

0.07

0.04

——

20.04

0.07

——

0.49*

0.19

Trinidad

andTobago

——

20.02

0.02

——

20.01

0.03

——

20.02

0.11

OtherEnglish

——

20.07*

0.03

——

20.06

0.07

——

20.32

0.29

Immigration

status(years)

0–10

——

0.08

0.04

——

0.04

0.05

——

20.00

0.26

11–20

——

0.04

0.02

——

20.00

0.02

——

0.17

0.15

211

——

20.01

0.03

——

20.07

0.04

——

0.00

0.34

Expectation

——

——

——

——

20.12

0.18

0.05

0.26

Satisfaction

——

——

——

——

20.27

0.20

20.70**

0.15

R2

0.054

0.356

0.192

0.248

0.148

0.509

Note.Excluded

categoriesforregion

(0¼

SouthforAfrican

Americans;0¼

South,North

Central,and

WestforCaribbean

Blacks),gender(0

¼male),relation

ship

status(0

¼cohabiting/nevermarried),em

ploym

entstatus(0

¼em

ployed),respon-

denthaschildren(0

¼no

children),partnerhaschildrenfrom

apreviousrelation

ship

(0¼

nochildren),countryof

origin

(0¼

Jamaica),andim

migration

status(0

¼U.S.born).

*p,

.05.**p,

.01.***p,

.001.


did not differ on this measure in regression analyses(not shown). To identify those relationship factorsthat are associated with longevity, we added the varia-bles relationship satisfaction and expectations of mar-riage to the model.

Among African Americans, age, parental status,and relationship status were significantly associatedwith the number of years in a romantic relation-ship (Table 2, Model 3), with longer relationshipsobserved among older African Americans, those withchildren, persons whose partner had no childrenfrom a previous relationship, and those who werecohabiting/never married (compared to those whowere previously married). Among Black Caribbeans,age, education, parental status, relationship status,country of origin, and relationship satisfaction weresignificantly associated with the number of years ina romantic relationship (Table 2, Model 3). OlderBlack Caribbeans, those with children, and thosewith less education had longer relationships com-pared to their counterparts. Those who were pre-viously married (not cohabiting) had shorterrelationships than those who cohabit/never mar-ried, whereas those who were never married (notcohabiting) had longer relationships than thosewho cohabit/never married. Regarding country oforigin, Black Caribbeans from Haiti had longernonmarital relationships than those from Jamaica.Finally, relationship satisfaction was negativelyassociated with relationship longevity. Specifically,Black Caribbeans who were less satisfied with theirrelationships spent more years in a nonmaritalromantic relationship compared to those who weremore satisfied.

Discussion

The goal of this paper was to shed light on thenature of romantic unions among unmarried AfricanAmericans and Caribbean Blacks by identifying thecorrelates of relationship satisfaction, expectations ofmarriage, and relationship longevity. Overall, thefindings indicated that there were no significant dif-ferences between African Americans and Black Car-ibbeans in these relationship outcomes. However,the pattern of relationships for the two groups dif-fered. In the remainder of this section, we will dis-cuss the findings for each of the dependent variablesseparately.

Relationship Satisfaction

The vast majority of respondents reported thatthey were either very or somewhat satisfied withtheir current romantic relationship. For both Afri-can Americans and Black Caribbeans, materialhardship was associated with decreased levels ofrelationship satisfaction. This finding is consistentwith previous research on African Americans indi-cating that financial strain (Cutrona et al., 2003)and lower incomes (Bryant & Wickrama, 2005)were associated with reduced levels of marital qual-ity. The present findings, in combination withprevious work on married samples, indicate thatmaterial hardship is detrimental to relationshipsatisfaction among both married and unmarriedcouples.

Among African Americans, gender was not sig-nificantly associated with relationship satisfaction,whereas among Black Caribbeans, women weremore satisfied with their relationship compared tomen. Both findings are inconsistent with priorresearch on marital satisfaction, which has foundhusbands to be more satisfied with their marriagesthan wives (e.g., Dillaway & Broman, 2001). Thefindings of the present analysis could be attributableto several factors. First, men are perceived to receivethe majority of the benefits of marital relationships,such as reduced levels of household and childcareresponsibilities. Nonmarital unions, however, maybe more egalitarian than marriages with respect tothe division of household labor and child care.According to economic explanations of maritalbehavior, women in nonmarital unions may havemore economic leverage and status and are, thus,able to negotiate more egalitarian arrangements. Sec-ond, it is important to consider the broader demo-graphic and social contexts within which bothmarital and nonmarital unions occur. Given that theoverall marriage rates of African Americans are low,many women may express higher levels of satisfac-tion with nonmarital romantic relationships becausethe alternative is not being in a relationship of anykind (Jayakody & Cabrera, 2002). In essence, thebroader marriage market for African Americanwomen may function to ‘‘inflate’’ their reports ofsatisfaction with nonmarital relationships (render-ing them indistinguishable from those of AfricanAmerican men).

Among Black Caribbeans, women indicatedhigher levels of satisfaction with their romantic


relationship than men. Unfortunately, there is verylittle previous literature on marital relationshipsamong Black Caribbeans, much less nonmaritalunions. In an analysis of married Jamaican womenwho migrated to New York, Foner (2005) foundthat they were satisfied with their marital relation-ships. Her analysis suggests that this was becausecouples who migrated from Jamaica spent moretime together in ‘‘couple’’ activities after they arrivedin the United States. In addition, once in the UnitedStates, the household roles were more egalitarian.Other analyses of the NSAL data indicated thatamong married respondents, Black Caribbean wo-men had higher rates of marital satisfaction thanAfrican American women. Collectively, these find-ings indicate that both married and unmarried BlackCaribbean women have higher levels of relationshipsatisfaction than Black Caribbean men.

Caribbean Blacks with children were less satis-fied with their relationships compared to thosewithout children. The presence of children oftenincreases the level of stress in a relationship anddecreases relationship happiness among cohabitors(Brown, 2003). In particular, having children isassociated with low levels of happiness in longcohabiting unions, perhaps because nearly most ofthese unions involve children from prior unions.In our study, 75% of respondents have childrenand more than 56% of their partners have chil-dren from a previous relationship. It is likely thatlevels of satisfaction are impacted by both thenumber of children between the couples and theresponsibilities that parents have to their biologicalchildren and any involvement with the parentfrom their previous relationship.

Caribbean Blacks from Jamaica had higher levelsof satisfaction with their relationships than thosefrom other English-speaking countries (e.g., Barbados,Bahamas). This finding is consistent with the ethno-graphic work of Foner (2005) on the high levels ofsatisfaction among Jamaican couples. This findingalong with other analysis of these data reveals thatboth married and unmarried Jamaicans note relativelyhigher levels of satisfaction with their relationships.

Expectations of Marriage

The majority of the respondents indicated that itwas highly or somewhat likely that they would getmarried or remarried. This is consistent with previ-ous research indicating that, despite low rates of

marriage, the majority of African Americans havea strong belief in the importance of marriage (Har-knett & McLanahan, 2004). Among both AfricanAmericans and Black Caribbeans, age was negativelyassociated with the expectations of marriage. This isconsistent with the notion that older respondents,who have considerable experience in relationships,are content with their current situation and do notwant to have the legal and social obligations of mar-riage. Conversely, younger adults likely view mar-riage as the next stage in the life course.

African American women had lower expectationsof marriage than African American men. Consistentwith economic explanations of marital behavior, thelower socioeconomic status of many African Ameri-can men may make them less attractive as potentialmarriage partners but suitable for a romantic rela-tionship. This finding contradicts research by South(1993) who found that African American men hada lower desire of marriage than African Americanwomen. This discrepancy is probably because of themajor differences in study samples; South’s samplewas restricted to young (ages 19 – 35 years), unmar-ried noncohabiting African Americans, whereas thisstudy includes all unmarried African Americanadults who are currently in a romantic relationship.

Among African Americans, those who reside inthe South had higher expectations that their currentrelationship will lead to marriage than residents ofthe West. This could be because of the more tradi-tional and more religious culture of the South(Taylor, Chatters, & Levin, 2004), where marriagewould be a more expected outcome of a romanticrelationship.

African Americans with higher incomes hadgreater expectations of marrying than their lowersocioeconomic status counterparts. This finding isconsistent with research on the importance of eco-nomic stability as a precursor to marriage (Sweeney,2002). This finding, in conjunction with the find-ings for relationship satisfaction among AfricanAmericans (Table 2), further demonstrates theimportance of economic factors as a precursor tomarriage among this group.

In comparison to Caribbean Black adults whoare in cohabiting/never-married relationships, thosewho have been previously married had lower expect-ations of marriage. This finding is consistent withresearch and general observations that previouslymarried adults are much more cautious about enter-ing into another marital union.


Relationship Longevity

There was considerable heterogeneity in the numberof years that respondents were in nonmarital unions,ranging from 1 to 60 years. For both African Ameri-cans and Caribbean Blacks, previously married(noncohabiting) respondents were in relationshipsthat were of shorter duration than respondents whowere cohabiting/never married. It is possible thatrespondents who have experienced marriage in thepast may be less invested in long-term nonmaritalrelationships. Conversely, many enter cohabitationwith the intent to marry. Thus, they are more likelyto have invested heavily in the relationship (bothmaterially and emotionally), which may help main-tain their relationships longer (Rusbult & Buunk,1993).

Our data show that African Americans andCaribbean Blacks with children spend more time ina nonmarital union compared to those with no chil-dren. This finding is consistent with studies examin-ing the role of children in the transition to marriageand relationship stability. Wu (1995) found thatcohabitating couples with children in the relation-ship are less likely to experience union disruptionthan couples without children. The financial respon-sibilities associated with raising children may be animpediment to marriage while at the same timeencouraging a commitment to a nonmarital union.

In contrast, African Americans whose partnerhad children from a previous relationship hadshorter relationships than those whose partner hadno children from a previous relationship. Becausewe are unable to gauge whether the relationship wasterminated or ended in marriage, we can only specu-late as to the meaning of this finding. Research indi-cates that if a mother has a child by a previousrelationship, it deters a visiting relationship but hasno effect on the decision to cohabit or marry. How-ever, a father’s children from a previous relationshipdeter cohabitation and marriage with their currentpartner but have no effect on visiting (Carlson,McLanahan, & England, 2004). So, it is likely thatthe father’s children from a previous relationshipmay have some bearing on the termination of therelationship, which would explain our finding ofshorter relationships for those African Americanswhose partner had children from a previous relation-ship. Future research should explore whether havingchildren compromises a man’s future marital andnonmarital relationships because of child support

obligations, his likely disinclination to commit toone woman, or because other women are wary of hispotential continued romantic involvement with themother of his children.

Among Caribbean Blacks only, fewer years ofeducation was associated with more years in a roman-tic relationship. Additionally, never-married respon-dents reported more years in their relationship,whereas previously married respondents reportedfewer years in their relationship (compared to cohab-iting/never-married Caribbean Blacks). The findingfor education is consistent with socioeconomicexplanations of marital behavior among CaribbeanBlacks. Studies examining the impact of educationalattainment and income on the decision to marryindicate that marriage is more likely to occur amongthose with greater economic stability and highersocial ranking (Goode, 1960).

Caribbean Blacks who were less satisfied withtheir relationship maintained their romantic rela-tionships longer than those who were more satisfied.This finding is consistent with studies of relationshipquality among cohabiting couples. Findings fromthese studies indicate that duration of a relationshiphas a negative impact on relationship happiness(Brown, 2003).

Interpretation of these findings should be consid-ered within the context of the study’s strengths andlimitations. First, the Black Caribbean sampleexcluded individuals who did not speak English;therefore, the study findings are not generalizablebeyond this group. Another limitation involves theuse of single items to measure our dependent varia-bles. Although there are occasions when a singleitem is insufficient for representing a particular con-struct (e.g., personality), they are suitable for narrowor otherwise unambiguous constructs (Sackett &Larson, 1990). In the current analysis, the depen-dent variables are unambiguous constructs that aresufficiently narrow and easily understood by respon-dents. Our results indicate that the measured itemshave face validity and function within the model aswould be theoretically expected. Finally, becausethese data are cross-sectional, it is impossible todetermine whether those respondents in a shorterterm relationship will end up in a longer term rela-tionship in the future. Admittedly, this is one of thelimitations associated with using cross-sectionaldata. Future studies using longitudinal data are nec-essary to make this distinction. Despite these limita-tions, this study provides some unique insights into


nonmarital romantic relationships among AfricanAmericans and Black Caribbeans.

Implications for Practice, Policy, and Research

The increasing prevalence of nonmarital romanticrelationships in the United States poses a challengefor policymakers regarding how to promote mar-riage while simultaneously protecting the rights ofindividuals who are in unions with no intent tomarry. Given the current trend, various governmen-tal and institutional entities in the United Statesmay eventually have to consider and debate the mer-its of providing unmarried relationships and fami-lies’ social and legal standing comparable to thatcurrently accorded to those who are married. ManyLatin American countries have long histories ofsocially accepted consensual unions, which may sub-stitute for formal unions in some groups (De Vos,1999). Laws about taxes, housing, and child supporttreat unmarried and married couples the same inSweden, where premarital cohabitation is practicallyuniversal (Hoem, 1995). Countries like Canada,France, and Sweden have already changed their legalcodes so that references to ‘‘spouse’’ also pertain tounmarried partners and so that partners who meetcertain criteria can gain ‘‘marital’’ rights as domesticpartners. Other countries like Australia, Mexico,South Africa, and the U.K. have either taken initialsteps toward broader recognition or are currentlyengaged in national debates about how to best rec-oncile the gap between current existing family con-figurations and those defined in legal terms. Incontrast, family law in the United States does notgive the rights of married couples to those who areunmarried, despite the increasing numbers ofromantic unions that do not end in marriage.

Some of the problematic issues about providingthe same legal rights to unmarried couples and theirfamilies as married individuals are distinguishingbetween (a) those who would prefer marriage, buttheir economic circumstances prevent them frommarrying; (b) those who seek a different type of rela-tionship, that is, an alternative to marriage; and (c)those who are using their relationship as a trialperiod during which they assess whether marriage isthe desirable outcome. Another problematic issue isthe potential for fraud. Married couples can provetheir legal status with a marriage license, but it ismuch more difficult to prove that an unmarriedcouple is in a committed relationship. Various forms

of documentation can be used, either separately orin combination, to prove long-term commitmentincluding, for example, residing together for sevenor more years, being beneficiaries of insurance poli-cies, being a beneficiary on each others will, or beinglegal co-owners of a place of residence. The hetero-geneity of nonmarital romantic relationships as illus-trated by our findings, coupled with the substantialvariation across states in the availability of domesticpartner registration, the eligibility rules, and thebenefits and responsibilities of registration demon-strate public disagreement about the meaning ofcohabiting and noncohabiting romantic relation-ships and their place in the U.S. family system. Theincrease in nonmarital romantic relationships amongAfrican Americans and Americans in general, overthe past few decades, suggests that these relationshipsmay become more formalized institutions in thefuture, but it is unlikely that they will have the pref-erential standing of marriage.

Evidence does suggest that unmarried couplesand families can benefit from the same or similarservices as married couples and families. A recentstudy indicates that unmarried couples may alsobenefit from relationship education (Carlson &McLanahan, 2006). Specifically, findings indicatethat higher levels of supportiveness in the couples’relationship (e.g., affection, encouragement) andlower levels of conflict (e.g., criticism, arguments)resulted in more positive parenting behaviors. Thesefindings suggest that interventions should focus onboth increasing supportive exchanges and reducingnegative interactions between couples. This observa-tion is underscored by studies of negative interac-tions among Black Americans (Lincoln, Chatters, &Taylor, 2003; Lincoln, Taylor, & Chatters, 2003).

Rarely do researchers acknowledge that unmar-ried adults in romantic relationships may not per-ceive themselves as part of the marriage process. Themajor finding from this investigation is the observeddiversity in the longevity and relationship character-istics of nonmarital unions. With lower rates of mar-riage overall, we will likely see more variety inrelationship types and the sequencing process. Socio-economic status was an especially strong predictor ofrelationship quality and stability for CaribbeanBlacks. Those with more education are more satisfiedand have longer relationships than their counterparts.Similarly, those Caribbean Blacks with more incomehave higher expectations of marriage. These findingssupport those of other studies reporting an association


between economic factors and marital quality and sta-bility. Overall, our findings are largely consistent withfindings from studies of married couples (e.g.,Cutrona et al., 2003). That is, our findings suggestthat the correlates of relationship satisfaction, expecta-tions of marriage, and relationship longevity for Afri-can Americans and Black Caribbeans in nonmaritalrelationships may not be markedly different fromthose factors that predict quality and stability for mar-ried couples. There was one important distinction,however. Neither African American nor Black Carib-bean men displayed the greater relationship satisfac-tion levels generally observed in marital research.

Overall, the findings of this study are consistentwith other work in affirming that the traditionalnotion of short-term courtships, followed by a long-term marriage, is no longer normative for a signifi-cant number of Americans. Lower rates of marriageand remarriage invariably increase the length of timein adulthood spent unmarried. Consequently, it isimperative that family researchers expand theirresearch to include the entire spectrum of long-termrelationships—including marriage, cohabitation,and involvements among those who do not reside inthe same household.

References

African American Healthy Marriage Initiative. (2000). Marriage, divorce,childbirth, and living arrangements among African American or Black pop-ulations. Washington, DC: Administration for Children and Families,U.S. Department of Health and Human Services.

Brown, S. L. (2003). Relationship quality dynamics of cohabiting unions.Journal of Family Issues, 24, 583–601.

Bryant, C. M., & Wickrama, K. A. S. (2005). Marital relationships of Afri-can Americans: A contextual approach. In V. McLoyd, N. Hill, &K. A. Dodge (Eds.), African American family life: Ecological and culturaldiversity (pp. 111–134). New York: Guilford Press.

Bumpass, L. L., & Lu, H. (2000). Trends in cohabitation and implicationsfor children’s family context in the U.S. Population Studies, 54, 29–41.

Carlson, M., McLanahan, S., & England, P. (2004). Union formation infragile families. Demography, 41, 237–261.

Carlson, M. J., & McLanahan, S. S. (2006). Strengthening unmarried fam-ilies: Could enhancing couple relationships also improve parenting?Social Service Review, 80, 297–376.

Clark-Nicolas, P., & Gray-Little, B. (1991). Effect of economic resourceson marital quality in Black married couples. Journal of Marriage andFamily, 53, 645–655.

Coppin, A. (2000). Does type of conjugal union matter in the labor mar-ket? Evidence from a Caribbean economy. Review of Black PoliticalEconomy, 28, 7–27.

Cutrona, C. E., Russell, D. W., Abraham, W. T., Gardner, K. T.,Melby, J. N., Bryant, C., & Conger, R. D. (2003). Neighborhoodcontext and financial strain as predictors of marital interaction andmarital quality in African American couples. Personal Relationships,10, 389–409.

De Vos, S. (1999). Comment of coding marital status in Latin America.Journal of Comparative Family Studies, 31, 79–93.

Dillaway, H., & Broman, C. L. (2001). Race, class, and gender differencesin marital satisfaction and divisions of household labor among dual-earner couples: A case for intersectional analysis. Journal of FamilyIssues, 22, 309–327.

Foner, N. (2005). In a new land: A comparative view of immigration. NewYork: New York University Press.

Goode, W. (1960). Illegitimacy in Caribbean social structure. AmericanSociological Review, 25, 21–30.

Harknett, K., & McLanahan, S. S. (2004). Racial and ethnic differences inmarriage after the birth of a child. American Sociological Review, 69,790–811.

Henriques, F. (1953). Family and colour in Jamaica. London: Macgibbon& Kee.

Hoem, B. (1995). Sweden. In H.-P. Blossfeld (Ed.), The new role of women:Family formation in modern societies (pp. 35–55). Boulder, CO: West-view Press.

Jackson, J. S., Torres, M., Caldwell, C. H., Neighbors, H. W., Nesse, R.M., Taylor, R. J., Trierweiler, S. J., & Williams, D. R. (2004). TheNational Survey of American Life: A study of racial, ethnic and culturalinfluences on mental disorders and mental health. International Journalof Methods in Psychiatric Research, 13, 196–207.

Jayakody, R., & Cabrera, N. (2002). What are the choices for low incomefamilies? Cohabitation, marriage, and remaining single. In A. Booth &A. Crouter (Eds.), Just living together: Implications of cohabitation forchildren, families, and social policy (pp. 85–96). Mahwah, NJ: LawrenceErlbaum.

Kielcolt, K. J., & Fossett, M. A. (1995). Mate availability and marriageamong African Americans: Aggregate and individual-level analyses. InM. B. Tucker & C. Mitchell-Kernan (Eds.), The decline of marriageamong African Americans (pp. 121–135).New York: Russell Sage.

Lincoln, K. D., Chatters, L. M., & Taylor, R. J. (2003). Psychological dis-tress among Black and White Americans: Differential effects of socialsupport, negative interaction and personal control. Journal of Healthand Social Behavior, 44, 390–407.

Lincoln, K. D., Taylor, R. J., & Chatters, L. M. (2003). Correlates of emo-tional support and negative interaction among Black Americans. Jour-nals of Gerontology: Social Sciences, 58B, S225–S233.

Logan, J. R., & Deane, G. (2003). Black diversity in metropolitan America.Report from the Mumford Center for Comparative Urban andRegional Research. Retrieved June 1, 2006, from State University ofNew York, Albany, NY Website: http://mumford1.dyndns.org/cen2000/BlackWhite/BlackDiversityReport/black-diversity01.htm

Matthews, D. B. (1952). Crisis of the West Indian family: A sample study.Westport, CT: Greenwood Press.

McKenzie, H. (1993). The family, class and ethnicity in the future of theCaribbean. In Race, class and gender in the future of the Caribbean.Mona, Jamaica: Institute of Social and Economic Research, The Uni-versity of the West Indies.

Raley, R. K., & Bumpass, L. L. (2003). The topography of the divorce pla-teau: Levels and trends in union stability in the United States after1980. Demographic Research, 8, 245–259.

Rodman, H. (1971). Lower class families: Culture of poverty in Negro Trini-dad. New York: Oxford University Press.

Ruggles, S. (1997). The rise of divorce and separation in the United States,1880–1990. Demography, 34, 455–466.

Rusbult, C. E., & Buunk, B. P. (1993). Commitment processes in closerelationships: An independence analysis. Journal of Social and PersonalRelationships, 10, 175–204.

Sackett, P. R., & Larson, J. R. (1990). Research strategies and tactics inindustrial and organizational psychology. In M. D. Dunnette & L. M.Hough (Eds.), Handbook of industrial and organizational psychology(pp. 419–428). Palo Alto, CA: Consulting Psychologists Press.

Schoen, R., & Owens, D. (1992). A further look at first marriages and firstunions. In S. South & S. Tolnay (Eds.), The changing American family:Sociological and demographic perspectives (pp. 109–117). Boulder, CO:Westview.

Smith, R. (1956). The Negro family in British Guiana: Family structure andsocial status in villages. London: Routledge & Kegan Paul.

Smock, P. J. (2000). Cohabitation in the United States: An appraisal ofresearch themes, findings, and implications. American Review ofSociology, 26, 1–20.


St. Bernard, G. (2003). Major trends affecting families in Central Americaand the Caribbean. Paper prepared for the United Nations Division ofSocial Policy and Development Department of Economic and SocialAffairs Program on the Family.

South, S. J. (1993). Racial and ethnic differences in the desire to marry.Journal of Marriage and the Family, 55, 357–370.

Sweeney, M. M. (2002). Two of family change: The shifting economicfoundations of marriage. American Sociological Review, 67, 132–147.

Taylor, R. J., Chatters, L. M., & Levin J. (2004). Religion in the lives of Afri-can Americans: Social, psychological and health perspectives. ThousandOaks, CA: Sage Press.

Taylor, R. J., Tucker, M. B., Chatters, L. M., & Jayakody, R. (1997).Recent demographic trends in African American family structure.

In R. J. Taylor, J. S. Jackson, & L. M. Chatters (Eds.), Family life inBlack America (pp. 14–62). Thousand Oaks, CA: Sage Publications.

Tucker, M. B., & Taylor, R. J. (1997). Gender, age and marital status asrelated to romantic involvement among African American singles. In R.J. Taylor, J. S. Jackson, & L. M. Chatters (Eds.), Family life in BlackAmerica (pp. 79–94). Thousand Oaks, CA: Sage Publications.

U.S. Census Bureau. (2005a). Census 2000 Special Report, We the people:Blacks in the United States. Washington, DC: Author.

U.S. Census Bureau. (2005b). Census 2000, Summary File 3, MatricesPCT15 and PCT1. Washington, DC: Author.

Waters, M. (1999). Black identities. Cambridge, MA: Harvard University Press.Wu, Z. (1995). The stability of cohabitation relationships: The role of chil-

dren. Journal of Marriage and the Family, 57, 231–236.


romantic relationships among unmarried african americans and caribbean blacks: findings from the...

Documents