Interracial Marriage and the U.S. Military: A Test of Status Exchange and Own Race Preferences

Abstract

What roles do racial social distance and individual identity play in determining intermarriage among Black and White individuals in the U.S. military? This research uses interracial marriage as a measure of social distance and hypothesizes that the military lessens the social distance between White and Black individuals. More specifically, this article examines the relative education levels of Black and White married individuals in the military to test theories of market and dyadic exchange using the 2015–2019 American Community Survey (ACS) data. Evidence of weaker racial group boundaries is found for Black male and White female pairings where at least one spouse is in the military. There is also evidence of stronger own race preferences among Black women in the military.

Keywords

marriage interracial marriage military race education status exchange racial group boundaries

This article is broadly interested in the roles that institutions and identity play in determining intermarriage among Black and White individuals in the U.S. military, where race and military status are used as key characteristics of an individual’s identity. Interracial marriage is often used as a measure of racial social distance and the military is hypothesized to lessen the social distance between White and Black individuals. More specifically, this article examines status exchange and own race preferences among Black and White individuals in the military using the methodology found in Gullickson (2006). This method distinguishes between market and dyadic exchange and allows for an investigation of own race preferences first developed by Fu (2001).

Social distance measures the interconnectedness of groups (Borgardus, 1925). In terms of social networks, there is little evidence that the social distance between Whites and Blacks has improved over time (Smith et al., 2014). Interracial marriage can also shed light on the degree of separation between White and Black individuals and over time interracial marriage has increased, although rates remain lower than those that would occur with random matching (Fryer, 2007; Kalmijn, 1993; Torche & Rich, 2017).

There are several indicators that the U.S. military has lessened the social distance between White and Black individuals in the United States. Intermarriage rates are higher among military members (Farley, 1999; Houseworth & Grayson, 2018; Lundquist, 2004), military members reside in more racially integrated areas (Farley & Frey, 1994), Blacks are overrepresented in the military (Kane, 2006), and there is evidence of higher occupational prestige status among Blacks in the military (Moskos & Butler, 1996). The military may also add an additional dimension to one’s identity (Akerlof & Kranton, 2010; Daley, 2013; Devries et al., 2012) that may be important in partner selection.

Marriage is often studied utilizing the concept of assortative mating provided by Becker (1974) where individuals are predicted to match up on important traits to maximize the gains from marriage. Status exchange, first developed by Davis (1941) and Merton (1941), incorporates the role of trading characteristics in match formation and the majority of researchers use education as a key matching characteristic (Fu, 2001; Gullickson, 2006; Gullickson & Torche, 2014; Kalmijn, 1993, 1998; Rosenfeld, 2005; Schwartz et al., 2016). Education is used here as well as it is a good predictor of favorable socioeconomic conditions, but less likely to be endogenously related to marriage decisions.

In terms of race and education, status exchange predicts that Black spouses trade higher levels of education for marriage to a White spouse. There is evidence of status exchange among Black males married to White females (Fu, 2001; Gullickson, 2006; Kalmijn, 1993; Qian, 1997), but little evidence for Black female and White male marriages. Status exchange, however, relies on the assumption that a White spouse is preferred, by both Black and White individuals. Own race preferences (“in group preference”) assume that race trumps other measurable characteristics in marital formation (Fu, 2001). Own race preferences predict that educational matches are first made within racial groups and those who intermarry are more likely to have lower levels of education. Status exchange assumes hierarchical racial group boundaries, where own race preferences imply self-imposed boundaries that may arise either from Becker’s theory regarding gains to marriage or from a desire to preserve Black culture (Barth, 1969).

I first hypothesize that the military is an institution that lessons the social distance between Black and White individuals in the United States. If the racial group boundaries are lower in the military, relative to the civilian population, there should be less evidence of status exchange among military members. Own race preferences should also be less likely in the military, given an increase in the contact with others outside of one’s race (Farley & Frey, 1994) and the additional dimension of identity (military) for which individuals may match (Akerlof & Kranton, 2010; Daley, 2013; Devries et al., 2012).

To test these hypotheses, I estimate log-linear models for Black male–White female (BM-WF) marriages separately from White male–Black female (WM-BF) marriages following Gullickson (2006) using 2015–2019 American Community Survey (ACS) data. The two hypotheses are examined by incorporating market exchange terms, or educational boundaries, and dyadic exchange terms developed in Gullickson (2006). The educational boundary variables measure the relationship between intermarriage and the individual’s level of education and allow for an investigation of own race preferences. The dyadic exchange terms measure status exchange, which is the relationship between intermarriage and the relative education of the spouse.

The study of intermarriage is one of the most tangible ways to study race relations given the level of intimacy associated with marriage and family formation. Furthermore, this study examines race relations within an important U.S. institution, the military, and highlights structural factors that affect racial inequality. This research also adds to the literature on interracial marriage, specifically by providing a test of status exchange and own race preferences for a subset of the U.S. population that is hypothesized to have lower racial group boundaries.

Background and Predictions

Identity, Social Distance, and Group Boundaries

There is a vast and deep literature on the theoretical foundations of identity where the work of Hume (1739/1986) and Kant (1781/2007) is considered foundational. Researchers have defined individual identity in a variety of ways, where most definitions provide a description of the relation of self to others. For instance, James (1890) defines identify as “consciousness of personal sameness” and Mead (1934) writes that “the self is something which has a development; it is not initially there, at birth, but arises in the process of social experience and activity.” Stryker (1968, 1980) focuses on the role of hierarchies in determining where one might fit on a number of dimensions. Identity is a social construct and affects all aspects of an individual’s life to some degree, including marital decisions.

This study is particularly interested in the racial dimensions of one’s identity and a special identifier that results from being a member of the military institution in the United States. The level of social interaction in the United States is determined by the degree of residential and educational segregation, occupational segregation, and the level of racism and discrimination that the individual must exist in. It is important to note these realities; however, examining all the factors that may determine an individual’s identity as it pertains to race and preferences for a life partner is outside the scope of this article. Rather, the focus is on the distinct dimension of identity for military members as compared with civilians.

The military may de-emphasize race as a part of one’s identity and/or create a new component of one’s identity associated with the military. How individuals define their “group” depends on several identifiers, including race and gender, occupation, and generation of birth. Akerlof and Kranton (2010) argue that the military can alter one’s identity more than civilian occupations for several reasons. More recent military recruits are voluntary given that the Draft expired in 1973, which makes them more likely to “buy into” the de-individualizing training. Furthermore, their isolation during training or in residence on military base allows for the group identity to solidify without outside interference. In addition, outcomes are generally group-dependent rather than based on individual effort. Devries et al. (2012) provide a list of commonalities among military members that help to create a military culture that is distinct from civilian culture: rank, hierarchy, the concept of “mission first,” a high level of self-reported spirituality, combat experience, and specific military values that vary by branch. Daley (2013) argues that military involvement shapes a dimension of one’s identity that is similar to their ethnic identity.

Social distance measures the interconnectedness of groups (Borgardus, 1925). Smith et al. (2014) find little change between the 1980s and 2004 with regard to race and racial networks in the United States, where race/ethnic divides are the leading determinants of social interactions. Marriage is also often studied as a measure of the social distance between racial groups (George & Yancey, 2004). Low rates of interracial marriage indicate a large distance between groups and hence a wide racial group boundary. Social distance determines group boundaries and researchers often use the two terms interchangeably.

There are several indicators that the U.S. military has lessened the social distance between White and Black individuals in the United States. First, intermarriage rates are higher among military members (Farley, 1999; Houseworth & Grayson, 2018; Lundquist, 2004) and have been increasing over time (Fryer, 2007).

In addition, a large percentage of military members are more likely to live in more integrated areas (Farley & Frey, 1994). Furthermore, Blacks are overrepresented in the military (Kane, 2006), increasing the opportunity for contact compared with civilian populations. Allport (1954) and Cohen (1982) argue that intergroup contact could have positive or negative effects on stereotypes and racial discrimination, depending on the equality status of minorities and the race of authority figures. The military environment is argued to be one of the more equitable environments for Blacks in the United States with regard to pay, promotion, on-the-job training, and respect of workplace hierarchy (Moskos & Butler, 1996; Segal et al., 2007). Furthermore, Moskos and Butler (1996) find that both Black and White individuals report a more favorable racial climate in the military than in the civilian population. Therefore, the military is likely an institution for which intergroup contact will lessen prejudice and discrimination compared with civilian workplaces.

Interracial Marriage Theory

People get married for a wide variety of reasons; financial security, family formation, the general gains from being in a partnership, love, and so on. Economists often study marriage utilizing the concept of assortative mating provided by Becker (1974) where individuals are predicted to match up on important traits to maximize the gains from marriage. This lens allows for researchers to use measurable characteristics to identify patterns in marriage. Education is often used as an important measurable trait upon which people match (Becker, 1974; Fryer, 2007; Furtado & Theodoropoulos, 2011; Rosenfeld, 2005; Schwartz & Mare, 2005). Education is used here as well as it is a good predictor of favorable socioeconomic conditions, but less likely to be endogenously related to marriage decisions, like other measurable characteristics such as earnings.

Status Exchange

Assortative mating does not fully address the complex nature of marital decisions in that it only provides an examination of traits that are matched among spouses. Other theories have emphasized the role of trading characteristics in match formation. Status exchange, first developed by Davis (1941) and Merton (1941), predicts that dissimilarity in one area, such as race, must be compensated for in another area, such as education (Grossbard-Schechtman & Fu, 2002). The underlying assumption is that the racial group boundaries are such that a White spouse is preferred. These preferences may be based on symbolic boundaries, where White is preferred based on decades of racism and discrimination resulting in White supremacy or on social boundaries, where access to resources influences marital decisions. Status exchange theory predicts that highly educated Blacks and low educated Whites are more likely to intermarry. There is evidence of status exchange among Black males married to White females (Fu, 2001; Gullickson, 2006; Kalmijn, 1993; Qian, 1997), but little evidence for Black female and White male marriages. These results may indicate that gender roles operate differently in Black and White households.

Gullickson’s (2006) method, the preferred method here, is based on Fu (2001) and allows for an investigation into both hypergamy (marrying up in terms of education) and hypogamy (marrying down in terms of education) while also separating out joint (relative education levels of the spouse) and marginal (the education of the individual) educational effects. Gullickson argues that education can affect interracial marriage in two ways: market exchange and dyadic exchange. The market exchange theory posits that education is traded for Whiteness and as education increases the probability of intermarriage decreases for Whites and increases for Blacks. With market exchange, the education level of the spouse is not relevant; only the currency of an individual’s education to garner a White partner is important. Market exchange predicts higher levels of intermarriage among highly educated Blacks and low educated Whites. Gullickson (2006) finds evidence of market exchange for Black spouses, but no clear significant pattern for Whites. Using the same methodology, Gullickson and Torche (2014) examine interracial couples in Brazil and find evidence of market exchange for both Black and White individuals.

Dyadic exchange, on the contrary, assumes that exchange depends on the education level of the individual relative to the spouse as well as the relative races of the partners. Dyadic exchange is the closest interpretation of Merton’s (1941) theory. Gullickson (2006) finds evidence of hypergamy utilizing his dyadic exchange terms for BM-WF couples, but finds no evidence of hypogamy. Gullickson and Torche (2014) find no evidence of dyadic exchange in Brazil.

Own Race Preferences

Status exchange alone does not fully capture the role of racial group boundaries in marriage decisions. Own race preferences (“in group preference”) assume that race trumps other measurable characteristics in marital formation (Fu, 2001). The “in group preference” predicts that the most attractive in each racial group marry the most attractive in their respective racial group. The “in group preference” model results in highly educated Whites (Blacks) marrying highly educated Whites (Blacks). Therefore, intermarried Whites and Blacks would be drawn from the lower end of each respective education distribution. Evidence of own race preferences is impossible to examine for White individuals, as the predictions for both status exchange and own race preferences predict higher educated Whites being more likely to marry Whites.

Own race preferences for Black individuals can be measured through the market exchange terms (Gullickson, 2006). Recall that market exchange rests on the assumption that Whiteness is preferred. A decrease in intermarriage as education increases for Black individuals may be indicative of own race preferences. There is little to no evidence of own race preferences for Blacks in the United States.

Military Predictions

The differences between military and civilian populations in relation to identity and racial group boundaries discussed above may affect marriage decisions through two avenues. First, the changes in identity that may occur with military involvement may also alter an individual’s preferences for life partner and add an additional dimension on which to match. Research shows a connection between different identity structures and social networks and it is likely that extends to romantic partnerships (Mok et al., 2007). In addition, an institution with lower racial group boundaries may also fundamentally affect an individual’s racial belief system as discussed above, also altering their preferences for a partner. These mechanisms reside within an individual and therefore will affect intermarriage regardless of if the spouse is also a military member. On the contrary, because individuals often match at work or school (Rosenfeld & Thomas, 2012), the military may operate as a marriage market itself with lower racial group boundaries, and therefore intermarriage may be more likely for individuals who meet their spouse during through military channels. The following hypotheses hold regardless of whether the military affects an individual’s psyche or externally through a more diverse marriage market.

First, there should be smaller differences in the relative education levels between the intermarried and intramarried for the military sample. If racial boundaries do not exist, there should be no difference in the relative education levels between intermarried and intramarried couples. It follows that lower racial group boundaries should result in smaller differences between intermarried and intramarried couples. This hypothesis is first addressed with simple statistics that show a smaller variation in the relative education levels between intermarried and intramarried in the military compared with the civilian population.

Second, there should be less evidence of status exchange among military members if the military is an institution with lower racial group boundaries. Evidence of status exchange is effectively evidence of hierarchical racial group boundaries and if military involvement changes that, we will see less exchange on race and education. Furthermore, if the military adds an additional dimension for which individuals may match, there should be fewer status exchange marriages based on education and race. This hypothesis is tested by comparing the coefficients on the dyadic and market exchange terms for the U.S. civilian sample compared with the military sample.

Own race preferences should also be less likely in the military, given an increase in the contact with others outside of one’s race and the additional dimension of identity for which individuals may match. This hypothesis is tested by comparing the coefficients on the market exchange terms for the U.S. civilian sample compared with the military sample for Blacks only given that it is impossible to test for Whites as discussed above. Evidence of own race preferences can be seen through positive coefficients on the market exchange terms for Blacks.

Data and Restrictions

Samples of native-born, non-Hispanic Black alone and White alone men and women born between 1966 and 1992 (individuals over the age of 25 from Generation X and Y) are drawn from the 2015–2019 ACS, where all tables and analyses are adjusted using the person weights provided by the ACS. I exclude individuals who list multiple races to provide a sharper definition of intermarriage. The sample is further restricted to the native-born to examine intermarriage among those most likely to be exposed to U.S. customs and norms. Individuals are identified as intermarried if they are married to a spouse who reports a different race, where race of spouse is also restricted to non-Hispanic Black or White alone. Samples are further restricted to only include those married once with spouse present. Those who have been married more than once are excluded as they may induce bias. Individuals who have been married more than once are more likely to be intermarried (Chiswick & Houseworth, 2011) and individuals in the military are more likely to be married more than once (Houseworth & Grayson, 2018).

Individuals are identified as “military” using a combination of the detailed employment status variable that identifies those employed in the armed forces, both at work and not at work but with a job, and the veteran status variable. Individuals who ever served in any branch of the U.S. military are considered military members. Branches include the Army, Navy, Marine Corps, Air Force, and Coast Guard.

The sample is restricted to include individuals between the ages of 25 and 67 from the two generations that will generate the largest number of women and minority populations in the military: Generation X: those born between 1966 and 1976; and Generation Y: those born between 1977 and 1992. I restrict to over 25 to allow for the majority of individuals sampled to have completed their highest level of education.

Descriptive Statistics

Table 1 summarizes the possible combinations of relative race and education for both spouses and is adapted from Schwartz et al. (2016) for the two-race scenario studied here, where HE = husband’s education, WE = wife’s education, HR = husband’s race, WR = wife’s race, educational hypogamy = wives “marry down,” educational homogamy = spouses share traits, educational hypergamy = husbands “marry down.” Cell row (i) and column (j) positions are shown in brackets [i,j]. There are three types of educational pairings: educational homogamy, HE = WE, where both spouses have the same level of education; WE > HE, where wives have a higher level of education; and HE > WE, where husbands have a higher level of education. Note that black husband is denoted with BH, white wife with WW, white husband with WH, and black wife with BW. There are three types of racial pairings: racial endogamy or intramarriage, where both spouses have the same race; relationships between Black men and White women (BH-WW); and relationships between White men and Black women (WH-BW).

Table 1.

Classification of Marriages by Spouse’s Relative Education and Race.

		Spouse’s relative race
Spouse’s relative education		BH-WW	WH-WW, BH-BW	WH-BW
	HE < WE (hypogamy)	[1,1]	[1,2]	[1,3]
		Wife-advantaged		Exchange of wife’s E for husband’s R
	WE = HE (homogamy)	[2,1]	[2,2]	[2,3]
	HE > WE (hypergamy)	[3,1]	[3,2]	[3,3]
		Exchange of husband’s E for wife’s R		Husband-advantaged

Source. Schwartz et al. (2016).

Note. Cell row (i) and column (j) positions are shown in brackets [i, j]. HE = husband’s education; WE = wife’s education; hypogamy = wives “marry down”; BH = black husband; WW = white wife; WH = white husband; BW = black wife; homogamy = spouses share traits; hypergamy = husbands “marry down.”

Table 2 shows the percent of individuals in an interracial marriage by gender, race, and military status. White men and women, and Black women in the military are all more likely to be intermarried, whereas Black men in the military are just as likely to be intermarried as the civilian sample. Intermarriage rates are 1.3 (1.7) times higher among Black (White) military women compared with their civilian counterparts. Intermarriage rates are 2.1 times higher for White men in the military.

Table 2.

Descriptive Statistics of Married Individuals by Group.

	Women				Men
	White		Black		White		Black
	Military	Civilian	Military	Civilian	Military	Civilian	Military	Civilian
Percent intermarried	1.80	1.07	7.25	5.41	0.78	0.37	14.60	14.61
Percent Black			14.33	5.64			8.46	5.46
Education
Less high school	0.49	2.06	0.27	3.14	0.54	3.26	0.67	5.08
High school	16.45	21.02	14.37	28.21	28.57	27.65	31.58	41.02
Some college	33.53	24.04	37.01	31.18	33.56	22.32	38.86	28.04
College plus	49.54	52.88	48.35	37.48	37.32	46.76	28.89	25.86
Sample size, in units of couple	15,277	1,332,698	2,556	79,587	123,723	1,224,252	11,428	70,715

Source. 2015–2019 American Community Surveys.

Note. Sample restricted to native-born White and Black individuals aged 25 to 51 years, married only once.

Table 2 also shows the education distributions of married individuals by gender, racial group, and military status. Individuals are grouped into one of four education groups: less than high school, high school diploma, some college, and those with a bachelor’s degree or higher. Table 2 shows married Black men and women in the military have higher levels of education than their civilian counterparts.

Tables 3 and 4 show the percentage of individuals within each type of marriage, defined by the racial and educational composition of the marriage (columns sum to 100) for the civilian and military samples separately. Racial group boundaries are likely to be lower in the military if there is a smaller difference in the relative education levels between intramarried and intermarried than the civilian sample. I compare the intermarried spouse with their respective intramarried counterparts. For instance, Black husbands in BH-WW marriages are compared with BH-BW marriages and White wives in BH-WW marriages are compared with WH-WW marriages. The differences between intermarried and intramarried are smaller for Black men and women in the military (Table 4) compared their civilian counterparts (Table 3) for all relative education categories. However, the opposite is true for White women. White men generally have smaller differences between intermarried and intramarried when we consider those in the military (Table 4) compared with civilians (Table 3), with the exception of hypergamous marriages, where the spouse has the lower level of education. Overall, when intermarried Black men, Black women, and White men are compared with their intramarried counterparts the variation in relative education is smaller in the military samples, which supports the theory of lower racial group boundaries in the military.

Table 3.

Percent of Marriages by the Racial and Educational Composition of the Marriage for Civilian Only (Columns Sum to 100).

Educational composition of the marriage	Racial composition of the marriage
Educational composition of the marriage	BH-WW	BH-BW	WH-WW	WH-BW
HE < WE (hypogamy)	31.45	26.24	35.32	30.97
WE = HE (homogamy)	50.63	59.03	50.57	52.77
HE > WE (hypergamy)	17.92	14.74	14.11	16.27

Note. HE = husband’s education; WE = wife’s education. BH = black husband; WW = white wife; WH = white husband; BW = black wife.

Table 4.

Percent of Marriages by the Racial and Educational Composition of the Marriage for Military Only (Columns Sum to 100).

Educational composition of the marriage	Racial composition of the marriage
Educational composition of the marriage	BH-WW	BH-BW	WH-WW	WH-BW
HE < WE (hypogamy)	30.11	28.78	34.21	31.26
WE = HE (acthomogamy)	47.09	51.27	46.83	47.46
HE > WE (hypergamy)	22.8	19.94	18.96	21.28

Source. 2015–2019 American Community Surveys.

Note. Sample restricted to native-born White and Black individuals aged 25 to 51, married only once. HE = husband’s education; WE = wife’s education. BH = black husband; WW = white wife; WH = white husband; BW = black wife.

Method

I use a log-linear model to examine the education and race patterns of those in the military following the methodology found in Gullickson (2006). The 2015–2019 ACS data, restricted as discussed above, are flattened into a 2 × 2 × 4 × 4 × 2 × 9 table of HR by WR by HE by WE by cohort by region. The base model includes variables accounting for husband’s race, wife’s race, husband’s education, wife’s education, the interaction between husband’s and wife’s race (racial endogamy), the interaction between husband’s and wife’s education level (assortative mating), and the distribution of education for each racial group. i and j represent the race of the husband and wife, respectively, and k and i represent the education level of the husband and wife, respectively.

The base model also includes a control for geographic region. Controlling for region ensures that the opportunity for contact for members of different races is accounted for, albeit very crudely. Geographic region, GR, is a variable accounting for nine regions, including the New England, Middle Atlantic, East North Central, West North Central, South Atlantic, East South Central, West South Central, Mountain, and Pacific Divisions. This variable is interacted with both husband’s and wife’s education, as well as both husband’s and wife’s race, and the interaction between both husband’s education and race and wife’s education and race. A control for generation Y, Y, is also included and interacted with both husband’s and wife’s race as well as both husband’s and wife’s education.

The base model also includes a control, x, for partners who both attended college. This variable is a crude attempt to measure the impact of college as a marriage market itself. A better measure would control for the educational institution and its location, however, that is outside the scope of this research. There are four education levels: less than high school (1), high school (2), some college (3), and college plus (4). X is equal to 1 if k equals 3 or 4 and l equals 3 or 4 and the couple is intermarried:

\begin{matrix} l o g (f r e q_{i j k l m}) = H R_{i} + W R_{j} + H E_{k} + W E_{l} + H R \times W R_{i j} + H E \times W E_{k l} + H R \times H E_{i k} \\ + W R \times W E_{j l} + G R_{m} + H R \times G R_{i m} + W R \times G R_{j m} + H E \times G R_{k m} + W E \\ \times G R_{l m} + H R \times H E \times G R_{i k m} + W R \times W E \times G R_{j l m} + Y_{G e n Y = 1} + H R_{i} \times Y \\ + W R_{j} \times Y + H E_{k} \times Y + W E_{l} \times Y + β x_{i j k l} . \end{matrix}

The data are effectively split into two samples and the above model is run separately because there may be differences in intermarriage based on the race and gender of the spouse. For instance, relative education levels differ between BM-WF and WM-BF marriages. Intermarriages where the Black spouse is a male are more common and are also more likely to provide evidence of status exchange (Gullickson, 2006; Kalmijn, 1993; Qian, 1997). Furthermore, the role of gender in status exchange is important to consider. It may be more socially acceptable for White females who marry outside of their race to marry up in terms of education than White males. Sample 1 includes all intramarriages combined with intermarriages where the male spouse is Black (BM-WF). Sample 2 includes all intramarriages combined with intermarriages where the female spouse is Black (WM-BF). Therefore, the BM-WF samples compare intermarriage among Black men with their intramarried counterparts and intermarried White females with their intramarried counterparts. The WM-BF samples compare intermarried White men with their intramarried counterparts and intermarried Black females with their intramarried counterparts. Each sample is also parsed into civilian and military samples.

To test the hypotheses discussed above, I incorporate market exchange terms, or educational boundaries, and dyadic exchange terms developed in Gullickson (2006) to the base model described above. Dyadic exchange terms developed by Gullickson (2006) are used to measure evidence of status exchange. Intermarriages are split into three categories: (a) Both partners have the same level of education; (b) the Black partner has more education than the White partner (White educational hypergamy); or (c) the Black partner has less education than the White partner (White educational hypogamy). I create four dichotomous variables for intermarried individuals reflecting status exchange marriages where the White spouse marries up or down in terms of education. Therefore, for White women in intermarriages there are two dichotomous variables: one equal to one if she married a Black spouse with a higher level of education (hypergamy) and one in which she married a spouse with a lower level of education (hypogamy). Corresponding variables are created for intermarried White men.

Market exchange terms (educational boundary variables) are also incorporated into the model. The educational boundary variables measure the change in the expected number of interracial marriages across three educational boundaries: high school compared with less than high school, some college compared with high school, and college compared with some college. Dichotomous variables are created for White males in WM-BF marriages, equal to one for each education level “step.” Dichotomous variables are also created for the Black females in these marriages, as well as the Black males and White females in BM-WF marriages. Using the same “stairstep” coding found in Gullickson, each coefficient measures the effect of the identified education level compared with all levels below. For instance, intermarried men with a college degree or more are compared with their counterparts with some college or below. This results in 12 dichotomous variables; three each for White men and women and Black men and women who identify as intermarried.

Analysis

Civilian Sample

To provide a comparison group for the military samples, I provide results for the civilian samples (Table 5). Negative coefficients indicate a decrease in the likelihood of intermarriage. The patterns for the civilian sample mostly match those found in Gullickson (2006). An examination of the educational boundary terms provides evidence of both own race preferences and status exchange for Black men and status exchange for Black women. The coefficients on the White educational boundaries are almost all significantly negative, either providing evidence of status exchange or own race preferences for White women and men.

Table 5.

Coefficients From Select Variables in Log Linear Models.

		BM-WF
		Civilian		Military
Black educational boundaries
HS vs. less than HS	−0.082**	(.012)	−0.907**	(.060)
Some college vs. HS	0.051**	(.012)	0.454**	(.029)
College vs. some college	0.328**	(.008)	0.149**	(.020)
White educational boundaries
HS vs. less than HS	−0.097**	(.013)	−0.98**	(.039)
Some college vs. high school	−0.117**	(.010)	−0.061*	(.025)
College vs. some college	−0.237**	(.007)	−0.188**	(.019)
College propinquity	−0.011	(.012)	−0.209**	(.031
Hypergamy	0.196**	(.010)	−0.074**	(.026)
Hypogamy	0.108**	(.009)	0.136**	(.023)
BIC	247,596		50,695
df	161		161
		WM-BF
		Civilian		Military
Black educational boundaries
HS vs. less than HS	0.094**	(.019)	1.011*	(.411)
Some college vs. high school	0.425**	(.021)	−0.257**	(.039)
College vs. some college	0.233**	(.014)	−0.235**	(.028)
White educational boundaries
HS vs. less than HS	−0.495**	(.029)	−1.258**	(.062)
Some college vs. HS	0.202**	(.017)	−0.334**	(.037)
College vs. some college	−0.273**	(.013)	0.071**	(.026)
College propinquity	−0.463**	(.022)	0.308**	(.043)
Hypergamy	−0.131**	(.016)	0.165**	(.033)
Hypogamy	0.158**	(.017)	0.283**	(.035)
BIC	238,734		49,028
df	161		161

Source. 2015−2019 American Community Surveys.

Note. Standard errors in parentheses. Sample restricted to native-born White and Black individuals aged 25 to 51, married only once. Hypergamy indicates White upward educational mobility and hypogamy indicates White downward educational mobility. Significance noted at .05 (*) and .01 (**). HS = high school; BIC = Bayesian information criterion.

The coefficients are positive on the dyadic terms for intermarried White women, both the term measuring White upward mobility, hypergamy—supporting status exchange, and the term measuring White downward mobility, hypogamy—in opposition to status exchange. Intermarried White men are less likely to marry up and more likely to marry down, contrary to what is predicted by status exchange and also in opposition to that found in Gullickson (2006).

Military Sample

This section provides a discussion of the differences between the military and civilian samples. Chi-square tests were conducted to determine whether these differences are statistically significant and are available upon request. I include a discussion of the results that are not statistically significant given the ongoing discussion on the relative importance of statistical significance. For instance, some researchers argue that relying only on statistical significance is misleading, especially if the differences are large enough to be meaningful in the context of the real world (Wasserstein et al., 2019). It is also important to note that the military has educational requirements (a high school diploma or general educational development [GED] is the unofficial standard) and less than 1% of the military sample reported less than a high school level of education. Therefore, a discussion of the lowest educational boundary for the military sample is excluded.

The patterns for the Black and White educational boundary terms are similar for civilian and military samples of BM-WF marriages. The Black educational boundary terms are positive for two higher categories and the White educational boundary terms are all negative, similar to the civilian sample. Recall that negative coefficients on the White boundary terms support either own race preferences or status exchange, where positive coefficients on the Black boundary terms support status exchange and negative coefficients support own race preferences.

A comparison of the BM-WF military and civilian samples provides mostly favorable evidence that status exchange is less likely among military members. For instance, the coefficient on the highest Black educational boundary for the military is half the size of the civilian sample. For intermarried White women in the military, the two highest educational boundaries are between 50% and 20% smaller than their civilian counterparts. Surprisingly, none of these differences are statistically significant.

Furthermore, the coefficient on the term measuring hypergamy is negative for the BM-WF military sample, providing strong evidence that status exchange is less likely among White women in the military. Consider that the probability of intermarriage decreases by 7.4% for White military women who are married to a male with a higher level of education, where the probability increases for their civilian counterparts by 19.6%. The coefficient on the term measuring hypogamy is positive, and larger than the civilian sample, indicating that White women in the military are more likely to marry down than their intramarried counterparts and at a higher magnitude for the military sample. This result also supports the theory that status exchange is less likely among military members. However, these differences are also not statistically significant.

The results for the WM-BF sample also provide some support that the military sample has fewer marriages that fit the definition of status exchange. White men in the military with a college degree are 36% more likely to be intermarried than those with some college. When we compare that with the civilian result that indicates a 24% reduction in the likelihood of intermarriage, we find support for the hypothesis that status exchange is less likely in the military. However, the educational boundary for some college versus high school is negative while the coefficient for the civilian sample is positive and this difference is statistically significant. This indicates that as education increases from high school to some college, White men in the military sample are less likely to be intermarried.

The dyadic terms for WM-BF military marriages capture mixed evidence as well; however, neither is statistically significantly different from the civilian sample. White men in the military are more likely to be intermarried if their spouse has a higher level of education, which may support status exchange as being more likely in the military. However, hypogamy is more likely for the sample of military WM-BF marriages, which does support that status exchange is less likely in the military.

Therefore, it appears that there is strong evidence of a decrease in the likelihood of status exchange in the military samples of BM-WF marriages and mixed evidence in the WM-BF samples. The mixed results for the WM-BF sample are not surprising, given that previous research only finds significant results for BM-WF marriages (Fu, 2001; Gullickson, 2006; Kalmijn, 1993; Qian, 1997). The weaker evidence of status exchange may be driven by lower racial group boundaries for individuals in the military or by the increase in the dimensions of one’s identity through military involvement. There may also be a lower level of institutional racism in the military that enables minority groups to obtain a higher social status than they would be otherwise able to obtain in the civilian labor market. Consider that Black men in the military are more likely to be proportionally represented in positions of authority than civilian Black men (Moskos & Butler, 1996). It is also important to consider that men and women enter the military for very different reasons. Education may not be the status that is exchanged; rank or job title may provide a better measure of exchange among military member given that education matters less for advancement in the military.

An interesting and important result is found for the military sample of WM-BF marriages when the Black educational boundaries are examined. The coefficients on the two highest Black educational boundaries are negative for the military sample and positive for the civilian sample and these differences are statistically significant at the 5% level. Highly educated Black females in the military are more likely to have own race preferences than their civilian counterparts. One possible explanation for this phenomenon may be that the military provides a more favorable marriage market for Black females who place a premium upon marrying within their group. There is a larger own race marriage market for Black women in the military as education increases given that there are proportionally more highly educated Black men in the military (Table 2). Consider that an increase in the number of marriageable Black men substantially increases the likelihood of marriage for Black women (Craigie et al., 2018; Testa & Kroghn, 1995). However, there is a disproportionate number of Black women in the military (14%) compared with only 8% for Black men (Table 2).

What is different among the military sample of Black men that makes it more likely for a Black woman to match with a highly educated Black male than for their civilian counterparts? Because an individual’s identity is shaped by the social conditions they develop in and not determined at birth, it is likely that racial preferences for a partner to some degree depend on the racial makeup of the neighborhood one grew up in. While examining the racial composition of an individual’s childhood/early adulthood neighborhood is outside of the scope of this research, I can provide information on one facet of socioeconomic background, parental education. While parental education is not an ideal measure of the racial makeup of the neighborhood one grew up in, there is a strong link between education level and the racial makeup of one’s neighborhood (Goldsmith, 2009; Wei et al., 2018).

Using the last interview of the 1997 National Longitudinal Survey of Youth (NLSY97) in which respondents were 28 to 32 years old, I find that the military is comprised of a relatively larger percentage of highly educated Black men and women from lower socioeconomic backgrounds, compared with their highly educated civilian counterparts (Table 6). Consider that 25% of fathers of highly educated (some college or more) Black men in the military had an education level equal to some college or more. For fathers of highly educated Black civilian men, this figure was 36%. The pattern is even starker for Black women in the military, where there are no fathers of highly educated Black women in the military with an education level greater than a high school diploma. I also examined mother’s education level and found similar patterns.

Table 6.

Distribution of Father’s Education for Highly Educated (Some College or More) Sons and Daughters.

	Black		White
	Military	Civilian	Military	Civilian
Men
Less HS	50	31.85	12.9	12.9
HS	25	32.48	41.94	23.42
Some college	16.67	19.11	19.35	19.35
College plus	8.33	16.56	25.81	44.32
Sample size	446	1569	829	1581
Women
Less HS	52.63	29.91	11.63	16.01
HS	47.27	39	34.88	28.23
Some college		15.54	13.95	20.4
College plus		15.54	39.53	35.35
Sample size	318	764	802	1450

Source. 1997 National Longitudinal Survey of Youth.

Note. Respondents born between 1980 and 1984, aged 28 to 32 years and restricted to respondents with some college or more only. HS = high school.

Furthermore, Carter et al. (2017) posit that White individuals are more likely to enlist during times of conflict as an act of patriotism, whereas Black individuals are more likely to enlist as a reaction to personal socioeconomic conditions. The opportunity cost of military enlistment is lower for minorities in the United States given the fewer opportunities they are presented with in the civilian labor force due to discrimination, both through personal prejudice and institutional discrimination. In addition, minorities with less access to funds may find the military as an avenue to higher education as the military offers unique educational opportunities for its members.

Given the racially segregated nature of the United States (Massey & Tannen, 2015), Black men and women from lower socioeconomic backgrounds are likely to have grown up in neighborhoods with a large Black population and therefore, on average, are more likely to have own race preferences if their identity was shaped by the environment in which they were raised.

Contrary to that found for military Black men, civilian Black men with higher levels education are more likely to come from more affluent backgrounds, given the multiple barriers to higher education for Black men in the United States. Table 6 shows that the percentage of highly educated Black men with fathers who are also highly educated is higher among the civilian sample. Affluent neighborhoods are more likely to be White (Reardon et al., 2015) and therefore civilian Black men with high levels of education may have weaker own race preferences.

These differences will effectively generate a more favorable marriage market for Black women in the military, given the higher opportunity for contact with highly educated Blacks with a greater likelihood of own race preferences. Black women are much less likely to intermarry compared with their male counterparts. Table 2 shows that roughly 7% and 5% of the military and civilian samples of Black women are intermarried compared with 14% of Black men in both the military and civilian samples. Furthermore, Crowder and Tolnay (2000) find that the increase in intermarriage among Black males negatively affects the likelihood of marriage for Black females.

Summary

I estimate log-linear models using 2015–2019 ACS survey data to examine status exchange and own race preferences for military members using the methodology found in Gullickson (2006). First, I hypothesize that lower racial group boundaries will result in smaller differences in the relative education levels between the intermarried and intramarried for the military sample. Overall, when intermarried Black men and women, and White men are compared with their intramarried counterparts, the variation in relative education is smaller in the military samples supporting the theory of lower racial group boundaries in the military.

I also find evidence of lower racial group boundaries in the military for BM-WF marriages. A lower likelihood of status exchange is seen through the market exchange terms. Furthermore, the dyadic exchange terms show that the overriding pattern for military members is the opposite of status exchange. There is less evidence for WM-BF marriages; however, evidence of status exchange is not previously found for these matches. These results may indicate that gender roles operate differently within marriages where at least one spouse is a member of the military and within Black and White households. Furthermore, military members may use other measures of status on which to exchange such as rank. Longitudinal data that provide information on rank and pay grade as well as information on the spouse’s employment and earnings would be ideal.

Previous studies on interracial marriage fail to find evidence of own race preferences. Typically research on interracial marriage assumes implicitly that Whiteness is preferred and that racial hierarchy in the United States supersedes own race preferences. However, I find evidence of own race preferences among Black women in the military, likely attributable to a more favorable marriage market. This result highlights an important social consequence of the socioeconomic inequality between Blacks and Whites in the United States; it is harder for highly educated Black women to form families within their race than it is for their White counterparts.

Footnotes

Acknowledgements

I would like to thank Eva Dziadula, Jonathan Fisher, Aaron Gullickson, Pamela Icyeza, Jo Beth Mertens, and Christine Schwartz for their helpful feedback, comments, and questions.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Author Biography

Christina A. Houseworth is an associate professor of economics at Hobart and William Smith Colleges. She earned her PhD and MA in economics from the University of Illinois at Chicago and holds a BA in economics from Southern Illinois University in Carbondale. Her research focuses on issues related to marriage, wages, immigration, and the military.

References

Akerlof

G. A.

Kranton

R. E.

(2010). Identity economics: How our identities shape our work, wages, and well-being. Princeton University Press.

Allport

(1954). The nature of prejudice. Addison-Wesley.

Barth

(1969). “Introduction” and “Pathan identity and its maintenance.” In Barth

(Ed.), Ethnic groups and boundaries: The social organization of culture difference (pp. 9–38, 117–134). Universitetsforlaget; Allen & Unwin.

Becker

G. S.

(1974). A theory of marriage: Economics of the family. University of Chicago Press.

Borgardus

E. S.

(1925). Measuring social distance. Journal of Applied Sociology, 9, 299–308.

Carter

Smith

Wojtaszek

(2017). Who will fight? The all-volunteer army after 9/11. American Economic Review, 107(5), 415–419.

Chiswick

B. R.

Houseworth

(2011). Ethnic intermarriage among immigrants: Human capital and assortative mating. Review of Economics of the Household, 9, 149–180.

Cohen

(1982). Expectation states and interracial interaction in school settings. Annual Review of Sociology, 8, 209–235.

Craigie

Myers

S. J. R.

Darity

W. A. J. R

. (2018). Racial differences in the effect of marriageable males on female family headship. Journal of Demographic Economics, 84(3), 231–256.

10.

Crowder

K. D.

Tolnay

S. T.

(2000). A new marriage squeeze for black women: The role of racial intermarriage by black men. Journal of Marriage and Family, 62(3), 792–807.

11.

Daley

(2013). Understanding the military as an ethnic identity. In Munson

Daley

(Eds.), Social work practice in the military. Routledge.

12.

Davis

(1941). Intermarriage in caste societies. American Anthropologist, 43, 376–395.

13.

Devries

M. R.

Hughes

K. H.

Watson

Moore

B. A.

(2012). Understanding the military culture. In Moore

B. A.

(Ed.), Handbook of counseling military couples (pp. 7–18). Routledge.

14.

Farley

(1999). Racial issues: Recent trends in residential patterns and intermarriage. In Smelser

N. J.

Alexander

J. C.

(Eds.), Diversity and its discontents: Cultural conflict and common ground in contemporary American Society (pp. 85–128). Princeton University Press.

15.

Farley

Frey

W. H.

(1994). Changes in the segregation of whites from blacks during the 1980s: Small steps toward a more integrated society. American Sociological Review, 59(1), 23–45.

16.

Fryer

R. G.

(2007). Guess who’s been coming to dinner? Trends in interracial marriage over the 20th century. The Journal of Economic Perspectives, 21(2), 71–90.

17.

(2001). Racial intermarriage pairings. Demography, 38(2), 147–159.

18.

Furtado

Theodoropoulos

(2011). Interethnic marriage: A choice between ethnic and educational similarities. Journal of Population Economics, 24(2), 1257–1279.

19.

George

Yancey

(2004). Taking stock of America’s attitudes on cultural diversity: An analysis of public deliberation on multiculturalism, assimilation, and intermarriage. Journal of Comparative Family Studies, 35(1), 1–19.

20.

Goldsmith

P. R.

(2009). Schools or neighborhoods or both? Race and ethnic segregation and educational attainment. Social Forces, 87(4), 1913–1941.

21.

Grossbard-Shechtman

(2002). Women’s labor force participation and status exchange in intermarriage: A model and evidence for Hawaii. Journal of Bioeconomics, 4(3), 241–268.

22.

Gullickson

(2006). Education and black-white interracial marriage. Demography, 43(4), 673–689.

23.

Gullickson

Torche

(2014). Patterns of racial and educational assortative mating in Brazil. Demography, 51(3), 835–856.

24.

Houseworth

Grayson

(2018). Intermarriage and the U.S. military. Armed Forces and Society, 45(4), 659–680.

25.

Hume

(1986). A treatise of human nature. Penguin. (Original work published 1739)

26.

James

(1890). The principles of psychology. Henry Holt.

27.

Kalmijn

(1993). Trends in Black/White Intermarriage. Social Forces, 72(1), 119–146.

28.

Kalmijn

(1998). Intermarriage and homogamy: Causes, patterns, trends. Annual Review of Sociology, 24, 395–421.

29.

Kane

(2006). Who are the recruits? The demographic characteristics of U.S. military enlistment, 2003–2005 (Data Analysis Report No. 06-09). Heritage Foundation Center.

30.

Kant

(2007). Critique of pure reason ( Weigelt

, Trans.). Penguin. (Original work published 1781)

31.

Lundquist

J. H.

(2004). When race makes no difference: Marriage and the military. Social Forces, 83(2), 731–757.

32.

Massey

Tannen

(2015). A research note on trends in black hypersegregation. Demography, 52(3), 1025–1034.

33.

Mead

G. H.

(1934). Mind, self and society. University of Chicago Press.

34.

Merton

R. K.

(1941). Intermarriage and the social structure: Fact and theory. Psychiatry, 4, 361–374.

35.

Mok

Morris

M. W.

Benet-Martínez

Karakitapoğlu-Aygün

(2007). Embracing American culture: Structures of social identity and social networks among first-generation biculturals. Journal of Cross-Cultural Psychology, 38(5), 629–635.

36.

Moskos

Butler

J. S.

(1996). All that we can be: Black leadership and racial integration the army way. The Twentieth Century Fund.

37.

Qian

(1997). Breaking the racial barriers: Variations in interracial marriage between 1980 and 1990. Demography, 34(2), 263–276.

38.

Reardon

S. F.

Fox

Townsend

(2015). Neighborhood income composition by household race and income, 1990–2009. The ANNALS of the American Academy of Political and Social Science, 660(1), 78–97.

39.

Rosenfeld

M. J.

(2005). A critique of exchange theory in mate selection. American Journal of Sociology, 110(5), 1284–1325.

40.

Rosenfeld

M. J.

Thomas

R. J.

(2012). Searching for a mate: The rise of the internet as a social intermediary. American Sociological Review, 77(4), 523–547.

41.

Schwartz

C. R.

Mare

R. D.

(2005). Trends in educational assortative marriage from 1940 to 2003. Demography, 44, 621–646.

42.

Schwartz

C. R.

Zeng

Xie

(2016). Marrying up by marrying down: Status exchange between social origin and education in the United States. Sociological Science, 3, 1003–1027.

43.

Segal

Thanner

Segal

(2007). Hispanic and African American men and women in the U.S. military: Trends in representation. Race, Gender & Class, 14(3–4), 48–64.

44.

Smith

J. A.

McPherson

Smith-Lovin

(2014). Social distance in the United States: Sex, race, religion, age, and education homophily among confidants, 1985 to 2004. American Sociological Review, 79(3), 432–456.

45.

Stryker

Serpe

R. T.

(1994). Identity salience and psychological centrality: Equivalent, overlapping, or complementary concepts? Social Psychology Quarterly, 57(1), 16–35.

46.

Stryker

. (1968). Identity salience and role performance: The relevance of symbolic interaction theory for family research. Journal of Marriage and the Family, 30, 558–564.

47.

Stryker

. (1980). Symbolic interactionism: A social structural version. Benjamin/Cummings.

48.

Testa

Kroghn

(1995). The effect of employment on marriage. In Tucker

M. B.

Mitchell-Kernan

(Eds.), The decline in marriage among African Americans: Causes, consequences, and policy implications (pp. 59–95). Russell Sage Foundation.

49.

Torche

Rich

(2017). Declining racial stratification in marriage choices? Trends in Black/White status exchange in the United States, 1980 to 2010. Sociology of Race and Ethnicity, 3(1), 31–49.

50.

Wasserstein

Schirm

A. L.

Lazar

N. A.

(2019). Moving to a world beyond “p<0.05.” The American Statistician, 73(suppl. 1), 1–19.

51.

Wei

Y. D.

Xiao

Simon

C. A.

Liu

Ni.

. (2018). Neighborhood, race and educational inequality. Cities, 73(1), 21–30.