Вы находитесь на странице: 1из 27

A Path toward Interracial MarriageRosalind Blackwell Publishing Ltd.Oxford, UK and Malden, USATSQThe Sociological Quarterly0038-02532007 Blackwell Publishing Ltd.

2007482343369RACE IN THE PUBLIC AND PRIVATE REALMS King and Jenifer L. Bratter

Berkowitz

The Sociological Quarterly ISSN 0038-0253

A PATH TOWARD INTERRACIAL MARRIAGE: Womens First Partners and Husbands across Racial Lines
Rosalind Berkowitz King*
National Institutes of Health

Jenifer L. Bratter*
Rice University

We examine interracial marriage as a culminating event in a sequence of intimate relationships across the life course. Using data from the 1995 National Survey of Family Growth, we analyze the background characteristics associated with selecting a rst sex partner and rst husband who differ in race/ethnicity from the respondent as well as the continuity across both outcomes. Our results show that respondents race/ethnicity, parents education, and region of birth are signicant predictors of both choices. Selecting partners across racial lines for rst sex is signicantly associated with the selection of a rst husband across race; the association between both outcomes is particularly strong for non-Hispanic black women, implying that social integration across race may be a life course phenomenon.

INTRODUCTION The increasing frequency of marriage across the color line is considered evidence of a larger cultural shift in American thinking on the permeability of racial/ethnic boundaries in social interaction (Portereld 1982; Spickard 1989; Root 2001). Although race continues to be salient to selecting partners in several types of intimate relationships (e.g., Laumann et al. 1994), our knowledge of interracial relationships comes largely from the investigation of one union, marriage, in isolation from others. Sociological insights into interracial mate selection emphasize the role of de facto marriage markets that operate in tandem with shifting attitudes toward cross-race interactions (Lee and Yamanaka 1990; Qian 1997; Fu 2001). Previous empirical studies have not focused on experiences preceding the marriage that may be linked to or predictive of the choice to marry interracially. We are aware of only a single study (Lewis, Yancey, and Bletzer 1997) that incorporates premarital dating histories when looking at interracial marriage, but their analyses of black/white married couples limit the use of premarital dating history to a control

*Direct all correspondence to Rosalind Berkowitz King, Demographic and Behavioral Sciences Branch, Center for Population Research, National Institute of Child Health and Human Development, National Institutes of Health, U.S. Department of Health and Human Services, 6100 Executive Blvd., Room 8B07, Bethesda, MD 20892-7510, e-mail: rozking@mail.nih.gov; to Jenifer L. Bratter, Department of Sociology, MS-28, Rice University, 6100 S. Main Street, Houston, TX 77005-1892; email: jlb1@rice.edu
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

343

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

variable and their sample is selective. Our research makes a unique contribution by linking the racial/ethnic homogeneity of two signicant partnering choices within an individuals life experience in a nationally representative sample of women. We step back from examining the characteristics of already married couples and instead focus on background and experience leading up to a womans choice of husband. We contend that marrying across racial lines should be viewed within a larger context of partnering choices. The choice to marry interracially may reect previous partnering choices grounded in an underlying propensity to seek mates outside of ones own racial group. We believe that this willingness is more likely to emerge from otherwise liberal or tolerant upbringings that inform both selecting the rst sexual partner and the rst spouse, thus justifying a departure from the norm of endogamous mate selection. In this study, we explore factors that are exogenous to two partnership choicesrst sexual partner and rst husbandusing a nationally representative data set of American women. We advance three research goals. First, we explore how a womans background affects her likelihood of choosing across the lines of race and ethnicity for her rst sexual relationship and which background characteristics among these also predict an interracial rst marriage. Second, we examine the potential continuity between selecting an interracial rst partner and selecting an interracial rst husband. Essentially, we ask whether or not selecting a partner across racial lines earlier in the life course predicts such a choice when marrying for the rst time. Finally, we explore how this continuity varies for different racial/ethnic groups. This study thus adds to the growing literature on interracial marriage by examining the association between premarital experiences and actual marital choices made later in life. In addition, our analysis expands the purview of methods and theory used to examine interracial marriage by employing event history models and a life course perspective. BACKGROUND AND SIGNIFICANCE Interracial marriage stands as the universal indicator of social proximity between the races. Residential and occupational integration by race are important steps toward assimilation (see Gordon 1964), but marriage across the color line implies a greater level of intimacy and social integration. Although still small in absolute terms, levels of intermarriage have risen sharply in recent years. In 1970, approximately 321,000 interracial marriages appeared in U.S. Census data. That number rose to about 1 million in 1980, 1.5 million in 1990, and 2.9 million in 2000 (Farley 2002). However, the prevalence of these marriages is small compared to what would be observed if men and women married across race at random (Schoen and Wooldredge 1989; Harris and Ono 2005). While many scholars of interracial relations emphasize the roles of growing racial diversity (Pollard and OHare 1999) and increasing importance of achieved traits such as earned education (Kalmijn 1991a,b, 1993, 1994, 1998; Heaton and Albrecht 1996; Qian 1997; Fu 2001), accounting for these factors does not explain why interracial marriage remains a rare event.

344

The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

Others view the prevalence of interracial marriage as a shift in thinking about race. One line of inquiry, derived from contact theory (e.g., Allport 1954), points to the relevance of previous interracial contacts for altering the propensity toward engaging in future interracial experiences. Emerson, Kimbro, and Yancey (2002) applied this framework to study how previous interracial contacts narrow racial social distance by generating diverse social networks. In extending their work, contact theory could be similarly useful for a study of interracial marriage. This inquiry would require situating interracial marriage within a longitudinal framework where previous interracial contact would predict a later interracial marriage. One noted criticism of this hypothesis is that selection effects may be present. That is, individuals may have preexisting attitudes that predispose them to seek out or avoid persons of other races. Powers and Ellison (1995) attempted to test for selection bias in contact and found no evidence to reject the assumption that these propensities are exogenous to the contact itself. Their discussion notes that longitudinal data may be useful for further addressing this issue. Toward a Life Course Approach for the Study of Interracial Interaction A life course perspective suggests that propensities to marry across racial lines would likely develop before adulthood. Social and cultural similarities between spouses do not emerge in a vacuum. Patterns of homogamy on education, class, and religion can be traced back to shared organized settings where couples rst meet (Kalmijn and Flap 2001). Although most individuals tend to live, worship, and go to school in racially segregated settings (Massey and Denton 1993; Emerson and Smith 2000; Moody 2001), contexts such as schools, neighborhoods, and churches commonly set the stage for future cross-racial relationships when those settings are integrated (Kouri and Lasswell 1993; Yancey 2002; Kouri 2003). The social patterning of interracial relationships in adolescence is an important component of the pattern of choices observed in adulthood, and previous research has shown that premarital dating behaviors are relevant to future mating behavior. Romantic relationships during adolescence signify progress toward autonomy and independence from parents (Gray and Steinburg 1999). These partner choices may serve as the rst signicant chance to adhere to or reject parentally suggested preferences. As is the case in marriage, most adolescent romantic relationships occur between partners of the same race (Ford, Sohn, and Lepkowski 2001, 2003). Contextual factors such as the racial composition of schools and neighborhoods play a role in shaping tendencies toward interracial relationships (Mouw and Entwisle 2001; Quillian and Campbell 2003). But Joyner and Kao (2000) show that the racial composition of potential friends in school does not entirely explain whether or not an adolescent participates in an interracial friendship. They nd that the prevalence of interracial friendships largely mirrors that of the current interracially married population. For example, Hispanics and Native Americans are more likely to have an interracial friendship than Whites are, even within the same opportunity structure.

The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

345

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

Background Characteristics of Interracial Partnering Given the requisite racial social distance between adolescents, what factors might predict selecting a partner across racial lines? Children do not choose their social contexts. Rather, parents determine what school a child will attend or the neighborhood in which they will live. Contexts that generate diminished propensities toward ascribed homogamy and greater racial tolerance may predict later development of an interracial partnership. Parental social class and religious afliation may be particularly important contexts of this type. Education is one of the most powerful predictors of racial attitudes, with increases in education correlated with increasingly tolerant stances toward interracial interaction (Schuman, Steeh, and Bobo 1985; Bobo 2001). Parents education may be symbolic of an ideological environment that facilitates interracial contact, or at the very least does not discourage it. Religious afliation is also predictive of future mating choices (Kennedy 1944; Murstein 1986; Ellman 1987; Kalmijn 1998), although religious intermarriage has increased among cohorts born after World War II (Sherkat 2004). It is also important to consider religious afliation in studies of interracial marriage because some denominations may have specic teachings about interracial relations and because limiting the eld of potential spouses to those within the same religious community often effectively implies racial or ethnic selection. For example, a Jewish person who aims to marry another Jew is almost always narrowing the pool of available mates to those who are non-Hispanic White (NHW). An NHW Catholic individual who intends to marry a Catholic decreases her chances of marrying an African American, although she may increase her chances of marrying a Hispanic. While the number of multiracial congregations is currently growing (Emerson and Chai Kim 2003), our data reect childhoods and marriages from the 1960s through the early 1990s when congregations were largely segregated. The racial/ethnic composition of a household may also serve as an important contextual trait that promotes future interracial marriage. Alba and Golden (1986) suggested that having a multiethnic background encourages intermarriage. They theorized that families with multiple ethnic backgrounds would weaken afnities to any particular ethnicity for future generations. Accordingly, the U.S. population should become increasingly intermixed as more individuals are shifted into the mixed ancestry category with each new generation. In their analyses, this occurred primarily in the European-ancestry population where the salience of ethnic identity is highly variable (Alba 1990; Waters 1990) and therefore may have a dubious impact on marital behavior. Meanwhile, racial identity remains one of the most salient traits for individual identity in the contemporary United States. Nevertheless, it is worth examining whether multiracial ancestry is related to tendencies toward crossing racial lines. Connecting Two Partnership Moments Background characteristics that shape the propensity toward romantic interracial relationships within adolescence may or may not continue to predict interracial marriage later in life. On the one hand, research repeatedly highlights the importance of previous
346
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

arenas for future contact (e.g., Kouri and Lasswell 1993; Emerson et al. 2002; Kouri 2003). However, there may be a signicant disconnection between the initial contexts that produced interracial dating and those that produce a marriage because dating differs from marriage in very important ways. Fujino (1997) suggests that the likelihood of dating interracially is higher than the likelihood of marrying across race. Yancey (2002) further explains that dating demands less commitment and planning, and has less impact on family formation (Schoen and Weinick 1993). Harris and Ono (2000) nd that interracial contact is more common among cohabiters than among spouses, conrming the notion that marriages across race may be less common because marital relationships require more investment. Despite the differences between dating and marriage, those who date interracially may view the pool of available mates in a different way than those whose initial partnerships were within their own race/ethnicity do. The central question is the consistency of the racial composition of partnerships over time. Are current patterns of interracial marriage in part a function of the underlying preferences of the spouses? Interracial interactions such as dating could serve as predictive indicators of or shape future partnership choices, including selection of a husband, but this basic premise has yet to be tested with a nationally representative sample. Intermarriage has so far been examined as a cross-sectional phenomenon where previous experiences are not taken into account. Rather than examining the pattern of traits of pairs of individuals who have already married, we examine the continuity of choices made by a single individual as a function of background characteristics that may inuence propensities toward crossing racial/ethnic boundaries across time. RESEARCH AIMS We begin with the premise that our analyses will echo the ndings from previous research. Observed patterns of interracial marriage should reect differences in relative sizes of racial groups, so smaller racial/ethnic groups should have higher rates of interracial marriage than whites. Building on this premise, we test three hypotheses. Hypothesis 1: Background characteristics of individuals which have been noted to be associated with tolerant attitudes are associated with the likelihood of interracial partnering at multiple points in the life course. First, we explore the inuence of background characteristics on encouraging interracial romantic relationships, specically the rst sexual partner and rst spouse. While rst sex does not always occur within the context of a committed romantic relationship, the importance attributed to this event (Thompson 1995) indicates that the rst partner for voluntary intercourse is a signicant actor in an individuals life history. Also, interracial sex has historically been legally regulated along with interracial marriage, and thus, the two types of partnering have a shared philosophical history in American culture (Berry 1991; Pascoe 1996). Higher levels of parental education, having no religious afliation, being born in the Western United States, and prociency in English should be positively related to choosing an interracial signicant other.
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

347

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

Hypothesis 2: Signicant interracial partnering choices at one point in time are indicators of the likelihood of signicant interracial partnering choices later in life. Next, we explore consistency in racial/ethnic similarity of partner choice across the two relationships. We hypothesize that having had a racially heterogeneous rst partner is associated with an increased likelihood of later interracial marriage. We cannot disentangle whether our indicator is causal or a joint product, but we can show whether or not predictive value exists, thus making a case for further explication in the future. Hypothesis 3: The connection between interracial partnering choices across the life course varies by race/ethnicity. As social distance varies by race, so does the tendency toward interracial interactions. Members of groups who experience the least social distance from other groups should be most likely to select both an interracial partner and a rst husband. Asians and Hispanics are the least residentially segregated as well as the most favored minority neighbors according to recent research on racial attitudes (Emerson, Chai, and Yancey 2001). Hispanics and Native Americans are also more likely to have friends across racial lines (Joyner and Kao 2000), so interracial interaction is more likely a xture of their daily lives than it is for non-Hispanic Blacks (NHBs) or NHWs. The populations who are likely to have more contact across racial groups in both adolescence and adulthood should exhibit this continuity of choices more so than groups for whom interracial interactions are rarer. METHODS Data The data used in these analyses come from the 1995 National Survey of Family Growth (NSFG), a nationally representative sample of 10,847 women ages 1544. Conducted by the National Center for Health Statistics, the survey is a repeated cross-sectional study designed to provide information on the fertility behaviors and reproductive health of American women. The sample is stratied by age, race and ethnicity, marital status, and parity and includes oversamples of racial/ethnic minority groups (National Center for Health Statistics 1997). The survey questionnaire includes detailed information on the rst sexual partner (if voluntary), husbands, and cohabiting partners, but limited detail on other sexual partnerships and no information on dating or romantic relationships. We excluded 1,438 respondents from the sample because of not having had rst intercourse, having had an involuntary rst intercourse experience, having had the rst intercourse experience outside of the United States, or not providing information on the race/ ethnicity of their rst partner. We removed an additional 193 women in racial/ethnic groups that proved too small to provide reliable estimates in the multivariate analyses with interaction terms (non-Hispanic American Indian/Alaskan Native, non-Hispanic Other, Hispanic Asian/Pacic Islander, Hispanic Black, Hispanic Other). Our analyses are thus based on 9,216 women. The 1995 panel, like all previous panels, only includes women, so we are unfortunately unable to include men in our analyses. We would prefer to have data on men to model their choices as well, but the available data sets that include both genders either
348
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

lack a retrospective relationship history (e.g., the U.S. Census), lack information on sex partners (e.g., the NLSY79), or have not yet followed respondents beyond their mid-20s (e.g., the National Longitudinal Study of Adolescent Health). We think that the unique nature of our analyses makes them worthwhile even when restricted to a single sex, and we hope to conduct future analyses with newer data on both genders (e.g., the 2002 NSFG, which became available recently). Measures Demographic and Background Variables Our key demographic indicator is the respondents racial/ethnic categorization. First, we cross-classied the ve-category race variable (Asian/Pacic Islander, Black, White, American Indian/Alaskan Native, Other) with a dichotomous indicator of Hispanic ethnicity. We then selected only those race/ethnicity groups with at least 100 observations, leaving the following categories: Non-Hispanic Asian/Pacic Islander (NHA), NHB, NHW, Hispanic American Indian/Alaskan Native (HAI), and Hispanic White (HW).1 Second, the NSFG allows respondents to report multiple races so we coded a dummy variable, mixed race, as 1 for respondents who chose multiple races and set all others to 0. Respondents who chose more than one race were asked to choose a single race that best described them. For those who refused, we coded the rst race that they had offered or used the interviewers impressions (whether the respondent looked White, Black, or Other). The racial/ethnic categorizations of the respondents rst sex partner and rst husband were coded in the same manner, with the exception of the option for interviewer perception. Other background variables represent childhood inuences: respondents parents education (the highest reported value was used, with the mean substituted for missing values; we included dummy variables in the model to indicate missing values on mothers education or fathers education.), childhood religious afliation, nativity (born in the United States or elsewhere), and limited prociency in English. Foreign born is a dichotomous variable, and we also created a time-varying measure indicating the period when foreign-born respondents still lived in their country of origin and used that to exclude person-years spent outside the American dating and marriage markets. Previous research has avoided including immigrant women because researchers could not differentiate between marriages contracted abroad and here (e.g., Hwang and Saenz 1990; Qian 2000). Limited English prociency is based on whether or not the survey was taken in Spanish (see Lloyd 2001). Possible effects of the wording of the race/ethnicity questions on responses provided by Hispanic respondents have not been analyzed, so whether or not this factor generates differences remains an unknown question (R. Groves, personal communication, May 19, 2005). We did not include measures of the respondents education for two reasons. First, the respondents own education is endogenous to the timing of marriage. Education may delay the decision to marry, and marriage often coincides with the end of schooling. Parents education is a largely unchanging characteristic during the respondents life
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

349

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

course and an element of the context within which early and later partnership choices are made. Second, variation in parental education explains a large portion of the variation in respondents education. Stratication scholars have noted the continuity across generations with respect to education and the attainment of socioeconomic status. Factors Affecting Marriage Timing We also include age and age squared in our models to reect the curvilinear pattern of the likelihood of marriage over the life course. Marriage rates increase sharply across the late teens and 20s and then drop off across the 30s and into older ages. Birth cohort accounts for the increasing delays in marriage timing as well as the increased incidence of interracial marriage over recent decades. Intact during childhood indicates whether the respondents residential family at age 14 included both biological or adoptive parents. Research indicates that girls raised in nonintact households tend to marry earlier (Teachman 2003). Finally, we include the continuous measure age at rst intercourse since the timing of rst intercourse is linked to the timing of marriage. We also include this variable as a control variable in the models for rst sex partner. Geography We include geography as a context for interracial relationships through dummy variables for the respondents region of birth (Northeast, Midwest, South, and West) for those born in the United States. Previous research has noted regional variation in attitudes toward racial diversity, religious diversity, and individualism (Ellison, Burr, and McCall 1997). Region can serve as a proxy for racial ideology, racial diversity, and the inuence of a particular history of interracial relations (Harris and Sim 2002). We focus on region of birth because it is a background characteristic and thus ts theoretically with our other background variables. We are also limited to this measure because the NSFG does not include time-varying measures of location, so we cannot track changes across childhood and adolescence that could be preferable for a model that predicts transition to marriage.2 Childhood regional location is also unfortunately not available for the 1,117 foreignborn respondents (so immigrants in our subsample are coded foreign born, as described above). We do not use the other available variable, current region of residence, because it may reect postmarital migration choices. Interracial couples may move to a region where the ideological climate is friendlier. The four regions vary considerably in their racial composition. For reference, we calculated an entropy index (White 1986) based on the racial composition of each region in 1990 (the NSFG region variable matches the U.S. Census denition of region). For this index, values closer to zero represent less diversity and closer to one represent greater diversity. The Midwest had a value of .30, followed by the Northeast at .42, the South at .53, and the West at .57. Diversity in the South is generated largely by NHBs while the West has a greater presence of Hispanics and NHAs (U.S. Bureau of the Census 1996). Although geographic attributes at the metropolitan or county level have been applied in other research to take into account the available local opportunities for
350
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

meeting potential spouses (e.g., Cready and Saenz 1997; Hwang, Saenz, and Aguirre 1997; Harris and Ono 2000), our approach better suits our needs for examining the role of background characteristics, as opposed to contemporaneous opportunities for contact, in making choices to intermarry. Dependent Variables We have two dependent variables: homogeny between the race/ethnicity of the respondent and her rst sex partner, rst partner different race, and homogeny between the race/ ethnicity of the respondent and her rst husband, rst husband different race. The decision to focus on rst sex partner was driven by its theoretical importance and the data available to us. The former variable is coded 1 if the rst sex partner or husband does not match the respondent on race and Hispanic origin and 0 if the two match on both dimensions. First husband different race has two additional coding possibilities, described in the next section. We do not model for specic race-by-race or ethnicity-by-ethnicity combinations between partners but instead look at interracial interactions globally. Although more detailed analyses would add an important dimension to a larger conceptualization of race- and ethnic-specic social distances, our sample size does not permit us to estimate such multivariate models and obtain robust results. We discuss the implications of this choice in our conclusion. Statistical Methods We rst generate descriptive statistics for Tables 1 and 2; we use the survey commands in Stata (svyprop and svymean; Statacorp 2001) to account for the complex sampling design of the NSFG. Our subsequent multivariate analysis has two steps: assessing a womans choice of rst partner and assessing her choice of rst husband. First, we model whether the respondents rst sex partner was of the same or different race/ethnicity as the respondent. Since no respondents in our analytic sample are censored, these models use logistic regression (svylogit generates correct standard errors accounting for stratication and clustering). Some respondents in the sample married their rst sex partners. Obviously, these women create difculties when examining consistency over time in partnership choices because the same man is chosen both times. For the models that predict rst partnership choices, we created an independent variable, married rst sex partner, to differentiate these women. Whether or not these women entered a sexual relationship with their rst partner with the understanding that he would become her husband can only be assumed for women who were engaged or were married at the time of their rst sex. An earlier series of models that included this detailed information revealed no signicant effects, so we retained only the single indicator. For the models of rst marriage, we differentiated these women through the dependent variable, as described below. The next step in our analysis was to model the respondents entry into her rst marriage. Because we retained unmarried women in the analytic sample, there are four possible values for the dependent variable: (1) married partner of the same race/ethnicity
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

351

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

TABLE 1. Sample Characteristics by Type of First Marriage Same-race married N All women First sex partner different race Race/ethnicity Non-Hispanic White Non-Hispanic Asian Non-Hispanic Black Hispanic American Indian Hispanic White Mixed race Parents education Less than high school High school diploma Some college College degree Dads education missing Moms education missing Childhood religious afliation Protestant No afliation Catholic Jewish Other Family structure Intact during childhood Birth cohort 1950 to 1954 1955 to 1959 1960 to 1964 1965 to 1969 1970 to 1974 1975 to 1979 Region of birth West Northeast Midwest South Foreign born Limited English prociency N Age at marriage/censoring Age at rst intercourse 2,987 2,987 Interracially married Never married Married rst partner Percent

Percent N

Percent N

Percent N

2,987 100.0 178 5.8 2,201 12 643 14 117 24 578 1,291 443 675 160 32 1,822 182 906 43 34 1,788 561 827 798 522 248 31 481 630 818 957 101 17 84.6 0.6 11.7 0.3 2.8 0.7 16.5 44.4 15.0 24.2 4.8 1.1 58.2 6.9 32.0 1.8 4.2 61.6 17.8 26.4 25.2 18.9 10.2 1.4 16.5 22.0 28.7 29.6 3.2 0.5 Mean 23.0 16.9

337 100.0 155 43.7 160 16 32 17 112 10 56 126 63 92 24 3 155 24 152 3 3 176 51 79 109 66 29 3 117 57 65 77 21 0 N 337 337 57.5 5.9 6.3 3.3 27.0 3.1 13.1 39.1 18.3 29.6 5.2 0.8 45.5 9.2 42.2 1.2 1.8 53.5 16.3 21.7 30.5 20.5 9.5 1.5 37.0 16.4 19.7 20.9 6.0 0.0 Mean 23.6 16.8

2,898 100.0 408 15.3 1,420 45 1,127 46 263 38 521 1,103 490 784 265 22 1,717 250 846 33 52 1,388 150 262 348 563 847 728 426 594 795 930 153 23 N 2,898 2,898 63.5 2.0 25.1 1.5 7.9 1.3 14.4 36.8 17.7 31.1 7.5 0.8 54.8 9.4 32.2 1.5 2.2 51.7 4.5 8.3 10.8 18.5 29.5 28.3 15.9 21.9 28.2 28.5 5.4 0.8 Mean 25.0 17.0

2,994 100.0 265 8.3 2,027 69 429 49 420 21 841 1,247 403 503 118 9 1,688 130 1,092 28 56 2,023 843 796 695 431 200 29 435 489 766 987 317 86 N 2,994 2,994 77.0 3.2 8.2 1.1 10.5 0.9 25.4 43.2 13.9 17.5 3.5 0.2 56.6 4.6 35.3 1.1 2.3 68.6 28.5 26.0 21.4 15.0 7.7 1.3 14.4 17.5 26.4 32.4 9.4 2.3 Mean 20.3 18.8

Note: Ns are unweighted; percentages and means are weighted.


352
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

TABLE 2. Race of First Partner and First Husband by Race of Respondent Non-Hispanic Asian/Pacic Islander A. First partner N Same race and ethnicity Same race Same Hispanic ethnicity B. First husband N Same race and ethnicity Same race Same Hispanic ethnicity 139 56.3 57.1 92.8 103 69.4 69.7 98.5 Black 2,231 93.7 94.7 97.8 1,104 95.0 95.5 98.8 White 5,808 92.0 95.4 95.8 4,397 94.0 96.8 96.8 Hispanic Alaskan Native/ American Indian 126 54.7 56.5 69.4 84 59.6 60.8 73.6 White 912 67.1 93.7 70.2 678 68.3 94.4 70.9

Note: Ns are unweighted; percentages are weighted. Only respondents who did not marry their rst sex partner are included.

who was not the rst sexual partner (coded 0); (2) married partner of a different race/ ethnicity who was not the rst sexual partner (coded 1); (3) never married (coded 2); and (4) married rst sexual partner (coded 3). We retain the never married population in our analytic sample because of racial and ethnic differentials in the likelihood of staying in the never married state. Removing the never married population may lead us to misestimate the likelihood of intermarrying over time if we ignore the possibility that some who are not intermarrying are also not marrying anyone. For example, the low likelihood of interracial marriage among African-American women is partially because of the relatively small percentage of married African-American women who have husbands of different races (Qian 1997) and partially because of African-American womens relatively lower likelihood of marrying at all (Lichter et al. 1992; Goldstein and Kenney 2001). Because these data are censored since the respondents may one day marry, we use event history methods. We generate a person-year le from the respondent-based data and analyze it using multinomial logistic regression (svymlogit). Multinomial logit models for event history analysis allow the analyst to designate one outcome as the base. Once that base is chosen, the coefcients and signicance tests produced refer to the contrast between each of the other given outcomes and the base outcome. Our primary concern in this study is differences between same-race and different-race marriage. Therefore, we designated same-race marriage as the base category, and we present only the contrast between different-race marriage and same-race marriage. The outcomes of never marrying and marrying the rst sex partner are simultaneously estimated; these contrasts are available upon request from the authors. Simultaneous estimation of all outcomes in the multinomial logit framework generates parameter estimates that are more efcient than those generated by a series of binomial logits (Long 1997).
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

353

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

RESULTS Sample Characteristics Table 1 shows the characteristics of the respondents by marital outcome: (1) intraracially married, (2) interracially married, (3) never married, (4) married rst sex partner.3 The percentage of women whose rst partner was of a different race clearly varies by marital status, with the largest percentage among those who marry interracially (44 percent) and the smallest among women who either married within their race (6 percent) or married their rst partner (8 percent). As hypothesized, women who eventually cross racial lines in marriage appear more likely to have done so in previous relationships. The distribution by race/ethnicity also varies across these outcomes. NHW women are overrepresented among women who married within their race or married their rst partner, while they represented a relatively smaller share of interracially married women and never married women. The smaller minority groups of NHA, HAI, and HW women are overrepresented among the interracially married. In contrast, NHB women represent relatively smaller shares of all three marital categories and are overrepresented among the never married. These distortions reect differing marriage propensities among racial and ethnic groups in the United States. The married proportion of the adult population is highest among NHWs, followed by Asians and then Hispanics, and lowest among NHBs (Current Population Survey Reports 1998; Cohen 1999). Only 1 percent of the sample identied themselves as mixed race. These women were overrepresented among the interracially married and were underrepresented in all other categories. The subsequent sections of Table 1 show the distribution of background characteristics for women within marital outcome groups. Women with college-educated parents are disproportionately prevalent among the interracially married (30 percent) and never married (31 percent), while women with parents who earned less than a high school education are most prevalent among those who married their rst partner (25 percent). Interracially married women were less likely to have identied as Protestant in their childhood and were somewhat more likely to have been raised Catholic or unafliated with any religion. Tabulations (not shown) reveal that the driving factor behind this religious difference is the greater likelihood of Catholic NHA and NHB to marry interracially. This difference may reect greater contact with Whites and Hispanics through the Church during childhood since Asians and Blacks in the United States are predominantly Protestant. Interracially married and never married respondents were less likely to have lived with both parents at age 14. Regionally, respondents born in the Western United States represent 37 percent of the interracially married, compared to less than 17 percent among all other groups. Interestingly, foreign-born women are most prevalent among women who marry their rst partner (9 percent) compared to any other outcome, particularly the same-race married, only 3 percent of whom are foreign born. Meanwhile, limited English prociency appears to form a barrier to intermarriage for women as Lloyd (2001) observed for Latino men. The ndings from this table suggest that those who marry their rst partner are distinctive from the rest of the sample. The mean age at marriage for this group is nearly
354

The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

two years younger than those who marry within race. One-quarter of their parents earned less than a high school education. They are the most likely to have had some religious afliation during childhood (95 percent) and have lived in an intact family (69 percent). These traits may indicate a distinctive ideological contextwe speculate in a more conservative direction than averagethat enhances the propensity to marry within their own racial group as well as avoid searching beyond the rst partner for a husband. This table also shows the number of observations underlying each category. Sample limitations reduced the numbers in some race/ethnicity categories that had enough observations to pass our initial cut, and the reader should keep this in mind when interpreting the multivariate results. For example, the effects of mixed race status are based on fewer than 100 women. Another area of potential concern is that only 126 respondents are Hispanic with limited English prociency, and none of those women were interracially married. Mate Selection by Race and Ethnicity We assert that it is important to consider race for respondents of Hispanic ethnicity and whether or not respondents are Hispanic within each racial group. Race and Hispanic ethnicity have interactive effects on partnering choices that are only revealed when the two factors are brought together in analyses. Table 2 shows the prevalence of homogeny with rst partners and rst husbands by racial/ethnic categorization both separately and simultaneously. Overall, these proportions show high levels of partnering within race and Hispanic ethnicity. (For non-Hispanic respondents, same Hispanic ethnicity indicates that both members of the couple were non-Hispanic.) Racial and ethnic homogeny with rst sexual partners was greatest for NHB and NHW women. Levels of all types of homogeny for these two groups were always over 90 percent. The majority of NHA and HAI women partnered homogeneously by all three denitions, but sharing their partners Hispanic ethnicity was more common than was sharing their race, particularly for NHA women. However, race rather than Hispanic ethnicity appeared more relevant for HW respondents. Over two-thirds of HW women had rst partners of the same race and ethnicity as themselves and over 90 percent shared their partners race (i.e., both were White). The bottom panel of the table shows that these trends in the selection of rst sexual partners are reected in the selection of rst husbands, with tendencies toward homogeny either remaining the same or increasing for marriage. (Levels of homogeny below 90 percent generally increase by 510 percent when respondents who married their rst partners are included in the table; levels already above 90 percent remain about the same. A revised table with these results is available from the authors.) Based on these results, we conduct subsequent analyses with a denition of same-race marriage that requires similarity on both racial and ethnic dimensions. We would only have enough observations to model specic differences across race within Hispanic ethnicity for Whites and American Indians and across Hispanic ethnicity for Whites. Future research should investigate such topics as whether the ethnic differences observed for
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

355

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

Whites also hold for Asians and whether Hispanic groups of all races exhibit the same racialized patterns as those shown by non-Hispanic groups. Similarities in Choices of First Sex Partner and First Husband Multinomial logit analyses of marriage type revealed a statistically and substantively signicant positive association between having a rst sex partner of a different race and later interracial marriage (odds ratio [OR] = 4.4, p < 0.000, only this variable and age effects included). The apparent predictive effect of this indicator variable could represent endogeneity in the processes of mate selection. A womans view of her potential partnership choices could have been shaped prior to her sexual initiation by her own race/ethnicity, her parents education, her religious background, or her nativity characteristics. To see the interconnections between demographic and family background characteristics, rst partner choice, and marriage type, we now turn to results from regression models in Tables 3 and 4. Table 3 shows the effects of race/ethnicity and background characteristics on the likelihood of having a rst sex partner of a different race. We present the results in the form of odds ratios, where effects above 1.00 indicate an increase in the likelihood of having a partner of a different race and effects below 1.00 represent a decrease in that likelihood. Model 1 is our base model, adjusting only for race/ethnicity, whether or not the respondent married her rst sex partner, and age. Model 2 introduces controls for parental education, childhood religious afliation, childhood family structure, birth cohort, region of birth, and English prociency. Across both models, marrying the rst sex partner is negatively associated with an interracial relationship. Marrying ones rst sex partner may be viewed as a conservative choice, which could go along with a broader tendency toward conservatism such as limiting potential intimate partners to members of ones own race/ethnicity. The relative propensity toward interracial rst partnerships among different race/ethnicity groups appears stable across models. With and without demographic and family background controls, NHA, HAI, and HW respondents were signicantly more likely to have a heterogeneous rst partner than were NHW respondents. Respondents with mixed racial backgrounds were also more likely to have an interracial rst partner relative to those who reported a single race. However, NHB respondents were not signicantly more or less likely than their white counterparts to have an interracial rst sex partner once background characteristics were taken into account. The odds ratios for race/ethnicity in these analyses are notably large. Their magnitude reects the relative rarity of the events of interracial partnership and marriage for the NHW majority in the sample. For example, the apparently huge odds ratio on NHA occurs because this group is much smaller and the event is more frequent among them. We have condence in our result because, as shown in Table 2, nearly half of the NHA sample (44 percent) has a rst sex partner of a different race and ethnicity compared to less than 10 percent of the NHW sample. Results in the second model showed that parents education, family structure, and region of birth are also associated with having had an interracial rst partner. The inclusion of parents education results in a reduction of the effects of race/ethnicity; for
356
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

TABLE 3. Odds Ratios Predicting Interracial First Partner versus Same-Race First Partner Model 1 OR Constant 0.20*** Age at rst intercourse 0.96* Married rst partner 0.55*** Race/ethnicity (reference = Non-Hispanic White) Non-Hispanic Asian 11.09*** Non-Hispanic Black 0.68** Hispanic American Indian 9.58*** Hispanic White 6.30*** Mixed race 3.06*** Highest parents education (reference = less than high school) High school Some college College Childhood religious afliation (reference = Protestant) No religion Catholic Jewish Other religion Family background Intact during childhood Birth cohort (reference = 1950 to 1954) 1955 to 1959 1960 to 1964 1965 to 1969 1970 to 1974 1975 to 1979 Region of birth (reference = West) Northeast Midwest South Born outside of the United States Limited English Prociency F-test for added variables 84.7*** Model 2 OR 0.12*** 0.99 0.68*** 10.21*** 0.82 11.71*** 7.28*** 2.59** 1.43* 1.60** 1.79*** 1.28 0.96 1.32 0.89 0.83* 1.05 1.09 1.09 1.40* 1.46* 0.51*** 0.47*** 0.43*** 0.43*** 0.13*** 5.7***

Note: N = 9,216, p < .10, *p < .05, **p < .01, ***p < .001. Analyses also control for missing values on parental education. These dummy variables are not statistically signicant.

example, the baseline odds of an interracial partnership for NHA respondents are now only slightly greater than random (0.12 * 10.21 = 1.23). Despite the limitations on our contextual information, indicators of geography and acculturation also showed effects consistent with our hypotheses. Respondents born in any region other than the Western
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

357

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

TABLE 4. Odds Ratios Predicting Interracial First Husband versus Same-Race First Husband (With Never Married and Married First Sex Partner as Competing Risks, Not Shown) Model I OR Sig. Model II OR Sig. Model III OR Sig. Model IV OR Sig.

Constant 0.06*** 0.05*** Age 0.99 0.99 1.00 1.00 Age2 Race/ethnicity Non-Hispanic Asian 19.94*** 23.77*** Non-Hispanic Black 0.76 0.95 Hispanic American Indian 15.99*** 19.64*** Hispanic White 15.32*** 19.51*** Mixed race 3.12* 1.95 Highest parents education High school 1.93** Some college 2.80*** College or above 3.02*** Childhood religious afliation No religion 1.01 Catholic 0.89 Jewish 0.80 Other religion 0.59 Family background Intact during childhood 0.72* Birth cohort 1955 to 1959 0.88 1960 to 1964 1.12 1965 to 1969 1.12 1970 to 1974 0.85 1975 to 1979 0.88 Region of birth Northeast 0.63* Midwest 0.64* South 0.58* Born outside of the United States 0.47* First Partner Different Race Age at First Intercourse First Partner DR Interactions with Race of Respondent First Partner DR*NHA First Partner DR*NHB First Partner DR*HW First Partner DR*HAI F-test for added variables 42.0*** 4.3***

0.03*** 1.01 1.00 11.76*** 1.05 15.11*** 15.35*** 1.89 1.94** 2.80*** 3.19*** 1.05 0.91 0.82 0.57 0.69* 0.93 1.21 1.26 0.99 1.14 0.65 0.67* 0.60* 0.48* 3.21***

0.14*** 1.04 1.00 13.55*** 0.66 16.84*** 12.54*** 1.54 1.86* 2.75*** 3.18*** 1.05 0.90 0.78 0.48 0.77 0.88 1.19 1.21 0.87 1.00 0.65 0.68* 0.63* 0.56 3.39*** 0.92*** 1.45 2.49*** 1.49 0.57 5.6***

133.3***

Note: N = 214,255, p < .10, *p < .05, **p < .01, ***p < .001. Analyses also control for missing values on parental education. These dummy variables are not statistically signicant.
358

The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

United States were less likely to have an interracial rst partner than those born in that area. F-tests revealed no signicant differences between regional coefcients (analyses not shown), indicating that the regional variation in propensity toward interracial rst partners occurs mainly between Western and non-Western residents. Limited English prociency among Hispanic respondents was also strongly associated with the likelihood of staying within race when choosing a rst partner. Relative to respondents who completed the survey in English, those who took it in Spanish were .13 times as likely to have an interracial rst sex partner. Nevertheless, the odds of an interracial rst partnership remain substantially higher for Asian and Hispanic women than for NHW women, even when these background factors are included. The question of whether the same background factors that predict the selection of an interracial rst partner also predict the selection of an interracial rst husband remains. In Table 4, we present odds ratios for having an interracial rst marriage versus same-race rst marriage. Again, we begin with a baseline model, showing the initial differences in the likelihood of interracial marriage across racial/ethnic groups. We then employ essentially the same set of variables from the preceding analysis in model II (with the exception of married rst partner, which is now accounted for in the construction of the dependent variable; also, we do not include English prociency in these models because no respondents with limited English prociency married interracially). Models III and IV introduce information on the rst partner. The results show that many of the characteristics linked to an interracial rst sexual relationship also have signicant associations with interracial rst marriage. The relative magnitude of the effects of racial/ethnic identication on crossing racial lines in marriage were similar to those found for rst sex partnerships and remained steady across models. The odds ratio for NHB respondents relative to NHW respondents was still statistically insignicant. Interestingly, the effect of being mixed race, while strongly predictive of having an interracial rst partner, was not signicant for predicting an interracial rst marriage once background characteristics were controlled. Potentially, a mixed race background leads to a greater heterogeneity of relationships prior to marriage, but mixed race individuals may display marital selection tendencies similar to the group that best describes them in adulthood because of a perceived greater seriousness or permanence of marriage. Parental education continued to yield strong positive associations with interracial marriage. Region of birth, still the region of residence for the majority of respondents, also retained an association with interracial partnering. Continuity across Time 1 and Time 2 Outcomes Finally, we added information about the rst sex partner to models III and IV to observe whether having an earlier interracial relationship mediates and interacts with factors related to the selection of an interracial rst spouse. Model III includes an indicator for having had a rst partner of a different race (F-test = 133.3, p < 0.001). This variable partially mediates the variation in interracial marriage across race/ethnicity. While NHA, HAI, and HW women are still more likely than NHW women to have interracial spouses, the magnitudes of the odds ratios are sharply reduced. Parental education effects remain
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

359

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

positive and strong. Controlling for the race of the rst partner appears to attenuate mildly the effect of region of birth. Model IV shows how the continuity between the two outcomes varies by race/ ethnicity by employing interactions between rst partner and race/ethnicity. The results show that having a rst partner of a different race is especially predictive of interracial marriage for NHB women. This is surprising given that the prior analyses on NHB womens tendencies toward interracial partnering and marriage revealed behavior essentially similar to NHW respondents. This interaction effect suggests that the racial nature of early intimate relationships among black women is more indicative of later mate selection than it is among other women. DISCUSSION Interracial marriage remains an important barometer of both interracial relations and the relative importance of ascribed versus achieved characteristics in mate selection. However, few studies of interracial marriage explore inuences that precede entry into the marriage market. Our initial goal was to explore a womans background as either encouraging or discouraging of intimate interracial interactions, specically rst sexual partnership and rst marriage. We then explored whether or not selecting a rst partner across racial lines was associated with the later choice of rst husband. Finally, we identied how this link may vary across racial groups. These results conrm a signicant variation across race and ethnicity in the likelihood of interracial romantic relationships. For both outcomes, NHA, HW, and HAI women have higher odds of intimate interracial interaction relative to NHW respondents. Meanwhile, overall tendencies among NHB women toward having interracial rst partners or rst husbands do not differ signicantly from those of NHW women, despite the much smaller representation of NHB women in the overall population. We further nd that multiple factors in a womans backgroundrace/ethnicity, parents education, region of birth, and prociency in Englishimpact the likelihood that she will engage in an intimate interracial relationship. Regarding parental education, we tested a model in which we added respondents own time-varying education and found that the effects of parental education remained. (Own education itself had marginal to no effect on the contrast between same-race and interracial marriage.) In addition, we nd that women whose rst partners are of a different race are more likely to have rst husbands of a different race as well. Surprisingly, this link is particularly strong for a group that is highly unlikely to intermarry: NHB women. Below, we discuss the implications of this nding for placing interracial marriage within a broader framework of relationships and how this nding ts into the literature on the racial composition of perceived marriage markets. As we noted, the primary limitations of our analyses are that the data do not include time-varying information on the demographic characteristics of the local area and that they are limited to women. Our results would be strengthened if we had measures of who was available for partnering when respondents became involved with their rst sex
360
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

partner and rst husband. The results would also be strengthened if we were able to replicate them for men, which we hope to do in the future with the release of the 2002 NSFG. Nevertheless, our ndings present a piece of the puzzle largely unexamined thus far, highlighting the romantic life course. On a Path toward Interracial Marriage We began with the premise that rst sexual partnerships and rst marriages across the color line are the result of preferences formed prior to entering either of those relationships and that evidence of these preferences or tendencies may be found in an individuals background characteristics. In this line of thought, these factors can be considered indicators of endogenous underlying propensities toward contact across racial lines. Our analyses show that women who were born in the Western United States and whose parents were educated beyond high school demonstrate a heightened propensity toward having intimate interracial relationships. As shown by Lloyd (2001), a lack of prociency in English forms a barrier to intimate relationship contact between Hispanics and nonHispanics. The impact of mixed racial background is highly relevant for the selection of an interracial rst sex partner, although it appears to fade later in the life course. These associations conrm our initial expectations that background environments that tend toward liberal racial attitudes or demonstrate a greater degree of racial diversity encourage later engagement in interracial interactions. We also nd that women who eventually interracially marry have distinctive partnering experiences, particularly from those who marry their rst partner, because the path toward interracial marriage is more likely to include multiple intimate partners. In Table 3, women who did not marry their rst partner were more likely to be involved across the color line than those who did. The results of Table 4 show that having a different-race rst partner is highly predictive of an interracial marriage (estimated simultaneously with the possibility of having married the rst partnera competing risk, not shown). Thus, individuals who eventually intermarry are likely to have been exposed to a variety of relationships beyond their initial intimate encounter. If that initial intimate encounter was interracial, it is strongly associated with a hypothetical turn in their trajectory from the default direction of a same-race marital partner toward the more raried outcome of an interracial husband. This result also indicates that the population that is likely to enter an interracial marriage is largely a subset of the population that has premarital sexual experience. This explains the family structure effect that women who lived within an intact family are less likely to have partners of a different race since living in a single parent home is positively related to having had a premarital sexual experience (Albrecht and Teachman 2003). These analyses also reveal selected continuities between the relationship events of rst sexual partner and rst marriage. For nonblack women, having had an interracial rst sex partner increases their odds of entering an interracial marriage rather than a same race marriage by a multiplier of 3.4 compared to other women of their race/ethnicity (considering only the main effect of race of rst partner since the interaction terms between race of rst partner and race/ethnicity are not signicant for these groups). For black women,
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

361

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

having had an interracial rst sex partner increases their odds of entering an interracial marriage rather than a same race marriage by a multiplier of 8.5 (3.4 * 2.5) compared to other black women. This strong likelihood of continuity in the choices made by black women conrms ndings from the ethnographic literature that the interracially married tend to have had previous interracial interactions independent of those with their spouse (Lewis et al. 1997; Root 2001; Kouri 2003). This literature emphasizes the need for placing the event of an interracial marriage within the context of a larger story about how individuals overcame preexisting racial stereotypes. The implication of the combined results from the ethnographic work and our study is the need to include more information on the types of premarital interracial contact that individuals experience. These data cannot address why the continuity of partner choice appears particularly important for NHB women. Potentially, the choice to date interracially for NHB adolescents proxies for other tendencies that not only separate them from the larger black majority but also lead these women into marriage markets where black men are less frequently present. Overall, the picture painted by these results is that NHA and HW women are able to cross the lines of race and ethnicity in partner and spousal choice relatively freely, but NHB women are either unwilling or unable to do so (Rosenfeld 2002 draws similar conclusions). If the situation is the latter, this nding suggests an extreme racial caste disadvantage beyond being nonwhite that extends specically to being black. This suggestion calls to mind Daviss (1941) claim that racial caste disadvantage was the only type of caste difference that could never be overcome. However, what looks like racial caste disadvantage is more likely a combination of opposition toward interracial relating and structural constraints that inuences the degree of racial and social distance across the life course. As blacks experience more social distance from whites on an array of dimensions (Massey and Denton 1989; Clark 1992; Farley and Frey 1994; Bobo and Zubrinsky 1996; Zubrinsky and Bobo 1996), the amount of meaningful social interaction that could potentially result in interracial friendship and dating declines over the life course. For black women, early partnering experiences are highly segregated. As well, many black women voice substantial opposition to interracial romance, particularly since the black power movement of the late 1960s and early 1970s (e.g., Kennedy 2003; Romano 2003:21647). Chito Childs (2005) links this phenomenon to white racism, the need for black communal solidarity, and cultural standards of beauty that devalue black women. Thus, these low propensities toward interracial interaction likely reect a combination of cultural, social, structural, and historical factors. Meanwhile, Asian and Hispanic women, who are comparably less segregated and come from groups with their own unique histories of racial and ethnic conict in the United States, are more likely to have interacted with whites over the life course and thus have greater opportunities for interracial relationships. Race and ethnicity do inuence partnering between other nonblack groups. The results of Table 2 suggest that NHA and HW women negotiate the lines of race and ethnicity selectively. For example, NHA women more commonly cross the lines of race (marrying other non-Asians) than ethnicity (partnering with Hispanics). This supports prior work noting that Asian women who intermarry more often marry NHW than other
362
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

groups (Hwang, Saenz, and Aguirre, 1995; Qian 1997). One explanation is structural in nature. Places where both Asians and Hispanics reside in large numbers are also large metropolitan areas where both groups have many potential partners of their own race/ ethnicity. Areas where Asians intermarry may necessarily be areas where not only few Asians live, but also few Hispanics as well (Rosenfeld 2001). Prejudice may also play a role. Like Blacks, Hispanics are also subject to negative stereotypes that could impact the impressions that other groups have of them and ultimately affect the likelihood of contact (Mindiola, Flores Niemann, and Rodriguez 2003). Meanwhile, a greater percentage of HW women date and marry within racial lines (i.e., other whites) than within ethnic ones (i.e., other Hispanics), according to Table 2. Previous research has shown that race, as well as ethnicity, inuences the intermarriage of Hispanics (Qian and Cobas 2004). Together, these two patterns support the notion that Hispanics and Asians are not simply more prone to intermarriage but are also more likely to engage in certain types of intermarriage. Future Implications One clear implication of these ndings is the need to link histories of interracial contact during adolescence to patterns of contact in adulthood. Blau, Blum, and Schwartz (1982) noted, To be sure, most casual encounters do not lead to marriage. Yet such supercial contacts are a necessary, though not a sufcient, condition for more intimate relations to develop (p. 47). A growing ethnographic literature on interracial marriage positioned within the narrative of life story emphasizes the need to view interracial contact as a process rather than a isolated event (Root 2001; Kouri 2003). Further research also needs to delve deeper into the connection between specic racial and ethnic combinations in interracial partnerships. For example, the category of Asian should be unpacked into country-of-origin groups to reect the cultural differences between them. Different communities have varying expectations of their members; for example, the range of acceptable marriage partners likely differs sharply for a third-generation Chinese woman than a rst-generation woman from Japan. Given that the largest number of interracial and interethnic partnerships in the general population involve one NHW person, our analyses are most meaningful to comment on the social distance between the specic combinations of non-Hispanic non-White women and NHW men, HW women and NHW men, and NHW women with Hispanic or Asian partners. The presence of past interracial rst sex partners provides insight into the likely selection of future partnerswhat about neighbors, friends, and casual dating? As Brown, Feiring, and Furman (1999) state, the study of adolescent romance is a recent development, and the study of interracial romance is particularly nascent (e.g., Joyner, Wang, and Kao 2003; Joyner and Kao 2005). Racially segregated environments are undoubtedly less encouraging of interracial adolescent relationships. Although this study does not examine any of the standard indicators of intergroup opportunities for interaction (see Blau and Schwartz 1984; Blau 1994), our ndings show some convergence of early opportunity and preference conditioning future behavior in the larger marriage market. To this end,
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

363

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

social histories of interaction patterns of black females in particular may reveal further evidence of life course continuity in interracial social contact. ACKNOWLEDGMENTS Previous versions of this article were presented at the 2001 annual meeting of the Population Association of America and the 2001 annual meeting of the American Sociological Association. We would like to thank Kathleen Mullan Harris, J. Richard Udry, R. Kelly Raley, Tracey X. Karner, and anonymous reviewers at TSQ for their helpful comments and suggestions. Work by the rst author was supported by a National Research Service Award (T32 HD07168). The views expressed in the article do not represent the views of the NICHD, the National Institutes of Health, the U.S. Department of Health and Human Services, or the United States. NOTES
1

Contrary to our expectations, most Hispanic respondents chose a race. Twenty-nine respondents refused any racial self-identication; 23 of these women self-identied as Hispanic. Few Hispanic respondents selected races of African American or Asian. African descent persons of Hispanic origin from Puerto Rico, the Caribbean, and South and Central America make up more than 12 percent of Hispanic population and almost 2 percent of the U.S. population as of 2000 (Guzman 2001). However, less than 1 percent of the U.S. population listed both Black race and Hispanic ethnic origin in the U.S. Census, which makes this population difcult to capture in a nationally representative sample. 2 On the positive side, 5,984 respondents have always lived in the same state. On the negative side, 74 percent of respondents who report moving from their state of birth did so at least ve years before they were interviewed. 3 As noted earlier, these categories are mutually exclusivewomen in categories (1) and (2) have had at least one sexual partner prior to their husbands.

REFERENCES
Alba, Richard. 1990. Ethnic Identity: Transformation of White America. New Haven, CT: Yale University Press. Alba, Richard D. and Reid M. Golden. 1986. Patterns of Ethnic Marriage in the United States. Social Forces 65:20223. Albrecht, Chris and Jay Teachman. 2003. Childhood Living Arrangements and the Risk of Premarital Intercourse. Journal of Family Issues 24(7):86794. Allport, Gordon. 1954. The Nature of Prejudice. Cambridge, MA: Perseus Books. Berry, Mary Frances. 1991. Judging Morality: Sexual Behavior and Legal Consequences in the Late Nineteenth-Century South. The Journal of American History 78 (December):83556. Blau, Peter M. 1994. Structural Contexts of Opportunity. Chicago, IL: University of Chicago Press. Blau, Peter M., Terry C. Blum, and Joseph E. Schwartz. 1982. Heterogeneity and Intermarriage. American Sociological Review 47:4562.
364
The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

Blau, Peter M. and Joseph E. Schwartz. 1984. Crosscutting Social Circles: Testing a Macro-Structural Theory of Intergroup Relations. New York: Academic Press. Bobo, Lawrence. 2001. Racial Attitudes and Relations at the Close of the Twentieth Century. Pp. 264301 in America Becoming: Racial Trends and Their Consequences, vol. 1, edited by Neil J. Smelser, William Julius Wilson, and Faith Mitchell. Washington, DC: National Academy Press. Bobo, Lawrence and Camille L. Zubrinsky. 1996. Attitudes on Residential Integration: Perceived Status Differences, Mere In-Group Preference, or Racial Prejudice? Social Forces 74:883909. Brown, B. Bradford, Candice Feiring, and Wyndol Furman. 1999. Missing the Love Boat: Why Researchers Have Shied Away from Adolescent Romance. Pp. 33063 in The Development of Romantic Relationships in Adolescence, edited by Wyndol Furman, B. Bradford Brown, and Candice Feiring. Cambridge, England: Cambridge University Press. Chito Childs, Erica. 2005. Looking behind the Stereotypes of the Angry Black Woman: An Exploration of Black Womens Responses to Interracial Relationships. Gender and Society 19:54461. Clark, William A. V. 1992. Residential Preferences and Residential Choices in Multiethnic Context. Demography 29:45166. Cohen, Phillip. 1999. Racial-Ethnic and Gender Differences in Returns to Cohabitation and Marriage: Evidence from the Current Population Survey. Presented at the annual meeting of the Population Association of America, March 2527, New York, NY. Retrieved February 18, 2002 (http://www.census. gov/population/www/documentation/twps0035/twps0035.html). Cready, Cynthia and Rogelio Saenz. 1997. The Nonmetro/Metro Context of Racial/Ethnic Outmarriage: Some Differences between African Americans and Mexican Americans. Rural Sociology 62:33562. Current Population Survey Reports. 1998. Marital Status and Living Arrangements: March 1998 (Update). Retrieved February 18, 2002 (http://www.census.gov/prod/99pubs/p20-514u.pdf). Davis, Kingsley. 1941. Intermarriage in Caste Societies. American Anthropologist 43:37695. Ellison, Christopher, Jeffrey A. Burr, and Patricia McCall. 1997. Religious Homogeneity and Metropolitan Area Suicide Rates. Social Forces 76:27399. Ellman, Yisrael. 1987. Intermarriage in the United States: A Comparative Study of Jews and Other Ethnic and Religious Groups. Jewish Social Studies 49:126. Emerson, Michael O., Karen J. Chai, and George Yancey. 2001. Does Race Matter in Residential Segregation? Exploring the Preferences of White Americans. American Sociological Review 66(6):92235. Emerson, Michael O. and Karen Chai Kim. 2003. Multiracial Congregations: A Typology and Analysis of Their Development. Journal for the Scientic Study of Religion 42:21727. Emerson, Michael O., Rachel T. Kimbro, and George Yancey. 2002. Contact Theory Extended: The Effects of Prior Racial Contact on Current Social Ties. Social Science Quarterly 83:74561. Emerson, Michael O. and Christian Smith. 2000. Divided by Faith: Evangelical Religion and the Problem of Race in America. New York: Oxford University Press. Farley, Reynolds. 2002. Racial Identities in 2000: Response to the Multiple Race Response Option. Pp. 3961 in The New Race Question: How the Census Counts Multiracial Individuals, edited by Joel Perlmann and Mary C. Waters. New York: Russell Sage. Farley, Reynolds and William Frey. 1994. Changes in the Segregation of Whites from Blacks During the 1980s: Small Steps toward a More Integrated Society. American Sociological Review 59:2345.

The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

365

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

Ford, Kathleen, Woosung Sohn, and James Lepkowski. 2001. Characteristics of Adolescents Sexual Partners and Their Association with Use of Condoms and Other Contraceptive Methods. Family Planning Perspectives 33(3):100105, 132. . 2003. Ethnicity or Race, Area Characteristics, and Sexual Partner Choice among American Adolescents. The Journal of Sex Research 40(2):21118. Fu, Vincent Kang. 2001. Racial Intermarriage Pairings. Demography 38:14759. Fujino, Diane C. 1997. The Rates, Patterns, and Reasons for Forming Interracial Dating Relationships among Asian Americans. Journal of Social and Personal Relationships 14(6):80928. Goldstein, Joshua R. and Catherine Kenney. 2001. Marriage Delayed or Marriage Forgone? New Cohort Forecasts of First Marriage for U.S. Women. American Sociological Review 66:50619. Gordon, Milton T. 1964. Assimilation in American Life: The Role of Race, Religion, and National Origin. New York: Oxford University Press. Gray, Marjory R. and Laurence Steinburg. 1999. Adolescent Romance and the Parent Child Relationship: Bi-Directional Links across Settings. Pp. 23565 in The Development of Romantic Relationships in Adolescence, edited by Wyndol Furman, B. Bradford Brown, and Candice Feiring. Cambridge, England: Cambridge University Press. Guzman, Betsy. 2001. The Hispanic Population: Census 2000 Brief. U.S. Bureau of the Census. Issued May 2001. Retrieved April 7, 2006 (http://www.census.gov/prod/2001pubs/ c2kbr01-3.pdf). Harris, David R. and Hiromi Ono. 2000. Estimating the Extent of Intimate Contact between the Races: The Role of Metropolitan Area Factors and Union Type in Mate Selection. Presented at the annual meetings of the Population Association of America, March 23, Los Angeles, CA. . 2005. How Many Interracial Marriages Would There Be If All Groups Were of Equal Size in All Places? A New Look at National Estimates of Interracial Marriage. Social Science Research 34:23651. Harris, David and Jeremiah Joseph Sim. 2002. Who Is Multiracial: Assessing the Complexity of Lived Race. American Sociological Review 67:61427. Heaton, Tim and Stan L. Albrecht. 1996. The Changing Pattern of Interracial Marriage. Social Biology 43:20317. Hwang, Sean-Shong and Rogelio Saenz. 1990. The Problems Posed by Immigrants Married Abroad on Intermarriage Research: The Case of Asian Americans. International Migration Review 24:56376. Hwang, Sean-Shong, Rogelio Saenz, and Benigno E. Aguirre. 1997. Structural and Assimilationist Explanations of Asian American Intermarriage. Journal of Marriage and the Family 59:75872. . 1995. The SES Selectivity of Interracially Married Asians. International Migration Review 29:46991. Joyner, Kara and Grace Kao. 2000. School Racial Composition and Adolescent Racial Homophily. Social Science Quarterly 81:81025. Joyner, Kara and Grace Kao. 2005. Interracial Relationships and the Transition to Adulthood. American Sociological Review 70:56381. Joyner, Kara, Hongyu Wang, and Grace Kao. 2003. Stability of Interracial and Intraracial Dating Among Adolescents. Presented at the annual meeting of the American Sociological Association, August 1619, Atlanta, GA. Kalmijn, Matthijs. 1991a. Shifting Boundaries: Trends in Religious and Educational Homogamy. American Sociological Review 56:786800. . 1991b. Status Homogamy in the United States. American Journal of Sociology 97:496523.

366

The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

. 1993. Trends in Black/White Intermarriage. Social Forces 72:11946. . 1994. Assortative Mating by Cultural and Economic Occupational Status. American Journal of Sociology 100:42252. . 1998. Intermarriage and Homogamy: Causes, Patterns, Trends. Annual Review of Sociology 24:395421. Kalmijn, Matthijs and Henk Flap. 2001. Assortative Meeting and Mating: Unintended Consequences of Organized Settings for Partner Choices. Social Forces 79:1289312. Kennedy, Randall. 2003. Interracial Intimacies: Sex, Marriage, Identity, and Adoption. New York: Pantheon Books. Kennedy, Ruby Jo Reeves. 1944. Single or Triple Melting Pot? Intermarriage Trends in New Haven, 18701940. American Journal of Sociology 49:3319. Kouri, Krystan M. 2003. Black/White Interracial Couples and the Beliefs that Help Bridge the Racial Divide. Pp. 35572 in New Faces in a Changing America: Multiracial Identity in the 21st Century, edited by Loretta Windeters and Herman L. Debose. Thousand Oaks, CA: Sage Publications. Kouri, Krystan M. and M. Lasswell. 1993. Black-White Marriage: Social Change and Intergenerational Mobility. Marriage and Family Review 19(3/4):24155. Laumann, Edward O., John H. Gagnon, Robert T. Michael, and Stuart Michaels. 1994. The Social Organization of Sexuality: Sexual Practices in the United States. Chicago, IL: University of Chicago Press. Lee, Sharon M. and Keiko Yamanaka. 1990. Patterns of Asian American Intermarriage and Marital Assimilation. Journal of Comparative Family Studies 21:287305. Lewis, Richard Jr., George Yancey, and Siri S. Bletzer. 1997. Racial and Non-Racial Factors That Inuence Spouse Choice in Black/White Marriages. Journal of Black Studies 28:60 78. Lichter, Daniel T., Diane K. McLaughlin, George Kephart, and David J. Landry. 1992. Race and the Retreat from Marriage: A Shortage of Marriageable Men? American Sociological Review 57:78199. Lloyd, Kim M. 2001. Marriage Market Boundaries and Latino Mens Transition to First Marriage. Presented at the annual meetings of the Population Association of America, March 30, Washington, DC. Long, J. Scott. 1997. Regression Models for Categorical and Limited Dependent Variables. Thousand Oaks, CA: Sage Publications. Massey, Douglas S. and Nancy A. Denton. 1989. Hypersegregation in U.S. Metropolitan Areas: Black and Hispanic Segregation along Five Dimensions. Demography 26:37391. . 1993. American Apartheid: Segregation and the Making of the Underclass. Boston, MA: Harvard University Press. Mindiola, Tatcho, Jr, Yolanda Flores Niemann, and Nestor Rodriguez. 2003. Black-Brown: Relations and Stereotypes. Austin: University of Texas Press. Moody, James. 2001. Race, School Integration, and Friendship Segregation in America. American Journal of Sociology 107:679716. Mouw, Ted and Barbara Entwisle. 2001. A Country of Strangers? The Effect of Social Class, Residential Propinquity, and Mutual Activities on Multi-Racial Social Segregation in Schools. Presented at the annual meetings of the Population Association of America, March 29, Washington, DC. Murstein, Bernard I. 1986. Paths to Marriage. Beverly Hills, CA: Sage Publications.

The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

367

A Path toward Interracial Marriage

Rosalind Berkowitz King and Jenifer L. Bratter

National Center for Health Statistics. 1997. Public Use Data File Documentation: National Survey of Family Growth, Cycle 5: 1995, Users Guide. Hyattsville, MD: National Center for Health Statistics, Centers for Disease Control and Prevention, Public Health Service, U.S. Department of Health and Human Services. Pascoe, Peggy. 1996. Miscegenation Law, Court Cases, and Ideologies of Race in TwentiethCentury America. Journal of American History 83(June):4469. Pollard, Kelvin M. and William P. OHare. 1999. Americas Racial and Ethnic Minorities. Population Bulletin 54(3):147. Portereld, Ernest. 1982. Black-American Intermarriage in the United States. Marriage and Family Review 5:1737. Powers, Daniel and Christopher Ellison. 1995. Interracial Contact and Black Racial Attitudes: The Contact Hypothesis and Selectivity Bias. Social Forces 74(1):20526. Qian, Zhenchao. 1997. Breaking the Racial Barriers: Variations in Interracial Marriage between 1980 and 1990. Demography 34:26376. . 2000. Racial and Sex Variations in Interracial Marriage between Natives and Immigrants. Presented at the annual meetings of the Population Association of America, March 23, Los Angeles, CA. Qian, Zhenchao and Jose A. Cobas. 2004. Latinos Mate Selection: National Origin, Racial, and Nativity Differences. Social Science Research 33:22547. Quillian, Lincoln and Mary Campbell. 2003. Beyond Black and White: The Present and Future of Multiracial Friendship Segregation. American Sociological Review 68(4):54066. Romano, Renee. 2003. Race Mixing: BlackWhite Marriage in Post-War America. Cambridge, MA: Harvard University Press. Root, Maria P. 2001. Loves Revolution: Interracial Marriage. Philadelphia, PA: Temple University Press. Rosenfeld, Michael J. 2001. The Salience of Pan-National Hispanic and Asian Identities in U.S. Marriage Markets. Demography 38(2):16175. . 2002. Measures of Assimilation in the Marriage Market: Mexican Americans 19701990. Journal of Marriage and the Family 64:15262. Schoen, Robert and Robin M. Weinick. 1993. Partner Choice in Marriage and Cohabitation. Journal of Marriage and the Family 55(2):40814. Schoen, Robert and John Wooldredge. 1989. Marriage Choices in North Carolina and Virginia, 196971 and 197981. Journal of Marriage and the Family 51(2):46582. Schuman, Howard, Charles Steeh, and Lawrence Bobo. 1985. Racial Attitudes in America. Cambridge, MA: Harvard University Press. Sherkat, Darren. 2004. Religious Intermarriage in the United States: Trends, Patterns, and Predictors. Social Science Research 33:60625. Spickard, Paul. 1989. Mixed Blood Intermarriage and Ethnic Identity in Twentieth Century America. Madison: University of Wisconsin Press. Statacorp. 2001. Stata Statistical Software: Release 7.0. College Station, TX: Stata Corporation. Teachman, Jay. 2003. Premarital Sex, Premarital Cohabitation, and the Risk of Subsequent Marital Dissolution among Women. Journal of Marriage and the Family 65(2):44466. Thompson, Sharon. 1995. Going All the Way: Teenage Girls Tales of Sex, Romance, and Pregnancy. New York: Hill and Wang. U.S. Bureau of the Census. 1996. Statistical Abstract of the United States, 1995. Washington, DC: U.S. Government Printing Ofce.

368

The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

Rosalind Berkowitz King and Jenifer L. Bratter

A Path toward Interracial Marriage

Waters, Mary. 1990. Ethnic Options: Choosing Ethnic Identities in America. Berkeley: University of California Press. White, Michael J. 1986. Segregation and Diversity Measures in Population Distribution. Population Index 52(2):198221. Yancey, George. 2002. Who Interracially Dates? An Examination of the Characteristics of Those Who Have Interracially Dated. Journal of Comparative Family Studies 33(2):17990. Zubrinsky, Camille L. and Lawrence Bobo. 1996. Prismatic Metropolis: Race and Residential Segregation in the City of Angels. Social Science Research 25:33574.

The Sociological Quarterly 48 (2007) 343369 2007 Midwest Sociological Society

369

Вам также может понравиться